{
"version": "https://jsonfeed.org/version/1.1",
"title": "Readwyn blog",
"home_page_url": "https://readwyn.com/blog/",
"feed_url": "https://readwyn.com/blog/latest.json",
"language": "en",
"items": [
{"id":"https://readwyn.com/blog/agatha-christie-public-domain/","url":"https://readwyn.com/blog/agatha-christie-public-domain/","title":"Which Agatha Christie books are in the public domain?","summary":"In the US, the Christie books published up to 1930 are free. In the UK, Canada and Australia, none are: her copyright there runs to the end of 2046.","content_html":"<p>In the United States, every Agatha Christie book first published in 1930 or earlier is in the public domain: fourteen novels and story collections on Project Gutenberg, from <em>The Mysterious Affair at Styles</em> (1920) to <em>The Murder at the Vicarage</em> (1930), plus a few early stories. In the UK, Canada and Australia, none of her books are. Christie <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Agatha_Christie\">died on 12 January 1976</a>, and all three countries now protect a book for 70 years after its author's death, so her work stays in copyright there until 1 January 2047.</p>\n<h2 id=\"the-us-list\">The US list</h2>\n<p>American law counts an older book from the year it came out, not from its author's death (the <a href=\"/blog/when-do-books-enter-the-public-domain/\">general rules for when a book becomes free</a> have a post of their own). The <a rel=\"external\" href=\"https://guides.library.cornell.edu/copyright/publicdomain\">copyright term chart of Cornell University Library</a> puts every work published before 1931 in the public domain, and gives a book first published abroad from 1931 to 1977 a term of 95 years from publication. A British first edition of the 1930s therefore counts the same as an American one. Each term runs to the end of its calendar year (<a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/305\">17 U.S.C. § 305</a>), so a book becomes free on 1 January. The years of first publication below are those of <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Author:Agatha_Christie\">Christie's author page on Wikisource</a>, and each title links to its Project Gutenberg edition.</p>\n<table><thead><tr><th>Published</th><th>Book</th><th>Free in the US since</th></tr></thead><tbody>\n<tr><td>1920</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/863\"><em>The Mysterious Affair at Styles</em></a></td><td>before 2019</td></tr>\n<tr><td>1922</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1155\"><em>The Secret Adversary</em></a></td><td>before 2019</td></tr>\n<tr><td>1923</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/58866\"><em>The Murder on the Links</em></a></td><td>1 January 2019</td></tr>\n<tr><td>1924</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/61262\"><em>Poirot Investigates</em></a></td><td>1 January 2020</td></tr>\n<tr><td>1924</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/61168\"><em>The Man in the Brown Suit</em></a></td><td>1 January 2020</td></tr>\n<tr><td>1925</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/65238\"><em>The Secret of Chimneys</em></a></td><td>1 January 2021</td></tr>\n<tr><td>1926</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/69087\"><em>The Murder of Roger Ackroyd</em></a></td><td>1 January 2022</td></tr>\n<tr><td>1927</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/70114\"><em>The Big Four</em></a></td><td>1 January 2023</td></tr>\n<tr><td>1928</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/72824\"><em>The Mystery of the Blue Train</em></a></td><td>1 January 2024</td></tr>\n<tr><td>1929</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/75288\"><em>The Seven Dials Mystery</em></a></td><td>1 January 2025</td></tr>\n<tr><td>1929</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/78342\"><em>Partners in Crime</em></a></td><td>1 January 2025</td></tr>\n<tr><td>1930</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/78220\"><em>The Murder at the Vicarage</em></a></td><td>1 January 2026</td></tr>\n<tr><td>1930</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/78084\"><em>The Mysterious Mr. Quin</em></a></td><td>1 January 2026</td></tr>\n<tr><td>1930</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/78262\"><em>Giant's Bread</em></a> (as Mary Westmacott)</td><td>1 January 2026</td></tr>\n</tbody></table>\n<p>Duke's Center for the Study of the Public Domain named <em>Roger Ackroyd</em> on its <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2022/\">list for 2022</a> and <em>The Murder at the Vicarage</em>, the first Miss Marple novel, on its <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2026/\">list for 2026</a>. The other dates follow from the same 95-year rule. Project Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/author/451\">Christie page</a> also has a few early Poirot stories as separate ebooks, among them \"The Missing Will\" and \"The Plymouth Express Affair\". <a href=\"/blog/standard-ebooks/\">Standard Ebooks</a> has edited editions of thirteen of them (not <em>The Mysterious Mr. Quin</em>). One early book is missing from both sites: <em>The Road of Dreams</em>, a <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Agatha_Christie_bibliography\">collection of her poems</a> from 1924, is free in the US by the same rule but has no Gutenberg edition.</p>\n<p><em>The Sittaford Mystery</em>, published in 1931, joins them on 1 January 2027: Standard Ebooks <a rel=\"external\" href=\"https://standardebooks.org/ebooks/agatha-christie/the-sittaford-mystery\">gives that date</a> on its page for the book. The famous titles are further off: the Standard Ebooks page for <a rel=\"external\" href=\"https://standardebooks.org/ebooks/agatha-christie/murder-on-the-orient-express\"><em>Murder on the Orient Express</em></a> gives 1 January 2030, as the book came out in 1934. By the same count, <em>And Then There Were None</em> (1939) follows on 1 January 2035.</p>\n<p>What becomes free is the text as first published. A modern edition's introduction, notes or cover is new work with its own copyright, and so is a film or a stage version. Agatha Christie Limited, the company Christie <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Agatha_Christie\">set up in 1955</a> to hold her rights, still owns the rights to more than 80 of her novels and short stories and to 19 plays: everywhere for the later books, and outside the US for the early ones.</p>\n<h2 id=\"the-uk-canada-and-australia\">The UK, Canada and Australia</h2>\n<p>Here the year of publication doesn't matter, only the year of death (the <a href=\"/blog/public-domain-differs-by-country/\">differences between countries</a> have a post of their own). The UK government's guidance on <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">how long copyright lasts</a> gives a written work 70 years after the author's death. Australia's <a rel=\"external\" href=\"https://www.legislation.gov.au/C1968A00063/latest/text\">Copyright Act 1968</a>, section 33, ends the copyright of a work made public in the author's lifetime \"70 years after the calendar year in which the author of the work died\". For Christie, who died in 1976, the Australian term runs to the end of 2046, so her books become free there on 1 January 2047. The UK page names no last day, but <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Agatha_Christie\">Wikipedia</a> gives the start of 2047 for every country with a term of life plus 70 years.</p>\n<p>Canada is the near miss. Until 2022 Canada protected a book for the author's life plus 50 years, which would have freed all of Christie on 1 January 2027. The term became 70 years on 30 December 2022 (<a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/acts/C-42/section-6.html\">section 6 of the Copyright Act</a>, in force by <a rel=\"external\" href=\"https://orders-in-council.canada.ca/attachment.php?attach=42930\">order in council</a>). The change revived no copyright that had already run out, but Christie's still had four years to go, so Canadian readers now wait for 2047 with the British and the Australians.</p>\n<p>That leaves a reader in the UK, Canada or Australia with no free Christie at all, even on a site that lists her. Project Gutenberg marks its <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/78220\"><em>Murder at the Vicarage</em></a> \"public domain in the USA\", and its <a rel=\"external\" href=\"https://www.gutenberg.org/policy/license.html\">licence</a> asks readers in other countries to check their own law first.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/anne-of-green-gables-books-in-order/","url":"https://readwyn.com/blog/anne-of-green-gables-books-in-order/","title":"The Anne of Green Gables books in order","summary":"Story order runs Green Gables, Avonlea, the Island, Windy Poplars, House of Dreams, Ingleside, Rainbow Valley, Rilla. Two of the eight came out in the 1930s.","content_html":"<p>Read L. M. Montgomery's Anne books in the order of Anne's life: <em>Anne of Green Gables</em>, <em>Anne of Avonlea</em>, <em>Anne of the Island</em>, <em>Anne of Windy Poplars</em>, <em>Anne's House of Dreams</em>, <em>Anne of Ingleside</em>, <em>Rainbow Valley</em>, and <em>Rilla of Ingleside</em>. That is also the numbering of the series table on Wikipedia, which adds <em>The Blythes Are Quoted</em> as a ninth, and of Faded Page, which labels <em>Windy Poplars</em> book 4 and <em>Ingleside</em> book 6. It isn't the order they were written in: <em>Rilla</em> came out in 1921. Montgomery had tired of Anne, and she came back to her after a 15-year break, writing <em>Windy Poplars</em> (1936) and <em>Ingleside</em> (1939) to fill gaps she had left in Anne's life, as the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Lucy_Maud_Montgomery\">Wikipedia article on Montgomery</a> records.</p>\n<h2 id=\"the-eight-novels\">The eight novels</h2>\n<p>Anne's ages come from the “Timeline year” column of the series table in the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Anne_of_Green_Gables\">Wikipedia article on <em>Anne of Green Gables</em></a>. The word counts come from the Project Gutenberg text of each book, counted the same way as in <a href=\"/blog/how-many-words-in-a-novel/\">how many words are in a novel</a>, and the reading time assumes 250 words per minute, the <a href=\"/blog/reading-speed-words-per-minute/\">starting speed Readwyn uses</a> before it has measured yours.</p>\n<table><thead><tr><th>No.</th><th>Book</th><th>Published</th><th>Anne's age</th><th>Words</th><th>At 250 wpm</th></tr></thead><tbody>\n<tr><td>1</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/45\">Anne of Green Gables</a></td><td>1908</td><td>11 to 16</td><td>103,200</td><td>about 7 h</td></tr>\n<tr><td>2</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/47\">Anne of Avonlea</a></td><td>1909</td><td>16 to 18</td><td>91,100</td><td>about 6 h</td></tr>\n<tr><td>3</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/51\">Anne of the Island</a></td><td>1915</td><td>18 to 22</td><td>77,300</td><td>about 5 h</td></tr>\n<tr><td>4</td><td>Anne of Windy Poplars</td><td>1936</td><td>22 to 25</td><td>not counted</td><td></td></tr>\n<tr><td>5</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/544\">Anne's House of Dreams</a></td><td>1917</td><td>25 to 27</td><td>80,400</td><td>about 5½ h</td></tr>\n<tr><td>6</td><td>Anne of Ingleside</td><td>1939</td><td>34 to 40</td><td>not counted</td><td></td></tr>\n<tr><td>7</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/5343\">Rainbow Valley</a></td><td>1919</td><td>41 to 43</td><td>81,700</td><td>about 5½ h</td></tr>\n<tr><td>8</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/3796\">Rilla of Ingleside</a></td><td>1921</td><td>49 to 53</td><td>105,100</td><td>about 7 h</td></tr>\n</tbody></table>\n<p>The six books on Gutenberg come to 538,800 words, or about 36 hours at that pace. In the last two, Anne steps back and the story turns to her children and their friends.</p>\n<h2 id=\"story-order-or-publication-order\">Story order or publication order</h2>\n<p>Read the ages in publication order and you see why Wikipedia and Faded Page number by story: Anne is 22 at the end of <em>Anne of the Island</em> and 25 at the start of <em>Anne's House of Dreams</em>, the three years that <em>Windy Poplars</em> later filled, and she jumps from 27 to 41 before <em>Rainbow Valley</em>, with <em>Ingleside</em> (34 to 40) written later to go between. Publication order still has a case. It's the order in which Montgomery's first readers met the books, with <em>Anne's House of Dreams</em> following straight on from <em>Anne of the Island</em>, and it puts the two books she wrote in her sixties after the six that came out from 1908 to 1921. Neither order is wrong. The <a href=\"/blog/find-the-reading-order-of-a-series/\">general advice on reading orders</a> applies here too: decide which you want before you number your copies.</p>\n<p>One more wrinkle for <em>Windy Poplars</em>: Montgomery's American publisher asked for cuts, and the edited text came out as <em>Anne of Windy Poplars</em> in the US and Canada. Her British publisher printed the uncut text under her own title, <em>Anne of Windy Willows</em>, according to the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Anne_of_Windy_Poplars\">Wikipedia article on the book</a>. Check which one you are getting.</p>\n<h2 id=\"which-ones-are-free\">Which ones are free</h2>\n<p>Everything published in the US before 1931 is <a href=\"/blog/when-do-books-enter-the-public-domain/\">now in the American public domain</a>, so the first six books published, up to <em>Rilla</em> in 1921, are free on Project Gutenberg. <a href=\"/blog/standard-ebooks/\">Standard Ebooks</a> has its own editions of the first three, <em>Green Gables</em>, <em>Avonlea</em>, and <em>the Island</em>; the other three are listed there but not yet produced. The two late novels are a different matter, because the US counts from publication, not from the author's death. Standard Ebooks' placeholder pages give the dates: <em>Anne of Windy Poplars</em> <a rel=\"external\" href=\"https://standardebooks.org/ebooks/l-m-montgomery/anne-of-windy-poplars\">enters the US public domain</a> on 1 January 2032, and <em>Anne of Ingleside</em> <a rel=\"external\" href=\"https://standardebooks.org/ebooks/l-m-montgomery/anne-of-ingleside\">on 1 January 2035</a>.</p>\n<p>Outside the US the picture changes. Montgomery died on 24 April 1942, and in the UK a written work is protected for <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">70 years after the author's death</a>, so the books she published in her lifetime have been free there since the start of 2013. Faded Page offers both late novels as public domain in Canada: <a rel=\"external\" href=\"https://www.fadedpage.com/showbook.php?pid=20100608\"><em>Anne of Windy Poplars</em></a> and <a rel=\"external\" href=\"https://fadedpage.com/showbook.php?pid=20260315\"><em>Anne of Ingleside</em></a>, each as EPUB, MOBI, PDF, and plain text. Project Gutenberg Australia also has <a rel=\"external\" href=\"https://gutenberg.net.au/ebooks01/0100281h.html\"><em>Anne of Ingleside</em></a>. Each of these sites follows the law of its own country, so check yours before you download; <a href=\"/blog/public-domain-differs-by-country/\">public domain differs by country</a> explains why.</p>\n<h2 id=\"the-books-around-the-eight\">The books around the eight</h2>\n<ul>\n<li><em><strong>Chronicles of Avonlea</strong></em> (1912) is twelve stories set in Anne's village. Anne stars in only one, \"The Hurrying of Ludovic\", and turns up in a few others (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Chronicles_of_Avonlea\">Wikipedia</a>). It's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1354\">on Gutenberg</a>: 66,600 words.</li>\n<li><em><strong>Further Chronicles of Avonlea</strong></em> (1920) was put together from stories Montgomery had left out of the first collection and published without her permission. She sued her publisher, L. C. Page, and won after nearly nine years, by the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Further_Chronicles_of_Avonlea\">Wikipedia account</a>. Anne narrates one story, \"The Little Brown Book of Miss Emily\". It's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/5340\">on Gutenberg</a>: 71,000 words.</li>\n<li><em><strong>The Blythes Are Quoted</strong></em> is the book Montgomery finished near the end of her life: stories, poems credited to Anne and her son Walter, and scenes of the family talking about them. Her publisher received it on the day she died and didn't publish it. A shortened version, <em>The Road to Yesterday</em>, appeared in 1974, and the full text only in 2009, edited by Benjamin Lefebvre (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Blythes_Are_Quoted\">Wikipedia</a>). Its stories reach past the end of <em>Rilla</em> to Anne in old age, so read it last.</li>\n</ul>\n<p>If you keep the series in Readwyn, open each book's “Edit details…” (Cmd+E) and give it its number beside the series. The page of the series then shows an empty place, “Not in this library”, for any number you don't have yet, which makes the gap at 4 or 6 easy to spot if the late novels aren't free where you live.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/books-a-year-at-30-minutes-a-day/","url":"https://readwyn.com/blog/books-a-year-at-30-minutes-a-day/","title":"How many books can you read in a year at 30 minutes a day?","summary":"Half an hour a day is about 2.7 million words a year at 250 words a minute: some 22 books the length of Pride and Prejudice, or 56 as short as The Great Gatsby.","content_html":"<p>Thirty minutes a day comes to 182½ hours over a year of 365 days. At 250 words a minute that's about 2.7 million words, which is roughly 22 novels the length of <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/42671\">Pride and Prejudice</a>, or 56 as short as <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/64317\">The Great Gatsby</a>. At 350 words a minute the same half hour covers 3.8 million words, and the count of Pride and Prejudices rises to 31.</p>\n<p>The answer depends far more on the length of the books than on your speed. Going from 250 to 350 words a minute adds nine books the length of Pride and Prejudice, while choosing Gatsby over War and Peace moves the count from 4 to 56.</p>\n<h2 id=\"six-books-three-speeds\">Six books, three speeds</h2>\n<p>The word counts are those of the Project Gutenberg texts, counted the same way as in <a href=\"/blog/how-many-words-in-a-novel/\">how many words are in a novel</a>. The speeds are the ones that <a href=\"/blog/reading-speed-words-per-minute/\">reading speed in words per minute</a> uses: 250 words a minute is the starting figure there, and 300 and 350 come from its examples. They're figures for working the sums, not averages from a survey of readers.</p>\n<table><thead><tr><th>Book</th><th>Words</th><th>At 250 wpm</th><th>At 300 wpm</th><th>At 350 wpm</th></tr></thead><tbody>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/64317\">The Great Gatsby</a></td><td>48,600</td><td>56</td><td>67</td><td>78</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/84\">Frankenstein</a></td><td>75,200</td><td>36</td><td>43</td><td>50</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/42671\">Pride and Prejudice</a></td><td>122,500</td><td>22</td><td>26</td><td>31</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1260\">Jane Eyre</a></td><td>187,200</td><td>14</td><td>17</td><td>20</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/145\">Middlemarch</a></td><td>318,600</td><td>8</td><td>10</td><td>12</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2600\">War and Peace</a>, translated by Louise and Aylmer Maude</td><td>565,400</td><td>4</td><td>5</td><td>6</td></tr>\n</tbody></table>\n<p>Each figure is the number of whole books that fit in a year of 30 minutes a day, rounded down. A mixed shelf lands somewhere between the rows: at 250 words a minute, three Gatsbys for every Middlemarch comes to 23 books in the year, against 14 Jane Eyres.</p>\n<h2 id=\"one-book-at-a-time\">One book at a time</h2>\n<p>Turned the other way, the table gives the days that one book takes at half an hour a day. At 250 words a minute, Gatsby takes 6½ days and Frankenstein 10. Pride and Prejudice takes a little over 16, and Jane Eyre 25. Middlemarch needs about six weeks and War and Peace about 75 days, close to two and a half months.</p>\n<p>At 350 words a minute, every one of those shrinks by more than a quarter: Pride and Prejudice comes down to under 12 days, and War and Peace to about 54.</p>\n<h2 id=\"when-you-miss-days\">When you miss days</h2>\n<p>Half an hour on five days a week rather than seven makes 260 sessions a year, 130 hours in all. At 250 words a minute that's about 1.95 million words: nearly 16 books the length of Pride and Prejudice, or 40 the length of Gatsby.</p>\n<p>Speed can make up the gap. Five days a week at 350 words a minute reads about the same number of words as seven days a week at 250 (2.73 million against 2.74 million), so a faster reader who skips weekends keeps level with a slower one who never misses a day.</p>\n<h2 id=\"a-book-a-week\">A book a week</h2>\n<p>Fifty-two books in a year means one a week. At 30 minutes a day, that works only for short books: Gatsby needs 28 minutes a day at 250 words a minute to finish in a week. Frankenstein needs 43 minutes, and Pride and Prejudice 70. At 350 words a minute, Pride and Prejudice comes down to 50 minutes a day, which is still well over half an hour.</p>\n<h2 id=\"against-the-average\">Against the average</h2>\n<p>Half an hour a day adds up to more books than the polls report, even though the polls count a book that was only started. In <a rel=\"external\" href=\"https://news.gallup.com/poll/388541/americans-reading-fewer-books-past.aspx\">Gallup's poll of December 2021</a>, Americans said they had read an average of 12.6 books, all or part of the way through, in the past year, and 27% had read more than 10. Even at 250 words a minute, a year of daily half hours gets through 14 books as long as Jane Eyre. The averages of the other polls, and why the median is so much lower, are in <a href=\"/blog/how-many-books-does-the-average-person-read/\">how many books the average person reads in a year</a>.</p>\n<h2 id=\"your-own-numbers\">Your own numbers</h2>\n<p>These sums assume a speed that you hold for the whole half hour, and your speed varies from book to book. If you read in Readwyn, the Stats screen measures it from your own page turns: “Reading speed” gives your words per minute for the range you pick at the top of the screen, and “Books finished”, with the range set to Year, the count for the year. With those two figures, the sums above work for your own reading.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/borrow-a-book-from-the-internet-archive/","url":"https://readwyn.com/blog/borrow-a-book-from-the-internet-archive/","title":"How to borrow a book from the Internet Archive","summary":"Sign up for a free archive.org account, open the book's page, and click Borrow this Book. You read a 1-hour loan in the browser; a 14-day loan can also be downloaded for an LCP reader.","content_html":"<p>You need a free archive.org account and a web browser. Sign in, find the book, open its page, and click “Borrow this Book”. If every copy is out on a 14-day loan, you get “Join Waitlist” instead. The book then opens in BookReader, the Archive's reader in the browser, or, for some loans, as a file that you download and open in a separate app. All of this comes from the Archive's help page <a rel=\"external\" href=\"https://help.archive.org/help/borrowing-from-the-lending-library/\">Borrowing From The Lending Library</a>, which also covers loans made through OpenLibrary.org.</p>\n<h2 id=\"finding-a-book-you-can-borrow\">Finding a book you can borrow</h2>\n<p>The help page gives several ways in: the text icon at the top left of the black bar on archive.org, the “Books to Borrow” link, <a href=\"/blog/what-is-open-library/\">Open Library</a> with its suggested books and categories, the featured texts at the top of the page, or the Search bar, where a title or an author's name is enough. A book that you can borrow has the borrow button on its item page.</p>\n<p>Plenty of books show “Borrow Unavailable”. The Archive <a rel=\"external\" href=\"https://help.archive.org/help/why-are-so-many-books-listed-as-borrow-unavailable-at-the-internet-archive/\">counts more than 500,000</a> that it took out of lending because of Hachette v. Internet Archive, the publishers' suit against its scans of in-copyright books, and of removal requests from publishers that the suit did not name. The same page says that readers with print disabilities can still get those books through its accessibility programme. How the case got there is the subject of <a href=\"/blog/controlled-digital-lending/\">What is controlled digital lending</a>.</p>\n<h2 id=\"one-hour-or-fourteen-days\">One hour or fourteen days</h2>\n<p>The Archive lends a book for one hour or for fourteen days, and the length decides almost everything else.</p>\n<table><thead><tr><th></th><th>1-hour loan</th><th>14-day loan</th></tr></thead><tbody>\n<tr><td>Where you read it</td><td>BookReader, in the browser</td><td>BookReader, or a downloaded PDF or EPUB</td></tr>\n<tr><td>Waitlist</td><td>None</td><td>Yes, if every copy is out</td></tr>\n<tr><td>When it ends</td><td>Borrow it again if a copy is free</td><td>Borrow it again; no extension</td></tr>\n</tbody></table>\n<p>Most books have the 1-hour loan, and some have the 14-day loan as well. An hour is rarely enough for a whole book, so the help page's answer is to check the book out again when the hour runs out, as long as a copy is free. A 14-day loan can't be renewed: when it ends you borrow the book again, and if others are waiting, you join the queue behind them. The Archive emails you when your copy is ready.</p>\n<p>You can hold up to 10 books at a time. There are no late fees. A loan stops working when its period ends, and nothing else happens.</p>\n<h2 id=\"reading-the-download\">Reading the download</h2>\n<p>For a 14-day loan you pick the format at checkout, and the help page warns that not every book has every format. The PDF and EPUB files are protected by LCP, a form of DRM, so an ordinary ebook app won't open them. The page names Thorium on a computer and Cantook on iOS and Android. <a rel=\"external\" href=\"https://thorium.edrlab.org/en/\">Thorium</a> is free and open source, made by EDRLab, and runs on Mac, Windows and Linux. You don't create another account for it: the first time you open a borrowed book, the app asks for the email address and password that you use on archive.org.</p>\n<p>One step of the same help page still tells you to open downloaded formats in Adobe Digital Editions. Every other answer on the page names LCP readers, so start with Thorium on a computer or Cantook on a phone or tablet.</p>\n<h2 id=\"on-a-kindle-or-another-e-reader\">On a Kindle or another e-reader</h2>\n<p>A borrowed book doesn't go to a Kindle. The help page says that Kindle uses its own proprietary DRM, and it gives no way to read a borrowed LCP file on one. Books without restrictions are a different matter: if an item offers a PDF or MOBI file, you can download it and copy it to the Kindle as you would any other file.</p>\n<p>For a Nook, the page says to install an LCP reader such as Cantook on the device. It names no route for a Kobo. For every e-reader, its answer is the same fallback: read the book online in BookReader, in your browser.</p>\n<h2 id=\"seeing-and-returning-your-loans\">Seeing and returning your loans</h2>\n<p>Click your username, then “My Loans”. You return a BookReader loan from that page. You can return a downloaded LCP book early too, but only from archive.org: open it from your loans, and choose “Return” in BookReader. Thorium then shows “Loan Expired” next to the book and won't open it, and the three dots beside it give you “Delete”. If you don't return a book early, the loan ends by itself when its period runs out.</p>\n<p>For Kobo and Kindle owners whose public library lends ebooks through OverDrive, there's a second free route, covered in <a href=\"/blog/library-ebooks-on-kobo-and-kindle/\">How to read library ebooks on a Kobo or a Kindle</a>.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/free-opds-catalogues/","url":"https://readwyn.com/blog/free-opds-catalogues/","title":"Free OPDS catalogues to add to a reader app","summary":"Project Gutenberg, Wikisource, Unglue.it, and ManyBooks run free OPDS catalogues that any reader app can open without an account. Standard Ebooks asks most apps for a patron login.","content_html":"<p>In English, four general catalogues in the common OPDS 1 format open in a reader app with no account at all: Project Gutenberg, English Wikisource, Unglue.it, and ManyBooks. Paste one of the addresses from the table below into the app's list of catalogues and you can browse and download books on the device. Gutenberg's and ManyBooks' have a search; the other two are lists to browse. <a rel=\"external\" href=\"https://theanarchistlibrary.org/\">The Anarchist Library</a> is a fifth, for anarchist texts only. The Internet Archive has a much larger catalogue, but in the newer OPDS 2 format, which fewer apps read. Standard Ebooks also has an OPDS catalogue, but it asks most apps for a login that only its patrons, the people who produced one of its ebooks, and its corporate sponsors have.</p>\n<p>Each address below loaded as an OPDS feed when it was checked on 11 November 2026.</p>\n<table><thead><tr><th>Catalogue</th><th>Address</th><th>What is in it</th></tr></thead><tbody>\n<tr><td>Project Gutenberg</td><td><code>https://www.gutenberg.org/ebooks.opds/</code></td><td>Public-domain books</td></tr>\n<tr><td>Wikisource (English)</td><td><code>https://ws-export.wmcloud.org/opds/en/Ready_for_export.xml</code></td><td>Transcribed works that Wikisource marks ready for export</td></tr>\n<tr><td>Unglue.it</td><td><code>https://unglue.it/api/opds/</code></td><td>Free ebooks, many under Creative Commons licences</td></tr>\n<tr><td>ManyBooks</td><td><code>https://manybooks.net/opds</code></td><td>Free modern and classic ebooks</td></tr>\n<tr><td>The Anarchist Library</td><td><code>https://theanarchistlibrary.org/opds</code></td><td>An archive of anarchist texts</td></tr>\n<tr><td>Wolne Lektury</td><td><code>https://wolnelektury.pl/opds/</code></td><td>Polish literature, mostly public domain</td></tr>\n<tr><td>textos.info</td><td><code>https://textos.info/opds</code></td><td>Spanish-language books, free to download</td></tr>\n<tr><td>Chitanka</td><td><code>https://chitanka.info/catalog.opds</code></td><td>Bulgarian books, with lists by author, translator, and series</td></tr>\n</tbody></table>\n<h2 id=\"project-gutenberg\">Project Gutenberg</h2>\n<p>Gutenberg's root feed is short. It has three lists, Popular, Latest, and Random, and a search box that the app fills from Gutenberg's own <a rel=\"external\" href=\"https://www.gutenberg.org/catalog/osd-books.xml\">search description</a>. The <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/offline_catalogs.html\">offline catalogues page</a> tells apps to start at <code>https://www.gutenberg.org/ebooks/search.opds/</code> instead, which lists every book, 25 to a page. Both addresses work.</p>\n<p>Open a book and the entry offers several files. For Pride and Prejudice, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1342.opds\">the entry</a> has an EPUB 3 with images (labelled for e-readers and Send to Kindle), an older EPUB, a KF8 file for Kindles, and a MOBI file for older Kindles, plus versions of some of these without images. An app that reads EPUB should take the EPUB 3.</p>\n<p>Two things on that offline catalogues page are worth knowing. Gutenberg says its OPDS feeds are \"primarily intended for machine-to-machine communication\", so expect plain lists rather than a shop window. It also plans to retire the current XML feeds in 2027, in favour of a JSON feed in the newer OPDS 2 format that is still being tested. An address that works today may stop working then, so check that page if Gutenberg goes quiet in your app.</p>\n<p>There's more on the project itself in <a href=\"/blog/what-is-project-gutenberg/\">what Project Gutenberg is</a>.</p>\n<h2 id=\"wikisource\">Wikisource</h2>\n<p>Wikisource's catalogue is one long list: every work in the category \"Ready for export\", about 700 of them, in a single feed of about 1 MB with no search and no lists by author or subject. <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Help:OPDS\">Its help page</a> says the catalogue is rebuilt every day, and that other categories can have catalogues of their own on request. The post on <a href=\"/blog/what-is-wikisource/\">Wikisource</a> explains how those transcriptions are made. The server puts a bot check in front of requests that look like they come from a web browser, so if an app shows an error instead of the list, that check is the likely cause.</p>\n<h2 id=\"unglue-it\">Unglue.it</h2>\n<p>Unglue.it is a site for books whose creators have set them free under Creative Commons licences (<a rel=\"external\" href=\"https://unglue.it/about/\">its about page</a>). Its root has lists of popular and new free ebooks, a list of Creative Commons books, lists by subject (fiction, history, science, software, and more), and a list by format, so <code>https://unglue.it/api/opds/epub/</code> shows only the books that come as EPUB. It also lists the books that are still raising money to be made free.</p>\n<h2 id=\"manybooks\">ManyBooks</h2>\n<p>ManyBooks describes itself as a selection of modern and classic books, all free (<a rel=\"external\" href=\"https://manybooks.net/\">its home page</a>). Its root has lists of new titles, authors, titles, and genres, and a search. The books in it are not all old: the new titles in November 2026 included books published that year, and the one checked downloaded as an EPUB without an account.</p>\n<h2 id=\"internet-archive\">Internet Archive</h2>\n<p>The Internet Archive's catalogue at <code>https://archive.org/services/opds</code> is in OPDS 2, a JSON format, so it opens only in an app that reads OPDS 2. Thorium does: <a rel=\"external\" href=\"https://thorium.edrlab.org/en/docs/220_organizing/222_catalogs/\">its guide</a> says it reads OPDS 1 and 2 catalogues, and names the Archive's feed as one to browse and borrow from. The root has a search, and a group of open-access and public-domain ebooks that counted 3,667,217 items on 11 November 2026. Borrowing a book needs an archive.org account, as the guide also says; the steps are in <a href=\"/blog/borrow-a-book-from-the-internet-archive/\">borrowing from the Internet Archive</a>.</p>\n<h2 id=\"catalogues-in-other-languages\">Catalogues in other languages</h2>\n<p>Wolne Lektury is run by a Polish foundation and has more than 7,000 works. Its <a rel=\"external\" href=\"https://wolnelektury.pl/info/o-projekcie/\">about page</a> says that most of them are in the public domain and the rest are under free licences, and that nothing needs registration. The catalogue browses by author, kind, genre, and literary period. <a rel=\"external\" href=\"https://textos.info/\">textos.info</a> calls itself an open, free digital library of more than 6,000 books in Spanish, and its feed browses by author, title, and tag. <a rel=\"external\" href=\"https://chitanka.info/\">Chitanka</a>'s catalogue is in Bulgarian.</p>\n<p>The <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/OPDS\">MobileRead wiki</a> keeps a longer list of catalogues, grouped by language, with some stores among them. Several of its entries are marked as dead, so expect some addresses there to fail.</p>\n<h2 id=\"standard-ebooks\">Standard Ebooks</h2>\n<p>Standard Ebooks publishes careful new editions of public-domain books (<a href=\"/blog/standard-ebooks/\">more about them</a>), and its <a rel=\"external\" href=\"https://standardebooks.org/feeds\">feeds page</a> lists an OPDS catalogue at <code>https://standardebooks.org/feeds/opds</code>. Access to it is \"a benefit of membership in our Patrons Circle\", the page says, along with people who have produced a Standard Ebooks book and corporate sponsors. Without a login, the address answered most apps with HTTP 401 (Unauthorized) when it was checked on 11 November 2026, and an app then asks for a user name and password. The exception was a request that named KOReader or Readest as the app: the server sent the whole catalogue, with its download links, and asked for no login. A patron enters the email address and leaves the password blank. The free way to their books is the website, and the New Releases RSS and Atom feeds, which the same page says are open to everyone.</p>\n<h2 id=\"adding-a-catalogue\">Adding a catalogue</h2>\n<p>In KOReader, the <a rel=\"external\" href=\"https://github.com/koreader/koreader/wiki/OPDS-support\">OPDS page of its wiki</a> says to open the catalogues from the search (magnifying glass) menu of the file browser, not from inside a book, and then add a new one with the button in the top left corner. In Thorium Reader, <a rel=\"external\" href=\"https://thorium.edrlab.org/en/docs/220_organizing/222_catalogs/\">its guide</a> says to open the “Catalogs” tab of the library, choose “+ Add an OPDS feed”, give the catalogue a name, paste the address, and confirm with “Add”.</p>\n<p>A catalogue of your own books works the same way. Readwyn serves your library as an OPDS 1.2 catalogue on your home network, with a user name and password for each app (<a href=\"/blog/what-is-opds/\">what OPDS is</a>, and <a href=\"/blog/opds-catalogue-for-koreader/\">the steps in KOReader</a>). A book downloaded from Gutenberg can then sit in the same app as the rest of your library.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/hobbit-and-lord-of-the-rings-public-domain/","url":"https://readwyn.com/blog/hobbit-and-lord-of-the-rings-public-domain/","title":"When do The Hobbit and The Lord of the Rings enter the public domain?","summary":"In the UK and the EU, on 1 January 2044. In the US, the 1937 Hobbit comes free on 1 January 2033 and The Lord of the Rings in 2050 and 2051.","content_html":"<p>J. R. R. Tolkien <a rel=\"external\" href=\"https://www.tolkienestate.com/life/biography/\">died on 2 September 1973</a>, so in the UK, the EU and every other country that protects a book for its author's life plus 70 years, <em>The Hobbit</em> and <em>The Lord of the Rings</em> both come out of copyright on 1 January 2044. The United States counts from publication instead, which gives each book its own date: 1 January 2033 for the 1937 text of <em>The Hobbit</em>, and 2050 or 2051 for the three volumes of <em>The Lord of the Rings</em>. In countries that protect a book for only 50 years after its author's death, New Zealand among them, they're free already.</p>\n<table><thead><tr><th>Where</th><th><em>The Hobbit</em></th><th><em>The Lord of the Rings</em></th></tr></thead><tbody>\n<tr><td>UK, the EU, other life plus 70 countries</td><td>1 January 2044</td><td>1 January 2044</td></tr>\n<tr><td>Canada</td><td>1 January 2044</td><td>1 January 2044</td></tr>\n<tr><td>United States</td><td>1 January 2033 (the 1937 text)</td><td>1 January 2050 for the first two volumes, 1 January 2051 for the third (the first-edition text)</td></tr>\n<tr><td>Life plus 50 countries</td><td>Free since 1 January 2024</td><td>Free since 1 January 2024</td></tr>\n</tbody></table>\n<h2 id=\"britain-the-eu-and-other-life-plus-70-countries\">Britain, the EU and other life plus 70 countries</h2>\n<p>The UK government's page on <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">how long copyright lasts</a> gives written works 70 years after the author's death, and the EU sets the same term in <a rel=\"external\" href=\"https://www.wipo.int/wipolex/en/text/443261\">Directive 2006/116/EC</a>, which counts it from 1 January of the year after the death. Seventy years on from 1973 is the end of 2043. The pictures in the first <em>Hobbit</em> were Tolkien's own: he <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Hobbit\">drew the illustrations and the dust jacket</a> and <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Tolkien%27s_maps\">the maps</a>. The UK page puts artistic works in the same category, so his pictures come free on the same day as his words.</p>\n<h2 id=\"the-united-states-counts-95-years-from-publication\">The United States counts 95 years from publication</h2>\n<p>US law protects a book published from 1931 to 1977 for 95 years from its publication if it kept its copyright, and <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/305\">17 U.S.C. § 305</a> runs every term to the end of the calendar year. The <a rel=\"external\" href=\"https://guides.library.cornell.edu/copyright/publicdomain\">copyright term chart of Cornell University Library</a> gives a foreign book of 1931 to 1977 that same 95 years even if its publisher missed the American formalities, provided the book was still in copyright at home on 1 January 1996. Tolkien's were, so the British first editions count. George Allen &amp; Unwin published <em>The Hobbit</em> in 1937, and <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Lord_of_the_Rings\"><em>The Lord of the Rings</em></a> on 29 July 1954, 11 November 1954 and 20 October 1955. That puts <em>The Fellowship of the Ring</em> and <em>The Two Towers</em> in the US public domain on 1 January 2050, and <em>The Return of the King</em> a year later.</p>\n<p>Without the extension of the term that Congress passed in 1998, <em>The Hobbit</em> would have been free in the US in 2013. The blog Everybody's Libraries marked <a rel=\"external\" href=\"https://everybodyslibraries.com/2013/01/01/public-domain-day-2013-or-there-and-back-again/\">Public Domain Day 2013</a> by naming it among the 1937 works that Americans would otherwise have received that morning.</p>\n<h2 id=\"which-hobbit-comes-free\">Which Hobbit comes free</h2>\n<p>The text of 1937 is not the one that later editions print. For the second edition of 1951, <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Hobbit\">Tolkien rewrote \"Riddles in the Dark\"</a> to fit the Ring of the later books: in the first edition Gollum stakes his ring on the riddle game and the two part on good terms, while in the revision Gollum ends the chapter cursing Bilbo as a thief. A third edition followed in 1966.</p>\n<p>That revision matters only in the US. A later version carries a copyright of its own, but Duke's Center for the Study of the Public Domain <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2026/\">points out</a> that the rights extend only to the newly added material. On 1 January 2033 Americans get the 1937 text. The changes of 1951 have their own 95 years, which run to the end of 2046. The changes that Tolkien made in the 1960s have terms of their own too: those of the second edition of <em>The Lord of the Rings</em> (1965) run to the end of 2060, and those of the third edition of <em>The Hobbit</em> (1966) to the end of 2061. A US reader who wants the books in 2033 and 2050 gets the earlier texts. Under a life plus 70 term the question doesn't come up for the text: every edition is Tolkien's writing, and all of it comes free in 2044. The maps of <em>The Lord of the Rings</em> are another matter. <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Christopher_Tolkien\">Christopher Tolkien drew them</a>, and he died in 2020, so in life plus 70 countries his maps stay in copyright until the end of 2090.</p>\n<p>The 1966 edition had a reason behind it. In 1965 Ace Books brought out an American paperback of <em>The Lord of the Rings</em> without Tolkien's permission, because it <a rel=\"external\" href=\"https://blogs.loc.gov/law/2014/11/j-r-r-tolkien-paperbacks-and-copyright/\">believed the US copyright had expired</a>, as a post on the Law Library of Congress blog recounts. Tolkien revised <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Lord_of_the_Rings\"><em>The Lord of the Rings</em></a> and <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Hobbit\"><em>The Hobbit</em></a> so that the new texts could be protected, and authorised paperbacks from Ballantine followed. Under pressure from Tolkien's readers, <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Lord_of_the_Rings\">Ace withdrew its edition</a> and made him a small payment.</p>\n<h2 id=\"where-the-books-are-already-free\">Where the books are already free</h2>\n<p>Some countries protect a book for only 50 years after its author's death, the minimum that <a rel=\"external\" href=\"https://www.wipo.int/wipolex/en/text/283698\">Article 7 of the Berne Convention</a> sets. In those countries Tolkien's books came free on 1 January 2024, and the <a rel=\"external\" href=\"https://publicdomainreview.org/features/entering-the-public-domain/2024/\">Public Domain Review's list for that day</a> includes him. It names New Zealand and most of Africa and Asia as places with that term.</p>\n<p>Canada was one of them until the end of 2022. Its term became life plus 70 on <a rel=\"external\" href=\"https://orders-in-council.canada.ca/attachment.php?attach=42930\">30 December 2022</a>, a year before Tolkien's Canadian copyright would have run out. The change <a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/AnnualStatutes/2022_10/page-21.html\">revived no copyright that had already expired</a>, but Tolkien's had not, so Canadians wait until 2044 with Britain (<a href=\"/blog/public-domain-differs-by-country/\">the post on terms by country</a> gives the whole story of the change).</p>\n<p>A copy that is free in New Zealand is not free to share in London or New York. As Duke's page notes, you're governed by the law of the place where you use a work.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-long-is-war-and-peace/","url":"https://readwyn.com/blog/how-long-is-war-and-peace/","title":"How long is War and Peace?","summary":"The Maude translation of War and Peace is about 565,400 words: some 38 hours at 250 words a minute, in 15 books, two epilogues and 365 chapters.","content_html":"<p>Project Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2600\">War and Peace</a>, in the English translation by Louise and Aylmer Maude, runs to about 565,400 words. At 250 words a minute that is 37 hours and 42 minutes of reading; at 300 it's about 31½ hours, and at 350 about 27. The text comes in fifteen books and two epilogues, split into 365 chapters, so a chapter averages about 1,550 words, or six minutes at the slowest of those speeds.</p>\n<p>The count is the one in <a href=\"/blog/how-many-words-in-a-novel/\">how many words are in a novel</a>, made the same way: everything between Gutenberg's start and end lines, split at spaces and dashes, rounded to the nearest hundred. About 800 of those words are the title and the contents list. Another translation, or a count that splits at spaces only, gives a different total.</p>\n<h2 id=\"the-hours-at-three-speeds\">The hours at three speeds</h2>\n<p>The three speeds come from <a href=\"/blog/reading-speed-words-per-minute/\">reading speed in words per minute</a>: 250 words a minute is the starting figure there, and 300 and 350 appear in its examples.</p>\n<table><thead><tr><th>Speed</th><th>Whole book</th><th>One hour a day</th><th>Half an hour a day</th></tr></thead><tbody>\n<tr><td>250 words a minute</td><td>37 h 42 min</td><td>38 days</td><td>76 days</td></tr>\n<tr><td>300 words a minute</td><td>31 h 25 min</td><td>32 days</td><td>63 days</td></tr>\n<tr><td>350 words a minute</td><td>26 h 55 min</td><td>27 days</td><td>54 days</td></tr>\n</tbody></table>\n<p>The number of chapters gives a slower plan that happens to fit the calendar: one chapter a day takes exactly a year of 365 days in this file. Counts of chapters differ between editions. The <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/War_and_Peace\">Wikipedia article</a> gives the novel 361, so with another edition the plan can end a few days before the year does.</p>\n<p>If you read it in Readwyn, the book screen keeps the estimate up to date as you go. It shows the time left from the words you haven't read and your own measured speed, as explained in <a href=\"/blog/how-long-to-read-a-book/\">how long will this book take to read</a>, and a few hours of your own reading outweigh the starting figure of 250 words a minute.</p>\n<h2 id=\"the-parts-one-by-one\">The parts, one by one</h2>\n<p>The parts are far from even. Book Ten is more than three times as long as the second epilogue, and Books Thirteen and Fourteen are each under half the length of Book One. The table gives the chapters and words under each heading of the Gutenberg file, counted the same way, with the time at 250 words a minute.</p>\n<table><thead><tr><th>Part</th><th>Years</th><th>Chapters</th><th>Words</th><th>At 250 wpm</th></tr></thead><tbody>\n<tr><td>Book One</td><td>1805</td><td>28</td><td>48,100</td><td>3 h 12 min</td></tr>\n<tr><td>Book Two</td><td>1805</td><td>21</td><td>40,600</td><td>2 h 42 min</td></tr>\n<tr><td>Book Three</td><td>1805</td><td>19</td><td>44,200</td><td>2 h 57 min</td></tr>\n<tr><td>Book Four</td><td>1806</td><td>16</td><td>23,600</td><td>1 h 34 min</td></tr>\n<tr><td>Book Five</td><td>1806 to 1807</td><td>22</td><td>33,800</td><td>2 h 15 min</td></tr>\n<tr><td>Book Six</td><td>1808 to 1810</td><td>26</td><td>33,500</td><td>2 h 14 min</td></tr>\n<tr><td>Book Seven</td><td>1810 to 1811</td><td>13</td><td>22,900</td><td>1 h 32 min</td></tr>\n<tr><td>Book Eight</td><td>1811 to 1812</td><td>22</td><td>31,100</td><td>2 h 4 min</td></tr>\n<tr><td>Book Nine</td><td>1812</td><td>23</td><td>38,000</td><td>2 h 32 min</td></tr>\n<tr><td>Book Ten</td><td>1812</td><td>39</td><td>65,000</td><td>4 h 20 min</td></tr>\n<tr><td>Book Eleven</td><td>1812</td><td>34</td><td>52,100</td><td>3 h 28 min</td></tr>\n<tr><td>Book Twelve</td><td>1812</td><td>16</td><td>24,400</td><td>1 h 38 min</td></tr>\n<tr><td>Book Thirteen</td><td>1812</td><td>19</td><td>20,300</td><td>1 h 21 min</td></tr>\n<tr><td>Book Fourteen</td><td>1812</td><td>19</td><td>20,700</td><td>1 h 23 min</td></tr>\n<tr><td>Book Fifteen</td><td>1812 to 1813</td><td>20</td><td>24,400</td><td>1 h 38 min</td></tr>\n<tr><td>First Epilogue</td><td>1813 to 1820</td><td>16</td><td>24,300</td><td>1 h 37 min</td></tr>\n<tr><td>Second Epilogue</td><td></td><td>12</td><td>17,700</td><td>1 h 11 min</td></tr>\n</tbody></table>\n<p>The years in the headings show where the weight falls. The six books headed 1812 alone hold 220,500 words, about 39% of the novel, and they include the two longest. The three books of 1805 that open the novel add up to 132,900 words, close to a quarter. If you set yourself one book a week, the slowest weeks are Ten and Eleven; from Book Twelve to the end of the novel, no part reaches 25,000 words.</p>\n<p>Wikipedia describes the Russian text as four volumes of fifteen books plus an epilogue in two parts. The Gutenberg file has no volume headings, only the books and the two epilogues.</p>\n<h2 id=\"how-many-pages\">How many pages</h2>\n<p>A word count of 565,400 has no fixed page count, because a page of an ebook depends on the screen and the font. Two figures help as a guide. Wikipedia's infobox gives 1,225 pages for the first published edition. And at 250 words a page, the rate Readwyn uses to <a href=\"/blog/ebook-page-count-from-word-count/\">estimate pages from a word count</a>, the Maude text comes to about 2,260 pages. Readwyn splits words at spaces only, so the count and the pages that it shows for its own copy of the book can differ a little. A printed edition has its own count, set by its type and its page size, so check the edition you plan to buy.</p>\n<h2 id=\"serial-first-book-after\">Serial first, book after</h2>\n<p>Tolstoy's first readers didn't get it in one go. Wikipedia records that portions of an earlier version, titled <em>The Year 1805</em>, came out in The Russian Messenger from 1865 to 1867. Tolstoy rewrote the whole novel between 1866 and 1869, and it was published in full in 1869.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-long-to-read-the-bible/","url":"https://readwyn.com/blog/how-long-to-read-the-bible/","title":"How long does it take to read the Bible?","summary":"The King James Bible is about 790,400 words: some 53 hours at 250 words a minute, or about nine minutes a day to finish it in a year.","content_html":"<p>Read from Genesis to Revelation at 250 words a minute, the King James Version takes about 52 hours and 42 minutes. That's the time for the 790,400 words of Project Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/10\">King James Version of the Bible</a>, once the verse numbers are taken out. A faster reader at 350 words a minute needs about 37 hours and 38 minutes. Spread over a year, the whole book comes to roughly 2,170 words a day, which is nine minutes at the slower pace.</p>\n<h2 id=\"genesis-to-malachi-matthew-to-revelation\">Genesis to Malachi, Matthew to Revelation</h2>\n<p>The Gutenberg file holds 66 books and a contents list: the 39 books from Genesis to Malachi, then the 27 from Matthew to Revelation. The first part is by far the larger: 609,400 words against 180,600, so about 77% of the reading lies before Matthew.</p>\n<table><thead><tr><th>Part</th><th>Words</th><th>At 250 wpm</th><th>At 300 wpm</th><th>At 350 wpm</th></tr></thead><tbody>\n<tr><td>Genesis to Malachi</td><td>609,400</td><td>40 h 38 min</td><td>33 h 51 min</td><td>29 h 1 min</td></tr>\n<tr><td>Matthew to Revelation</td><td>180,600</td><td>12 h 2 min</td><td>10 h 2 min</td><td>8 h 36 min</td></tr>\n<tr><td>Whole Bible</td><td>790,400</td><td>52 h 42 min</td><td>43 h 55 min</td><td>37 h 38 min</td></tr>\n</tbody></table>\n<p>The first column uses 250 words a minute, the speed that Readwyn assumes until it has <a href=\"/blog/reading-speed-words-per-minute/\">measured your own</a>; 300 and 350 show how much a faster pace saves. The two parts add up to 790,000; the other 400 words are the contents list at the start of the file.</p>\n<p>Matthew to Revelation on its own is about 12 hours at 250 words a minute. At 180,600 words, it is a little under a third of <a href=\"/blog/how-long-is-war-and-peace/\">War and Peace</a> in the translation by Louise and Aylmer Maude (<a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2600\">565,400 words on Gutenberg</a>). The whole Bible is about 1.4 times the length of Tolstoy's novel.</p>\n<h2 id=\"a-year-or-less\">A year, or less</h2>\n<p>For a reading plan, the daily time is the more useful figure. These are the days the whole King James text takes at a fixed amount of reading each day:</p>\n<table><thead><tr><th>Speed</th><th>15 min a day</th><th>30 min a day</th><th>One hour a day</th></tr></thead><tbody>\n<tr><td>250 words a minute</td><td>211 days</td><td>106 days</td><td>53 days</td></tr>\n<tr><td>300 words a minute</td><td>176 days</td><td>88 days</td><td>44 days</td></tr>\n<tr><td>350 words a minute</td><td>151 days</td><td>76 days</td><td>38 days</td></tr>\n</tbody></table>\n<p>A plan of one year asks for less than any column of that table: the 2,170 words a day take 9 minutes at 250 words a minute, 7 at 300 and 6 at 350.</p>\n<h2 id=\"why-the-verse-numbers-matter\">Why the verse numbers matter</h2>\n<p>Every verse of the Gutenberg file starts with its chapter and verse, written like “1:1”. A counter that splits text at spaces takes each of these as a word. The file has 31,102 of them, so a count that keeps them gives 821,500 words instead of 790,400. At 250 words a minute those 31,100 extra “words” add about two hours that nobody spends reading. An average verse, by the same counts, is about 25 words long.</p>\n<p>The rest of the method is the same as in <a href=\"/blog/how-many-words-in-a-novel/\">how many words are in a novel</a>: everything between Gutenberg's start and end lines, split at spaces and at dashes, rounded to the nearest hundred. Another translation gives another total, and so does an edition with the Apocrypha. The King James Version of 1611 had <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/King_James_Version\">80 books</a>, 14 of them the books of the Apocrypha, and Wikipedia notes that later printings of the standard 1769 text nearly always left them out. Like those printings, the Gutenberg file has none of them: its catalogue page speaks of 80 books, but the file itself has the 66 headings from Genesis to Revelation. Whether you can copy that text freely depends on the country: see <a href=\"/blog/is-the-king-james-bible-public-domain/\">is the King James Bible in the public domain</a>.</p>\n<p>If you read it as an ebook in Readwyn, the book screen gives the time left from the words you haven't read yet and your own measured speed, as explained in <a href=\"/blog/how-long-to-read-a-book/\">how long will this book take to read</a>. After a few hours of your own reading, that estimate replaces the general figures above.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/is-1984-public-domain/","url":"https://readwyn.com/blog/is-1984-public-domain/","title":"Is 1984 in the public domain?","summary":"Orwell's Nineteen Eighty-Four is public domain in Canada, Australia, the UK and the EU, but in the United States it stays in copyright until 1 January 2045.","content_html":"<p>It depends on where you are. George Orwell's <em>Nineteen Eighty-Four</em> is in the public domain in Canada, Australia, the United Kingdom and the European Union, because those countries count copyright from the author's death, and Orwell died on <a rel=\"external\" href=\"https://www.orwellfoundation.com/the-orwell-foundation/orwell/biography/\">21 January 1950</a>. In the United States it stays in copyright until 1 January 2045, because American law counts a book of that age from the year it came out, and the novel came out in June 1949.</p>\n<h2 id=\"the-united-states-2045\">The United States: 2045</h2>\n<p>A US book published from 1931 to 1963 with a copyright notice, and whose copyright was renewed, is protected for 95 years after publication, according to <a rel=\"external\" href=\"https://guides.library.cornell.edu/copyright/publicdomain\">Cornell University Library's copyright term chart</a>. Many American books of that period are already free because nobody renewed them. A British book can't count on that gap: the Uruguay Round Agreements Act restored, from 1 January 1996, the US copyright of certain foreign works that had lapsed for want of renewal (<a rel=\"external\" href=\"https://www.copyright.gov/circs/circ15a.pdf\">Circular 15A</a>). <em>Nineteen Eighty-Four</em> was renewed anyway: Wikisource's <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Author:George_Orwell\">page for Orwell</a> lists its renewal, number R641953, and marks the novel as in copyright in the United States until 2045. A term runs to the end of the calendar year, so 1949 plus 95 gives the end of 2044. Standard Ebooks already has a <a rel=\"external\" href=\"https://standardebooks.org/ebooks/george-orwell/nineteen-eighty-four\">placeholder page for the novel</a> that says it will enter the US public domain on 1 January 2045, with no file to download yet.</p>\n<p>Four of Orwell's other books have renewals listed on the same Wikisource page, and follow the same arithmetic:</p>\n<table><thead><tr><th>Book</th><th>Published</th><th>Free in the US on</th></tr></thead><tbody>\n<tr><td><em>Down and Out in Paris and London</em></td><td>1933</td><td>1 January 2029</td></tr>\n<tr><td><em>Burmese Days</em></td><td>1934</td><td>1 January 2030</td></tr>\n<tr><td><em>Homage to Catalonia</em></td><td>1938</td><td>1 January 2034</td></tr>\n<tr><td><em>Animal Farm</em></td><td>1945</td><td>1 January 2041</td></tr>\n<tr><td><em>Nineteen Eighty-Four</em></td><td>1949</td><td>1 January 2045</td></tr>\n</tbody></table>\n<p>The rules behind the 95 years, and the many books that lost their copyright because nobody renewed it, are in the post on <a href=\"/blog/when-do-books-enter-the-public-domain/\">when a book enters the public domain</a>.</p>\n<h2 id=\"canada-since-2001\">Canada: since 2001</h2>\n<p>Canada protected a book for the author's life plus 50 years until 30 December 2022, when the term became life plus 70. The Canadian Intellectual Property Office's <a rel=\"external\" href=\"https://ised-isde.canada.ca/site/canadian-intellectual-property-office/en/guide-copyright\">guide to copyright</a> gives both terms and the date. <a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/acts/C-42/section-6.html\">Section 6 of the Copyright Act</a> counts the term from the end of the calendar year in which the author died. For Orwell, 50 years from the end of 1950 ran out at the end of 2000, so <em>Nineteen Eighty-Four</em> has been public domain in Canada since 1 January 2001. The 2022 change didn't bring it back: section 280 of the <a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/AnnualStatutes/2022_10/page-21.html\">Budget Implementation Act, 2022, No. 1</a> says the new term does not revive a copyright that had already expired.</p>\n<h2 id=\"australia-the-uk-and-the-eu\">Australia, the UK and the EU</h2>\n<p>Australia moved to life plus 70 in 2005, and didn't apply the change to works already free. The National Library of Australia's page on <a rel=\"external\" href=\"https://www.library.gov.au/services/copyright-library-collections/how-long-does-copyright-last\">how long copyright lasts</a> says that a published work whose creator died before 1 January 1955 is now in the public domain. Orwell qualifies with five years to spare. Project Gutenberg Australia first posted its <a rel=\"external\" href=\"https://gutenberg.net.au/ebooks01/0100021h.html\">ebook of <em>Nineteen Eighty-Four</em></a> in August 2001. Its <a rel=\"external\" href=\"https://gutenberg.net.au/licence.html\">licence</a> says it can check copyright only under Australian law, and that people outside Australia should check their own laws before they pass its ebooks on.</p>\n<p>Britain protects a written work for <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">70 years after the author's death</a>, and so does the EU. For an author who died in 1950, that ends with 2020. Duke's Center for the Study of the Public Domain named Orwell among the authors whose works entered the EU public domain <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2021/\">in 2021</a>, with Edgar Rice Burroughs and George Bernard Shaw.</p>\n<p>A copy that is free to share in Toronto or Sydney is still under copyright in Chicago, which is why the Australian site warns its foreign readers. The post on <a href=\"/blog/public-domain-differs-by-country/\">why a book can be public domain in one country and not another</a> covers that split. The novel has an odd place in ebook history too: in 2009 Amazon <a href=\"/blog/amazon-deleted-1984-from-kindles/\">deleted copies of <em>Nineteen Eighty-Four</em> from Kindles</a> after a seller with no rights to the book had listed them.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/is-peter-pan-public-domain/","url":"https://readwyn.com/blog/is-peter-pan-public-domain/","title":"Is Peter Pan in the public domain?","summary":"Yes, almost everywhere. In the UK the copyright ended in 2007, but a 1988 law still gives Great Ormond Street Hospital a royalty on the play and its adaptations.","content_html":"<p>Yes, almost everywhere. J. M. Barrie's novel <em>Peter and Wendy</em> (1911) and his play <em>Peter Pan</em> are public domain in the United States and the European Union, and the hospital charity's own <a rel=\"external\" href=\"https://www.gosh.org/about-us/peter-pan/theatre-performances/\">page on Peter Pan performances</a> says that Peter Pan is public domain \"in all countries outside of the UK\". The exception is the United Kingdom. There the copyright itself has run out, but an Act of Parliament from 1988 still lets the hospital's charity collect a royalty when the play, or an adaptation of it, is performed in public, published commercially or broadcast. Nobody needs permission to read the book. Even in the UK the hospital can't stop anyone using the story; it can only ask for a royalty on some uses.</p>\n<h2 id=\"a-gift-to-a-children-s-hospital\">A gift to a children's hospital</h2>\n<p>In 1929 Barrie was asked to sit on a committee to help Great Ormond Street buy land for a new wing. He declined, and, as the <a rel=\"external\" href=\"https://www.gosh.org/about-us/peter-pan/history/\">hospital charity tells it</a>, the board learned two months later that he had given it the rights to <em>Peter Pan</em> instead. Wikipedia's <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/J._M._Barrie\">life of Barrie</a> dates the gift to April 1929, and his will left the Peter Pan rights out of his estate because the hospital already had them.</p>\n<p>Barrie died on 19 June 1937. British copyright then lasted 50 years after the author's death, so, as the <a rel=\"external\" href=\"https://neverlandofficial.com/discover/peter-pan-copyright/\">charity's copyright page</a> puts it, the original copyright expired in 1987. The new copyright law of the following year wrote the hospital into it, a provision that the charity <a rel=\"external\" href=\"https://www.gosh.org/about-us/peter-pan/\">credits to Lord Callaghan and his wife Audrey</a>, then chair of the hospital's board of governors.</p>\n<h2 id=\"what-the-1988-act-gives-the-hospital\">What the 1988 Act gives the hospital</h2>\n<p><a rel=\"external\" href=\"https://assets.publishing.service.gov.uk/media/60180c2b8fa8f53fc62c5897/Copyright-designs-and-patents-act-1988.pdf\">Section 301 of the Copyright, Designs and Patents Act 1988</a> gives GOSH Children's Charity \"a right to a royalty\" for \"the public performance, commercial publication or communication to the public\" of the play, \"or of any adaptation of that work\", even though its copyright \"expired on 31st December 1987\". Schedule 6 of the Act fills in the detail:</p>\n<ul>\n<li><strong>The work</strong> is the play \"Peter Pan\" by Sir James Matthew Barrie. The novel isn't named, and the Act doesn't say whether it counts as an adaptation. The charity's page lists stage shows, publications, audiobooks, ebooks, radio, television and film among the uses that owe it a royalty.</li>\n<li><strong>It has no term in years.</strong> It ends only if the charity tries to assign it or pledge it as security, or if the hospital loses its separate identity or stops caring for sick children. That is why it is often called a perpetual copyright. It isn't one: the charity is owed a royalty, and if the two sides can't agree on the sum, the Copyright Tribunal sets it.</li>\n<li><strong>Some uses are free.</strong> No royalty is due for anything that the copyright exceptions of the Act would allow if the copyright still existed, such as quoting for criticism or review, or for anything that needed no licence before the end of 1987. A school play is one such use when pupils or teachers perform it as part of school life to pupils, teachers and others directly connected with the school (<a rel=\"external\" href=\"https://www.legislation.gov.uk/ukpga/1988/48/section/34\">section 34</a>); being a pupil's parent doesn't by itself make someone directly connected, so a show put on for the parents falls outside that exception.</li>\n<li><strong>Old deals are untouched.</strong> No royalty is due on arrangements made before the Act was passed.</li>\n</ul>\n<p>The UK term later grew to <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">70 years after the author's death</a>, in line with an EU directive. That brought Barrie's copyright back to life in Britain in 1996, and it ran out for good on <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Copyright,_Designs_and_Patents_Act_1988#Peter_Pan\">31 December 2007</a>. The royalty right stayed. The hospital's <a rel=\"external\" href=\"https://www.gosh.org/about-us/peter-pan/theatre-performances/\">theatre page</a> still licenses UK performances, whether the show is an original script, a pantomime or a school play.</p>\n<h2 id=\"the-american-dates\">The American dates</h2>\n<p>American law counts a book of this age from the year it was published, not from the author's death, and the two works took different routes. Duke University's Center for the Study of the Public Domain, in its <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2024/\">list for 1 January 2024</a>, notes that <em>Peter and Wendy</em> has been public domain in the US since 1967. The play had been on stage since 1904, but its script wasn't published for copyright purposes until 1928, so it only became free in the US on 1 January 2024, with the other works of 1928 that were still in copyright. <a href=\"/blog/public-domain-day/\">Public Domain Day</a> explains why that date comes round each January, and <a href=\"/blog/public-domain-differs-by-country/\">why a book can be free in one country and not another</a> covers the general rule behind splits like this one. <em>Winnie-the-Pooh</em> is another British children's classic whose dates differ on each side of the Atlantic; <a href=\"/blog/is-winnie-the-pooh-public-domain/\">its own post</a> has them.</p>\n<h2 id=\"where-to-get-the-book\">Where to get the book</h2>\n<p>Project Gutenberg has <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/16\"><em>Peter Pan [Peter and Wendy]</em></a>, marked \"Public domain in the USA\", as an EPUB3, an EPUB for older e-readers, two Kindle files and plain text. The novel runs to about 47,300 words. Printing and selling a UK edition is where, by the charity's reading, the 1988 Act comes in: its copyright page counts publications and ebooks among the uses that owe a royalty.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/is-the-king-james-bible-public-domain/","url":"https://readwyn.com/blog/is-the-king-james-bible-public-domain/","title":"Is the King James Bible in the public domain?","summary":"In the United States and most of the world, yes. In the UK it isn't: the Crown keeps a perpetual right over printing it, licensed to publishers by letters patent.","content_html":"<p>In the United States, yes, and in most other countries too. The King James Version came out in 1611, and American law treats every work published before 1931 as free, according to <a rel=\"external\" href=\"https://guides.library.cornell.edu/copyright/publicdomain\">Cornell University Library's copyright term chart</a>. The United Kingdom is the exception. There the text isn't protected by ordinary copyright at all but by <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/King_James_Version#Copyright_status\">a right of the Crown that never runs out</a>, and only publishers holding the Crown's letters patent, or those they authorise, may print it, publish it or import it into the country.</p>\n<h2 id=\"a-right-older-than-copyright\">A right older than copyright</h2>\n<p>Copyright in the UK lasts <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">70 years after the author's death</a>, which would have freed a translation by scholars of the early 1600s long ago. The King James Bible sits outside that system. The right to print it is part of the royal prerogative, the powers the Crown holds by custom rather than by statute, and the Copyright, Designs and Patents Act 1988 leaves such powers alone. Section 171(1)(b) says that nothing in the copyright part of the Act affects \"any right or privilege of the Crown subsisting otherwise than under an enactment\" (the <a rel=\"external\" href=\"https://assets.publishing.service.gov.uk/media/60180c2b8fa8f53fc62c5897/Copyright-designs-and-patents-act-1988.pdf\">consolidated text of the Act</a> published by the Intellectual Property Office, page 166).</p>\n<p>The arrangement goes back to the first printing. The title page of 1611 reads \"Imprinted at London by Robert Barker, Printer to the Kings most Excellent Majestie\", and Barker was the King's Printer. His father Christopher had been granted the perpetual royal privilege to print Bibles in England by Elizabeth I in 1589. Wikipedia's <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/King_James_Version#Copyright_status\">history of the King James Version</a> traces the office of the King's Printer and its Bible right back to at least 1577. In the 18th century John Baskett bought out the surviving shares of the monopoly, and the right passed through a line of printers until Cambridge University Press took over Eyre &amp; Spottiswoode in 1990 and became the King's Printer.</p>\n<h2 id=\"who-holds-the-right-today\">Who holds the right today</h2>\n<p>The letters patent split the UK in two, as the same article and its page on the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/King%27s_Printer\">King's Printer</a> describe:</p>\n<ul>\n<li><strong>England, Wales and Northern Ireland.</strong> The King's Printer, now Cambridge University Press, holds the patent. Separate grants of similar age let Cambridge and Oxford University Press print the Authorised Version whoever holds that office.</li>\n<li><strong>Scotland.</strong> The Scottish Bible Board licenses the printing of the Bible on behalf of the Crown, and Collins publishes the Authorised Version under its licence.</li>\n</ul>\n<p>The <em>Book of Common Prayer</em> is covered by the same kind of patent. Wikipedia describes the two books as the last remnant of the time when the Crown held a monopoly over all printing and publishing in the United Kingdom.</p>\n<p>In practice Cambridge licenses rather than forbids. Wikipedia summarises its published terms: up to 500 words may be quoted for liturgical and non-commercial educational use with the Press's acknowledgement, provided the quotation is no more than a quarter of the work quoting it and is not a complete book of the Bible. Anything beyond that is decided case by case.</p>\n<h2 id=\"which-text-and-which-edition\">Which text, and which edition</h2>\n<p>The words most people now read as \"the King James Bible\" aren't quite those of 1611. The Oxford edition of 1769, edited by Benjamin Blayney, standardised the spelling and punctuation and differs from the 1611 text in around 24,000 places. Most current printings reproduce it almost unchanged. Its age puts it in the same position as the 1611 text: free in the US, under the Crown's right in the UK.</p>\n<p>A modern printing can still carry rights of its own. The new introduction, notes and cross-references of an edition are new writing. In the UK the layout of a published edition is protected too, for <a rel=\"external\" href=\"https://assets.publishing.service.gov.uk/media/60180c2b8fa8f53fc62c5897/Copyright-designs-and-patents-act-1988.pdf\">25 years from the end of the year it first came out</a> (section 15 of the Act).</p>\n<p>The New King James Version is a different translation, not an edition of the old one. Bible Gateway gives its notice as <a rel=\"external\" href=\"https://www.biblegateway.com/versions/New-King-James-Version-NKJV-Bible/\">\"Copyright © 1982 by Thomas Nelson\"</a>, with permission to quote up to 500 verses on conditions, so it is in copyright everywhere.</p>\n<h2 id=\"getting-a-free-copy\">Getting a free copy</h2>\n<p>Project Gutenberg released <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/10\"><em>The King James Version of the Bible</em></a> on 1 August 1989 as its tenth ebook. The page marks it \"Public domain in the USA.\" and offers it as EPUB 3, as Kindle files, and as plain text. Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/policy/license.html\">licence</a> tells readers outside the United States to check the laws of the country they are in, and the UK is the country where that advice matters for this book. <a href=\"/blog/what-is-project-gutenberg/\">What Project Gutenberg is</a> covers the site and its files. <a href=\"/blog/public-domain-differs-by-country/\">Why a book can be public domain in one country and not another</a> explains how the terms of each country decide such cases. <em>Peter Pan</em> is a different UK case: there the copyright has run out, but an Act of Parliament still owes <a href=\"/blog/is-peter-pan-public-domain/\">Great Ormond Street Hospital's charity a royalty when the play is performed or published</a>.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/is-the-little-prince-public-domain/","url":"https://readwyn.com/blog/is-the-little-prince-public-domain/","title":"Is The Little Prince in the public domain?","summary":"The Little Prince is public domain in Canada and most of Europe, but stays in copyright in the US until 1 January 2039. France's date is disputed, and the English translation is free only in Canada.","content_html":"<p>In Canada, the UK and most of the EU, yes, but not in the United States or in France. Antoine de Saint-Exupéry <a rel=\"external\" href=\"https://www.antoinedesaintexupery.org/la-biographie/\">disappeared on a reconnaissance flight on 31 July 1944</a>, so <em>Le Petit Prince</em> has been free in Canada since 1 January 1995 and in the UK and most of the EU since 1 January 2015. The United States counts from publication instead, and the book stays protected there until 1 January 2039. France adds 30 years for an author who died for France, which puts its date somewhere in the 2030s or 2040s. That covers Saint-Exupéry's French text and drawings. Katherine Woods's 1943 English translation has its own term, and in the United States, the UK and the EU it stays protected until 1 January 2039.</p>\n<h2 id=\"published-first-in-new-york\">Published first in New York</h2>\n<p>The book <a rel=\"external\" href=\"https://www.antoinedesaintexupery.org/la-biographie/\">first came out in New York</a>, not Paris. Reynal &amp; Hitchcock <a rel=\"external\" href=\"https://www.antoinedesaintexupery.org/personne/le-petit-prince-1943-2/\">published it on 6 April 1943, in English and in French</a>, and Gallimard's first French edition followed only in April 1946, after the author's death. Both New York editions were registered for US copyright, and both were renewed in time, as the Stanford copyright renewal database records:</p>\n<table><thead><tr><th>Edition</th><th>Registered</th><th>Renewed</th><th>Renewal claimant</th></tr></thead><tbody>\n<tr><td><em>Le petit prince</em>, with the author's drawings</td><td>A172160</td><td><a rel=\"external\" href=\"https://exhibits.stanford.edu/copyrightrenewals/catalog/R495966\">R495966</a>, 30 November 1970</td><td>Consuelo de Saint Exupery, his widow</td></tr>\n<tr><td><em>The Little Prince</em>, translated by Katherine Woods</td><td>A172161</td><td><a rel=\"external\" href=\"https://exhibits.stanford.edu/copyrightrenewals/catalog/R504624\">R504624</a>, 31 March 1971</td><td>Harcourt Brace Jovanovich</td></tr>\n</tbody></table>\n<p>A US book published from 1931 to 1963 with a notice, and renewed, is protected for 95 years after publication, and every term runs to the end of its calendar year, according to <a rel=\"external\" href=\"https://guides.library.cornell.edu/copyright/publicdomain\">Cornell University Library's copyright term chart</a>. 1943 plus 95 is 2038, so the French text, the drawings and the Woods translation all enter the US public domain together on 1 January 2039. <a href=\"/blog/when-do-books-enter-the-public-domain/\">When a book enters the public domain</a> explains the renewal rule, and why so many American books of the same years are already free.</p>\n<h2 id=\"canada-the-uk-and-europe\">Canada, the UK and Europe</h2>\n<p>Countries that count from the author's death reached the book much sooner. Canada's term was the author's life plus 50 years, so Saint-Exupéry's work came free there at the end of 1994; <a rel=\"external\" href=\"https://communia-association.org/2015/01/23/the-little-prince-almost-in-the-public-domain\">COMMUNIA's 2015 piece</a> notes that Canadians had by then been free to use the book for about 20 years. A 2022 budget law raised the term in <a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/acts/C-42/section-6.html\">section 6 of the Copyright Act</a> to life plus 70, in force from <a rel=\"external\" href=\"https://orders-in-council.canada.ca/attachment.php?attach=42930\">30 December 2022</a>. That did not bring the book back: <a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/AnnualStatutes/2022_10/page-21.html\">section 280</a> of the budget law that made the change rules out any revival.</p>\n<p>The UK protects a written or artistic work for <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">70 years after the author's death</a>, and the <a rel=\"external\" href=\"https://publicdomainreview.org/features/entering-the-public-domain/2022/\">Public Domain Review</a> gives the same life plus 70 for most of the EU and South America. For an author who died in 1944, that ran out at the end of 2014. COMMUNIA marked the date in January 2015, calling the book public domain \"almost everywhere in Europe\". Which country's date applies to you depends on where you use the book, as <a href=\"/blog/public-domain-differs-by-country/\">why a book can be public domain in one country and not another</a> explains. <em><a href=\"/blog/is-1984-public-domain/\">Nineteen Eighty-Four</a></em> has the same split, free in Europe and Canada but protected in the United States until 1 January 2045.</p>\n<h2 id=\"the-english-translation-is-a-separate-work\">The English translation is a separate work</h2>\n<p>Saint-Exupéry's own text and drawings are free in those countries. A translation has an author of its own, and its term counts from the translator's death. Katherine Woods, who made the 1943 English version, died on 5 February 1968, according to her <a rel=\"external\" href=\"https://id.loc.gov/authorities/names/n82091851.html\">Library of Congress name record</a>. In Canada her life plus 50 ran out at the end of 2018, four years before the longer term came in, so her translation is free there too. In the UK and the other life plus 70 countries it stays protected until the end of 2038. A British publisher can reprint the French text today, but has to wait for 1 January 2039 to use the Woods wording, the same day as the Americans.</p>\n<h2 id=\"france-adds-30-years\">France adds 30 years</h2>\n<p>Saint-Exupéry was <a rel=\"external\" href=\"https://www.antoinedesaintexupery.org/la-biographie/\">recognised as <em>mort pour la France</em> in 1945</a>, and <a rel=\"external\" href=\"https://www.legifrance.gouv.fr/codes/article_lc/LEGIARTI000006278952\">article L123-10 of the Code de la propriété intellectuelle</a> extends an author's rights by 30 years in that case. How those 30 years combine with France's wartime extensions is disputed, so the date isn't settled. Lionel Maurel worked it out on his blog <a rel=\"external\" href=\"http://scinfolex.com/2014/04/05/la-malediction-du-petit-prince-ou-le-domaine-public-un-jour-dissous-dans-le-droit-des-marques\">S.I.Lex in 2014</a> and concluded \"not before 2032\". COMMUNIA's piece gives a range from 1 May 2033 to 1 January 2045.</p>\n<p>Maurel's post also points to trademarks: the heirs have registered the main characters, the prince, the rose and the fox among them, as marks. In a reply under the post, he adds that a trademark covers only the classes of goods it was filed for, and that the heir has said reprinting the text of the novel itself will stay possible.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/is-winnie-the-pooh-public-domain/","url":"https://readwyn.com/blog/is-winnie-the-pooh-public-domain/","title":"Is Winnie-the-Pooh in the public domain?","summary":"In the US, yes: the 1926 book has been free since 2022 and Tigger since 2024. In the UK, Milne's text comes free in 2027 and Shepard's drawings in 2047.","content_html":"<p>In the United States, yes. A. A. Milne's <em>Winnie-the-Pooh</em> came out in 1926 and entered the US public domain on 1 January 2022, together with E. H. Shepard's original illustrations, as Duke's Center for the Study of the Public Domain recorded on <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2022/\">its list for that year</a>. Pooh, Piglet, Eeyore, Rabbit, Kanga, Roo, Owl and Christopher Robin, as that book draws them, are free to copy and build on there. In Britain the answer is \"not yet\": Milne died on 31 January 1956, so his words come free on 1 January 2027, and Shepard's drawings stay protected until the end of 2046.</p>\n<h2 id=\"the-pooh-books-year-by-year\">The Pooh books, year by year</h2>\n<p>Duke's 2022 page is careful about one point: only the 1926 book went free that year, not the later stories or the films. Tigger, for example, isn't in it. He arrives in <em>The House at Pooh Corner</em> (1928), which had to wait until 2024. Each of the four Pooh books came free on a 1 January of its own:</p>\n<table><thead><tr><th>Book</th><th>Published</th><th>Public domain in the US from</th></tr></thead><tbody>\n<tr><td><em>When We Were Very Young</em></td><td>1924</td><td><a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2020/\">1 January 2020</a></td></tr>\n<tr><td><em>Winnie-the-Pooh</em></td><td>1926</td><td><a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2022/\">1 January 2022</a></td></tr>\n<tr><td><em>Now We Are Six</em></td><td>1927</td><td><a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2023/\">1 January 2023</a></td></tr>\n<tr><td><em>The House at Pooh Corner</em></td><td>1928</td><td><a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2024/\">1 January 2024</a></td></tr>\n</tbody></table>\n<p>Duke's lists for 2022, 2023 and 2024 name Shepard's illustrations with each book. Project Gutenberg released <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/67098\"><em>Winnie-the-Pooh</em></a> on 3 January 2022 and <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/73011\"><em>The House at Pooh Corner</em></a> on 21 February 2024. Both pages credit Shepard as the illustrator and mark the books \"Public domain in the USA.\" The <em>Winnie-the-Pooh</em> page offers EPUB and Kindle files with the pictures, and a separate EPUB without images for older e-readers. <a href=\"/blog/what-is-project-gutenberg/\">What Project Gutenberg is</a> covers the site and its files.</p>\n<h2 id=\"britain-europe-and-canada\">Britain, Europe and Canada</h2>\n<p>The UK protects a written or artistic work for <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">70 years after the author's death</a>. The <a rel=\"external\" href=\"https://publicdomainreview.org/features/entering-the-public-domain/2026/\">Public Domain Review</a> shows the arithmetic: in life plus 70 countries (it names the UK, Russia, and most of the EU and South America), work by people who died in 1955 became free on 1 January 2026. Milne <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/A._A._Milne\">died in 1956</a>, so his text follows on 1 January 2027.</p>\n<p>The illustrations are a separate work by a separate author. Shepard <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/E._H._Shepard\">died on 24 March 1976</a>, aged 96, so in those countries his Pooh drawings are protected until the end of 2046. From 2027 to 2046 a British publisher can print Milne's words freely but needs permission for Shepard's pictures.</p>\n<p>In Canada, Milne's text is already free, and has been since 1 January 2007. Canada's term was the author's life plus 50 years (the <a rel=\"external\" href=\"https://publicdomainreview.org/features/entering-the-public-domain/2022/\">Public Domain Review's 2022 list</a> still names it among those countries) until a 2022 amendment raised it to 70 (<a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/acts/C-42/section-6.html\">section 6 of the Copyright Act</a>). Milne's Canadian copyright had expired at the end of 2006, and section 280 of the <a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/AnnualStatutes/2022_10/page-21.html\">Budget Implementation Act, 2022, No. 1</a> says the new term revives no copyright that had already expired. Shepard's term was still running in 2022, so it now lasts to the end of 2046 there too. The general rule behind these splits is in <a href=\"/blog/public-domain-differs-by-country/\">why a book can be public domain in one country and not another</a>.</p>\n<h2 id=\"what-disney-still-owns\">What Disney still owns</h2>\n<p>Jennifer Jenkins, the director of Duke's centre, wrote <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2023/bcvpd/\">an essay on Pooh, copyright and trademarks</a> for Public Domain Day 2022. Its main points, for the US:</p>\n<ul>\n<li><strong>The book and its bear are free.</strong> Anyone can post the 1926 book online or print an edition, and can copy the bear's appearance in Shepard's original drawings.</li>\n<li><strong>Disney's later work is not.</strong> Disney still holds copyrights in its later books and films, but only in the material they added, not in what they took from Milne.</li>\n<li><strong>The red shirt has its own history.</strong> The literary agent Stephen Slesinger, who bought certain rights from Milne, was the first to draw Pooh in colour with the familiar red shirt.</li>\n<li><strong>Trademarks are narrower than they look.</strong> Disney keeps trademark rights in the words \"Winnie the Pooh\" on products, but trademark law stops a use only where it's likely to confuse buyers about who made or sponsored the product. A trademark can't be used to extend a copyright that has run out. People may use a trademark in the title of an expressive work if it is artistically relevant and doesn't explicitly mislead about who made the work, and Jenkins adds that a note saying the work is not produced or sponsored by Disney puts the matter beyond doubt.</li>\n</ul>\n<p>The essay predates <a rel=\"external\" href=\"https://www.supremecourt.gov/opinions/22pdf/22-148_3e04.pdf\"><em>Jack Daniel's v. VIP Products</em></a> (2023), in which the Supreme Court held that the title test no longer covers a name used as a brand for one's own goods.</p>\n<p>Duke's <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2023/\">2023 list</a> shows what the first year produced, from a Mint Mobile advert to the horror film <em>Winnie-the-Pooh: Blood and Honey</em>. Why a book from the 1920s comes free on a fixed 1 January is in <a href=\"/blog/when-do-books-enter-the-public-domain/\">when a book enters the public domain</a>, and the books that follow each year are in <a href=\"/blog/public-domain-day/\">Public Domain Day</a>.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/longest-classic-novels-by-word-count/","url":"https://readwyn.com/blog/longest-classic-novels-by-word-count/","title":"The longest classic novels, by word count","summary":"Clarissa leads at about 967,300 words, then The Vicomte de Bragelonne, Les Misérables and War and Peace, counted from Project Gutenberg's texts.","content_html":"<p>Samuel Richardson's <em>Clarissa</em> is the longest of the classic novels counted here, at about 967,300 words across the nine volumes of Project Gutenberg's edition. Nothing else comes close. Dumas's <em>The Vicomte de Bragelonne</em> is second at 721,500, and <em>Les Misérables</em> and <em>War and Peace</em> follow with about 567,500 and 565,400 in the English translations that Gutenberg carries. Below those come <em>The Count of Monte Cristo</em> and <em>Don Quixote</em>, and then six books packed close to 350,000 words, three of them by Dickens.</p>\n<h2 id=\"the-twelve-longest\">The twelve longest</h2>\n<p>Every count here comes from the plain-text files of Project Gutenberg, made by one method: everything between Gutenberg's start and end lines, split at spaces, dashes and double hyphens (a hyphenated word counts once), rounded to the nearest hundred. It's the method of <a href=\"/blog/how-many-words-in-a-novel/\">how many words are in a novel</a>, so the numbers can be compared with the ones there. The hours are at 250 words a minute, the default pace that <a href=\"/blog/reading-speed-words-per-minute/\">Readwyn starts from before it measures yours</a>.</p>\n<table><thead><tr><th></th><th>Novel</th><th>Words</th><th>At 250 wpm</th></tr></thead><tbody>\n<tr><td>1</td><td><em>Clarissa</em>, Samuel Richardson</td><td>967,300</td><td>64 h 29 min</td></tr>\n<tr><td>2</td><td><em>The Vicomte de Bragelonne</em>, Alexandre Dumas</td><td>721,500</td><td>48 h 6 min</td></tr>\n<tr><td>3</td><td><em>Les Misérables</em>, Victor Hugo, translated by Isabel Florence Hapgood</td><td>567,500</td><td>37 h 50 min</td></tr>\n<tr><td>4</td><td><em>War and Peace</em>, Leo Tolstoy, translated by Louise and Aylmer Maude</td><td>565,400</td><td>37 h 42 min</td></tr>\n<tr><td>5</td><td><em>The Count of Monte Cristo</em>, Alexandre Dumas</td><td>463,400</td><td>30 h 54 min</td></tr>\n<tr><td>6</td><td><em>Don Quixote</em>, Miguel de Cervantes, translated by John Ormsby</td><td>427,600</td><td>28 h 30 min</td></tr>\n<tr><td>7</td><td><em>Dombey and Son</em>, Charles Dickens</td><td>357,300</td><td>23 h 49 min</td></tr>\n<tr><td>8</td><td><em>David Copperfield</em>, Charles Dickens</td><td>357,200</td><td>23 h 49 min</td></tr>\n<tr><td>9</td><td><em>Bleak House</em>, Charles Dickens</td><td>355,600</td><td>23 h 42 min</td></tr>\n<tr><td>10</td><td><em>The Way We Live Now</em>, Anthony Trollope</td><td>354,200</td><td>23 h 37 min</td></tr>\n<tr><td>11</td><td><em>The Brothers Karamazov</em>, Fyodor Dostoyevsky, translated by Constance Garnett</td><td>351,900</td><td>23 h 28 min</td></tr>\n<tr><td>12</td><td><em>The History of Tom Jones, a Foundling</em>, Henry Fielding</td><td>351,800</td><td>23 h 27 min</td></tr>\n</tbody></table>\n<p>The Gutenberg texts: <em>Clarissa</em> in nine files, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/9296\">volume 1</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/9798\">2</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/9881\">3</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/10462\">4</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/10799\">5</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/11364\">6</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/11889\">7</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/12180\">8</a> and <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/12398\">9</a>; <em>The Vicomte de Bragelonne</em> in four, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2609\">2609</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2681\">2681</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2710\">2710</a> and <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2759\">2759</a>; then <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/135\">Les Misérables</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2600\">War and Peace</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1184\">The Count of Monte Cristo</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/996\">Don Quixote</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/821\">Dombey and Son</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/766\">David Copperfield</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1023\">Bleak House</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/5231\">The Way We Live Now</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/28054\">The Brothers Karamazov</a> and <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/6593\">Tom Jones</a>. The files of the two Dumas novels don't name a translator.</p>\n<p>The bottom half of the table is a near tie. Seventh and twelfth place are 5,500 words apart, about 22 minutes of reading, and <em>Anna Karenina</em> in Constance Garnett's translation (<a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1399\">1399</a>) misses the list by 1,100 words with 350,700. Another edition, or a count that splits words differently, could reorder all six.</p>\n<h2 id=\"how-sure-is-first-place\">How sure is first place?</h2>\n<p><em>Clarissa</em> holds up well among novels in English. Wikipedia's <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Clarissa;_or,_The_History_of_a_Young_Lady\">article on the novel</a> gives it about 970,000 words, within 3,000 of the Gutenberg count, and calls it one of the longest novels in English. It's a novel in letters. Its first two volumes came out in December 1747 and the other five in 1748, so Gutenberg's nine files don't follow the first edition's seven. Classics in other languages go further. Proust's <em>In Search of Lost Time</em>, which Wikipedia <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/In_Search_of_Lost_Time\">says</a> holds the Guinness World Record for the longest novel, and Madeleine de Scudéry's <em>Artamène</em>, which <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Artam%C3%A8ne\">Wikipedia</a> puts at about 1,954,300 words, are both longer. Neither is in this count.</p>\n<p>At 967,300 words it is about twenty times the length of <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/64317\">The Great Gatsby</a></em> (48,600), and more than twenty-four times the 40,000 words that the <a rel=\"external\" href=\"https://www.wsfs.org/wp-content/uploads/2026/06/WSFS-Constitution-as-of-2025-10-07-Clean_20260606.pdf\">Hugo rules</a> set as the shortest novel, as <a href=\"/blog/how-many-words-in-a-novel/\">how many words are in a novel</a> explains. At 250 words a minute, an hour a day takes about nine weeks.</p>\n<h2 id=\"why-some-books-are-on-the-list-and-some-aren-t\">Why some books are on the list and some aren't</h2>\n<p>Second place depends on what you call one novel. Dumas published <em>The Vicomte de Bragelonne</em> as one serial from 1847 to 1850, but <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Vicomte_of_Bragelonne:_Ten_Years_Later\">Wikipedia notes</a> that editions split it into three, four or more books. Gutenberg uses four, and the last is <em>The Man in the Iron Mask</em>. Taken one by one, none of the four would make this table: the longest file has 190,500 words. The <a href=\"/blog/three-musketeers-books-in-order/\">Three Musketeers reading order</a> gives each part.</p>\n<p>A series is not a novel, so it's left out. The six Gutenberg files of the d'Artagnan novels (<a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1257\">1257</a>, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1259\">1259</a> and the four above) come to 1,194,800 words, more than <em>Clarissa</em>, but that is three novels.</p>\n<p><em>Don Quixote</em> has some help: its count includes John Ormsby's preface and his notes, which a count of the story alone would leave out.</p>\n<h2 id=\"just-outside-the-twelve\">Just outside the twelve</h2>\n<p>Three Dickens novels make the twelve, and five more pass 300,000 words: <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/963\">Little Dorrit</a></em> (339,700), <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/968\">Martin Chuzzlewit</a></em> (338,200), <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/883\">Our Mutual Friend</a></em> (326,900), <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/967\">Nicholas Nickleby</a></em> (324,600) and <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/580\">The Pickwick Papers</a></em> (302,400).</p>\n<p>Others cross the same line. Trollope's <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/5140\">He Knew He Was Right</a></em> has 343,700 words, George Eliot's <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/145\">Middlemarch</a></em> 318,600 and <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/7469\">Daniel Deronda</a></em> 311,400, and Thackeray's <em><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/599\">Vanity Fair</a></em> 304,500.</p>\n<p><a href=\"/blog/how-long-is-war-and-peace/\">How long is War and Peace</a> gives the hours of Tolstoy's novel book by book. If you read in Readwyn, the app counts the words of your own copy, and the book screen <a href=\"/blog/how-long-to-read-a-book/\">estimates the time left</a> from the speed it measures as you read.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/oz-books-in-order/","url":"https://readwyn.com/blog/oz-books-in-order/","title":"The Oz books in order, and which are in the public domain","summary":"Read L. Frank Baum's fourteen Oz books in the order they came out, 1900 to 1920. All fourteen are in the public domain, and in the US so are some of the sequels by other writers.","content_html":"<p>L. Frank Baum wrote fourteen Oz novels, from <em>The Wonderful Wizard of Oz</em> in 1900 to <em>Glinda of Oz</em> in 1920, and the order to read them in is the order they came out. Every one of them is in the public domain in the United States, in the UK, and in any other country that protects a book for 70 years after its author's death or less: Baum died in 1919. The sequels that other writers carried on with after him are a different matter, and in the United States the line runs through the middle of Ruth Plumly Thompson's nineteen books.</p>\n<h2 id=\"baum-s-fourteen\">Baum's fourteen</h2>\n<p>The years are the ones that the <a rel=\"external\" href=\"https://onlinebooks.library.upenn.edu/webbin/book/lookupname?key=Baum%2C%20L%2E%20Frank%20%28Lyman%20Frank%29%2C%201856%2D1919\">Online Books Page</a> gives for each book, and the <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/961\">Project Gutenberg page</a> for <em>Glinda of Oz</em>. The word counts are from Gutenberg's plain-text files, counted the same way as in <a href=\"/blog/how-many-words-in-a-novel/\">how many words are in a novel</a>.</p>\n<table><thead><tr><th>#</th><th>Book</th><th>Year</th><th>Gutenberg</th><th>Words</th></tr></thead><tbody>\n<tr><td>1</td><td>The Wonderful Wizard of Oz</td><td>1900</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/55\">55</a></td><td>39,700</td></tr>\n<tr><td>2</td><td>The Marvelous Land of Oz</td><td>1904</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/54\">54</a></td><td>42,000</td></tr>\n<tr><td>3</td><td>Ozma of Oz</td><td>1907</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/486\">486</a></td><td>38,800</td></tr>\n<tr><td>4</td><td>Dorothy and the Wizard in Oz</td><td>1908</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/420\">420</a></td><td>42,100</td></tr>\n<tr><td>5</td><td>The Road to Oz</td><td>1909</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/485\">485</a></td><td>39,400</td></tr>\n<tr><td>6</td><td>The Emerald City of Oz</td><td>1910</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/517\">517</a></td><td>53,900</td></tr>\n<tr><td>7</td><td>The Patchwork Girl of Oz</td><td>1913</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/955\">955</a></td><td>57,600</td></tr>\n<tr><td>8</td><td>Tik-Tok of Oz</td><td>1914</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/956\">956</a></td><td>48,400</td></tr>\n<tr><td>9</td><td>The Scarecrow of Oz</td><td>1915</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/957\">957</a></td><td>45,900</td></tr>\n<tr><td>10</td><td>Rinkitink in Oz</td><td>1916</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/958\">958</a></td><td>49,300</td></tr>\n<tr><td>11</td><td>The Lost Princess of Oz</td><td>1917</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/959\">959</a></td><td>47,300</td></tr>\n<tr><td>12</td><td>The Tin Woodman of Oz</td><td>1918</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/960\">960</a></td><td>44,500</td></tr>\n<tr><td>13</td><td>The Magic of Oz</td><td>1919</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/419\">419</a></td><td>39,500</td></tr>\n<tr><td>14</td><td>Glinda of Oz</td><td>1920</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/961\">961</a></td><td>39,700</td></tr>\n</tbody></table>\n<p>Together they come to 628,100 words, more than <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2600\"><em>War and Peace</em></a> in the Maude translation (565,400). No single book is long. The shortest, <em>Ozma of Oz</em>, is under 39,000 words, and the longest, <em>The Patchwork Girl of Oz</em>, under 58,000. W. W. Denslow illustrated the first book, and John R. Neill drew the pictures for the other thirteen.</p>\n<p>The gap between book 6 and book 7 is there in the text. <em>The Emerald City of Oz</em> closes with Glinda making Oz invisible and a last message from Dorothy, printed in capitals: “You will never hear anything more about Oz, because we are now cut off forever from all the rest of the world.” Three years later, the prologue of <em>The Patchwork Girl of Oz</em> explains how Baum, “after making six books”, got back in touch: a child asked why he couldn't hear from Princess Dorothy “by wireless telegraph”, so he put up a high tower in his back yard.</p>\n<p>Baum also wrote <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Little_Wizard_Stories_of_Oz\"><em>Little Wizard Stories of Oz</em></a>, six short tales printed as booklets in 1913 and collected in 1914, which <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/List_of_Oz_books\">sits outside the numbered list</a>. It is in the public domain on the same grounds.</p>\n<h2 id=\"after-baum-the-famous-forty\">After Baum: the Famous Forty</h2>\n<p>The series went on after Baum's death. Its first forty books are known as the “Famous Forty”, and the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/List_of_Oz_books\">Wikipedia list</a> numbers them by year: Baum wrote 1 to 14, Ruth Plumly Thompson 15 to 33 (1921 to 1939), Neill 34 to 36, Jack Snow 37 and 38, Rachel Cosgrove Payes 39, and Eloise Jarvis McGraw and Lauren Lynn McGraw 40. Thompson's first, <em>The Royal Book of Oz</em>, carried Baum's name, though the <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/30537\">Gutenberg page</a> notes that she wrote it entirely.</p>\n<p>In the US, Thompson's books split into three groups:</p>\n<ul>\n<li><strong>1921 to 1930</strong>, <em>The Royal Book of Oz</em> to <em>The Yellow Knight of Oz</em>: free, because everything published in the US before 1931 is out of copyright, as the <a rel=\"external\" href=\"https://guides.library.cornell.edu/copyright/publicdomain\">Cornell University Library chart</a> shows. <em>The Yellow Knight of Oz</em> joined the others on <a href=\"/blog/public-domain-day/\">1 January 2026</a>.</li>\n<li><strong>1931 to 1934</strong>, <em>Pirates in Oz</em>, <em>The Purple Prince of Oz</em>, <em>Ojo in Oz</em> and <em>Speedy in Oz</em>: none is on Project Gutenberg. A US book of those years is protected for 95 years if its copyright was renewed. If these were, they come free one a year, on 1 January 2027, 2028, 2029 and 2030.</li>\n<li><strong>1935 to 1939</strong>, <em>The Wishing Horse of Oz</em> to <em>Ozoplaning with the Wizard of Oz</em>: Project Gutenberg has all five and marks each one “Public domain in the USA”. Its <a rel=\"external\" href=\"https://www.gutenberg.org/help/copyright.html\">copyright rules</a> accept a US book from 1931 to 1963 only after a check of the renewal records, and the Cornell chart puts a book of those years without a renewal in the public domain.</li>\n</ul>\n<p>Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/search/?query=Ruth+Plumly+Thompson\">list of Thompson's books</a> has fifteen Oz titles: all ten from 1921 to 1930, and the five from 1935 to 1939.</p>\n<p><a href=\"/blog/standard-ebooks/\">Standard Ebooks</a> draws a stricter line. Its <a rel=\"external\" href=\"https://standardebooks.org/collections/oz\">Oz collection</a> lists the Famous Forty, <em>Oz Stories</em> by Baum, and two more books by Thompson. It marks every book from <em>Pirates in Oz</em> onwards as not in the public domain, including the five that Gutenberg has, and so far it has released only books 1 to 5. Books 7 to 9 are in progress, book 6 is on its list of wanted books, and books 10 to 14 are not in its catalogue yet.</p>\n<h2 id=\"outside-the-united-states\">Outside the United States</h2>\n<p><a href=\"/blog/public-domain-differs-by-country/\">Countries that count from the author's death</a> give a simpler answer. Baum's books have been free in the UK for decades, since <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">UK copyright</a> lasts 70 years after the author's death. <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/John_R._Neill\">Neill died in 1943</a>, so his drawings are free there too. <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Ruth_Plumly_Thompson\">Thompson died in 1976</a>, so the usual UK term of 70 years after the author's death would run to the end of 2046 for her books. Canada's <a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/acts/C-42/section-6.html\">Copyright Act, section 6</a>, gives the same 70 years and counts them to the end of the calendar year, so her books are free there on 1 January 2047.</p>\n<p>If you keep the series in a Readwyn library, set the series and its number on each book in “Edit details”. The switch “Group series” of the Sort menu then shows the books as one tile, and the page of the series marks each number you don't have yet, with an empty place “Not in this library” or, for a longer gap, one place such as “Books 4 to 9”.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-ao3-fanfiction-on-kindle-or-kobo/","url":"https://readwyn.com/blog/read-ao3-fanfiction-on-kindle-or-kobo/","title":"How to read AO3 fanfiction on a Kindle or a Kobo","summary":"Download the work from AO3 as an EPUB. A Kobo takes it over USB as it is; a Kindle gets it through Send to Kindle, or as a MOBI or AZW3 file over USB.","content_html":"<p>Each work on the Archive of Our Own has a Download button with five formats: <a rel=\"external\" href=\"https://archiveofourown.org/faq/downloading-fanworks\">AZW3, EPUB, MOBI, PDF, and HTML</a>. For a Kobo, take the EPUB and copy it over a USB cable. For a Kindle, either send the EPUB through Amazon's Send to Kindle, which converts it, or take the MOBI or AZW3 and copy that over USB.</p>\n<h2 id=\"on-a-kobo\">On a Kobo</h2>\n<p>Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">list of supported file formats</a> gives EPUB, PDF, and MOBI among its book formats, and HTML among its text formats. AZW3 isn't on that list, so leave it for the Kindle. Of the formats AO3 offers, take the EPUB: it's the format that Kobo's own steps for sideloading are written for, and it's the one that the KEPUB conversion below starts from.</p>\n<p>Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">steps for a computer</a> are the same on a Mac and on Windows:</p>\n<ol>\n<li>Connect the Kobo with its USB cable, and tap Connect when the Kobo shows “Computer detected”.</li>\n<li>Open the drive named KOBOeReader on the computer and drag the EPUB onto it.</li>\n<li>Eject the drive, then unplug the cable. The Kobo shows that it's importing content.</li>\n<li>Open My Books from the Home screen and tap the work's title.</li>\n</ol>\n<p>The same page warns that a book added this way “will only appear on the device that you used to import the content”, so it won't turn up in the Kobo app on your phone. Kobo recommends plugging the reader straight into the computer rather than into a hub. If you'd rather skip the cable, a Kobo can also <a href=\"/blog/send-books-to-kobo-with-dropbox-or-google-drive/\">pull books from Dropbox or Google Drive</a>, and if the EPUB doesn't appear in My Books, <a href=\"/blog/epub-not-showing-on-kobo/\">the causes and their order</a> are in a post of their own.</p>\n<p>A plain EPUB opens, but a Kobo also reads KEPUB, Kobo's own flavour of EPUB (<a href=\"/blog/kepub-or-epub-for-kobo/\">KEPUB or EPUB</a> has the difference). Readwyn, the library app that this blog belongs to, converts an EPUB to a KEPUB when its “Send to device” writes the book to a Kobo. Its option “Convert EPUB to KEPUB when sending” is on by default, and the hint under it reads “Kobo’s own format: better page numbers and reading statistics. The text stays the same.” That's useful if you keep your fanfiction next to the rest of your books; Readwyn doesn't fetch anything from AO3 itself.</p>\n<h2 id=\"on-a-kindle\">On a Kindle</h2>\n<p>A Kindle doesn't open an EPUB that you copy to it by cable. Vellum's note on Send to Kindle says it plainly: <a rel=\"external\" href=\"https://blog.vellum.pub/2022/05/send-to-kindle-now-accepts-epub-files/\">“Kindles can't read EPUB files”</a>, so a file that you copy over USB has to be in a Kindle format. That leaves two routes.</p>\n<p><strong>Send the EPUB to Amazon.</strong> Send to Kindle has taken EPUB files by email since 2022. <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2022/04/29/kindle-ereaders-now-support-epubs-sent-by-email/\">The eBook Reader reported</a> that Amazon converts each one to AZW3 on the way, and that the file has to be free of DRM. The Send to Kindle apps for Mac and Windows took EPUB from <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2022/12/20/send-to-kindle-apps-for-pc-and-mac-now-support-epub/\">December 2022</a>, and the Send to Kindle web page accepts files of up to 200 MB. Amazon's page on <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TrpJES2mSHEg7bKinx\">sending documents</a> says they go to your Kindle library, on your Kindles and in the Kindle app. The limits of each route, and why a file can fail, are in <a href=\"/blog/send-epub-to-kindle/\">can you send an EPUB to a Kindle</a>.</p>\n<p>Don't send AO3's MOBI download this way. Amazon told users in 2023 that Send to Kindle would stop accepting MOBI, AZW, and PRC files <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2023/09/25/send-to-kindle-losing-mobi-support-probably-for-real-this-time\">by 20 December 2023</a>, by email, in the phone apps, and in the desktop apps.</p>\n<p><strong>Copy a MOBI or an AZW3 over USB.</strong> The same September 2023 report says that Kindles still read MOBI files copied by cable. The AZW3 download is KF8, the newer Kindle format, which the <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/AZW3\">MobileRead wiki</a> says every Kindle reads except the first and second generations and the Kindle DX; the Kindle 3 needs a software update, and a Kindle 4 needs firmware 4.1.0 or later. If you own one of those older models, take the MOBI.</p>\n<p>Drag the file into the Kindle's <code>documents</code> folder. On a Mac, a Kindle Scribe or a Kindle released in 2024 doesn't show up as a drive: <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's page on USB transfers</a> has you install Send to Kindle for Mac and use its USB File Manager (Tools &gt; USB File Manager). These models <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2024/11/01/how-to-connect-kindles-to-macs-to-transfer-connect-via-usb/\">use MTP</a>, which a Mac doesn't support natively. <a href=\"/blog/mobi-azw3-or-epub-for-kindle-usb/\">Which format to put on a Kindle over USB</a> compares the two in more detail, and <a href=\"/blog/kindle-not-showing-up-on-mac/\">why a Kindle doesn't show up on a Mac</a> covers an older Kindle that should mount and doesn't.</p>\n<h2 id=\"a-work-that-isn-t-finished\">A work that isn't finished</h2>\n<p>A download is a copy of the work as it stood when you saved it. When the author posts a new chapter, the file on your e-reader stays as it was: download the work again and replace the old copy.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/sherlock-holmes-books-in-order/","url":"https://readwyn.com/blog/sherlock-holmes-books-in-order/","title":"The Sherlock Holmes books in order, and where to get them free","summary":"Arthur Conan Doyle wrote nine Sherlock Holmes books, four novels and five story collections, from 1887 to 1927. All nine are free from Project Gutenberg and Standard Ebooks.","content_html":"<p>Arthur Conan Doyle wrote nine Sherlock Holmes books: four novels and five collections of short stories, which came out between 1887 and 1927. In order of publication they are <em>A Study in Scarlet</em>, <em>The Sign of the Four</em>, <em>The Adventures of Sherlock Holmes</em>, <em>The Memoirs of Sherlock Holmes</em>, <em>The Hound of the Baskervilles</em>, <em>The Return of Sherlock Holmes</em>, <em>The Valley of Fear</em>, <em>His Last Bow</em> and <em>The Case-Book of Sherlock Holmes</em>. Every one is free as an ebook from <a href=\"/blog/what-is-project-gutenberg/\">Project Gutenberg</a> and from <a href=\"/blog/standard-ebooks/\">Standard Ebooks</a>, whose <a rel=\"external\" href=\"https://standardebooks.org/collections/sherlock-holmes\">Sherlock Holmes collection</a> numbers them 1 to 9 in that order.</p>\n<h2 id=\"the-nine-books\">The nine books</h2>\n<p>The years are those of first publication in full: in a magazine for <em>A Study in Scarlet</em> (<em>Beeton's Christmas Annual</em>) and <em>The Sign of the Four</em> (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Sign_of_the_Four\"><em>Lippincott's</em>, February 1890</a>), and as a book for the other seven (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Canon_of_Sherlock_Holmes\">the canon on Wikipedia</a>). The <em>Memoirs</em> came out <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Memoirs_of_Sherlock_Holmes\">late in 1893, dated 1894</a>. The word counts are of the Project Gutenberg text of each book, and each title links to its Gutenberg page. The time is what each book takes at 250 words a minute.</p>\n<table><thead><tr><th>#</th><th>Book</th><th>Year</th><th>Kind</th><th>Words</th><th>At 250 wpm</th></tr></thead><tbody>\n<tr><td>1</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/244\">A Study in Scarlet</a></td><td>1887</td><td>Novel</td><td>43,500</td><td>2 h 54 min</td></tr>\n<tr><td>2</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2097\">The Sign of the Four</a></td><td>1890</td><td>Novel</td><td>43,200</td><td>2 h 53 min</td></tr>\n<tr><td>3</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1661\">The Adventures of Sherlock Holmes</a></td><td>1892</td><td>12 stories</td><td>104,700</td><td>6 h 59 min</td></tr>\n<tr><td>4</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/834\">The Memoirs of Sherlock Holmes</a></td><td>1894</td><td>11 or 12 stories</td><td>95,700</td><td>6 h 23 min</td></tr>\n<tr><td>5</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2852\">The Hound of the Baskervilles</a></td><td>1902</td><td>Novel</td><td>59,400</td><td>3 h 58 min</td></tr>\n<tr><td>6</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/108\">The Return of Sherlock Holmes</a></td><td>1905</td><td>13 stories</td><td>112,500</td><td>7 h 30 min</td></tr>\n<tr><td>7</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/3289\">The Valley of Fear</a></td><td>1915</td><td>Novel</td><td>58,100</td><td>3 h 52 min</td></tr>\n<tr><td>8</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2350\">His Last Bow</a></td><td>1917</td><td>7 or 8 stories</td><td>59,200</td><td>3 h 57 min</td></tr>\n<tr><td>9</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/69700\">The Case-Book of Sherlock Holmes</a></td><td>1927</td><td>12 stories</td><td>83,200</td><td>5 h 33 min</td></tr>\n</tbody></table>\n<p>Together that's about 660,000 words, or about 44 hours of reading at that pace. The 250 is the speed that Readwyn starts each library with, before your own page turns <a href=\"/blog/reading-speed-words-per-minute/\">take over</a>; a faster or slower reader can scale the column.</p>\n<p><em>A Study in Scarlet</em> first appeared in <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/A_Study_in_Scarlet\"><em>Beeton's Christmas Annual</em> for 1887</a>, the first appearance of Holmes and Watson. <em>The Hound of the Baskervilles</em> ran in <em>The Strand Magazine</em> from August 1901 to April 1902 before George Newnes printed it as a book in <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Hound_of_the_Baskervilles\">March 1902</a>, and <em>The Valley of Fear</em> ran from September 1914 to May 1915, with the first British book edition on <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Valley_of_Fear\">3 June 1915</a>. The 56 short stories also came out in magazines first; those of the <em>Case-Book</em> appeared from October 1921 to April 1927.</p>\n<h2 id=\"read-them-in-publication-order\">Read them in publication order</h2>\n<p>Publication order keeps the break that Doyle built into the series. He killed Holmes off in “The Final Problem”, the last story of the <em>Memoirs</em>, published in December 1893. He brought him back in “The Adventure of the Empty House”, the first story of the <em>Return</em>, which <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Adventure_of_the_Empty_House\"><em>The Strand</em> printed in October 1903</a> and where the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Arthur_Conan_Doyle\">death turns out to have been faked</a>. Read in order, the <em>Memoirs</em> end with a death and the <em>Return</em> opens with the surprise.</p>\n<p>Two of the novels sit outside that sequence. <em>The Hound of the Baskervilles</em> came out after Holmes's death in print but is <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Hound_of_the_Baskervilles\">set in 1889</a>, before it. <em>The Valley of Fear</em>, published in 1915, opens with a cipher from a <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Valley_of_Fear\">double agent of Professor Moriarty</a>, the villain of “The Final Problem”. Neither spoils the return, so they read well where they fall. For series where the two orders pull further apart, see <a href=\"/blog/find-the-reading-order-of-a-series/\">how to find the reading order of a series</a>.</p>\n<h2 id=\"the-story-that-moved\">The story that moved</h2>\n<p>“The Adventure of the Cardboard Box” is the reason two rows of the table give two story counts. The first British edition of the <em>Memoirs</em> left it out, and the first American edition printed it but soon dropped it. Doyle reportedly asked for that, <a rel=\"external\" href=\"https://standardebooks.org/ebooks/arthur-conan-doyle/the-memoirs-of-sherlock-holmes\">Standard Ebooks says</a>, because its theme of adultery made it unsuitable for younger readers. It came back in 1917 in <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Adventure_of_the_Cardboard_Box\"><em>His Last Bow</em></a>. Editions still disagree about where it goes.</p>\n<p>The two free libraries have also made different choices. On Project Gutenberg the story is in neither book: the contents of <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/834\">the <em>Memoirs</em></a> and of <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2350\"><em>His Last Bow</em></a> leave it out, and it has <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2344\">an ebook of its own</a>. So the word counts above don't include it. Standard Ebooks puts it back in <a rel=\"external\" href=\"https://standardebooks.org/ebooks/arthur-conan-doyle/the-memoirs-of-sherlock-holmes\">its <em>Memoirs</em></a>, as most modern British editions do, to return it to its place in the order. If you read the Gutenberg files, add that separate ebook to the <em>Memoirs</em>.</p>\n<h2 id=\"which-free-edition-to-get\">Which free edition to get</h2>\n<p>Gutenberg keeps plain text and HTML as the master files of a book, and its EPUB and Kindle files are <a rel=\"external\" href=\"https://www.gutenberg.org/help/file_formats.html\">generated automatically</a> from them. Standard Ebooks starts from an existing transcription, often Project Gutenberg's, proofreads it against scans of a printed edition, and gives each book a cover and a kepub file for Kobo next to the EPUB and the azw3 for Kindle. Its <a rel=\"external\" href=\"https://standardebooks.org/ebooks/arthur-conan-doyle/the-casebook-of-sherlock-holmes\"><em>Casebook</em></a> is an exception to the first part: the transcription came from Project Gutenberg Australia, and the scans from Google Books. Standard Ebooks spells the last title <em>The Casebook of Sherlock Holmes</em>; Gutenberg keeps the hyphen. <a href=\"/blog/kepub-or-epub-for-kobo/\">KEPUB or EPUB</a> explains what the Kobo file adds.</p>\n<h2 id=\"why-the-last-book-is-free-too\">Why the last book is free too</h2>\n<p>Doyle <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Arthur_Conan_Doyle\">died on 7 July 1930</a>. In the UK, a written work stays in copyright for <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">70 years after the author's death</a>, so all of Holmes went out of copyright there <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Sherlock_Holmes#Copyright_issues\">at the end of 2000</a>. It had already happened once, at the end of 1980 under the old 50-year term, before the longer term of 1996 brought the copyright back. The United States counts from publication instead, and the <em>Case-Book</em> of 1927 held out longest. Duke's Center for the Study of the Public Domain marked its release on <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2023/\">1 January 2023</a>, the day that all of Doyle's Holmes works became public domain in the US. Gutenberg released its edition on 3 January 2023.</p>\n<p>Duke's page also tells the story of the fight before that day. Doyle's estate had asked for licence fees for the characters even though most of the stories were already free. The editor Leslie Klinger went to court, a 2013 ruling held that the elements of the free stories are “free for public use”, and the appeals court agreed and awarded him his legal fees. Judge Posner wrote that it was “time the estate, in its own self-interest, changed its business model.” If you live outside the US and the UK, the <a href=\"/blog/public-domain-differs-by-country/\">rules of your own country</a> decide.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/short-classics-to-read-in-a-day/","url":"https://readwyn.com/blog/short-classics-to-read-in-a-day/","title":"Short classics you can read in a day","summary":"Twelve classics of 6,100 to 38,500 words, from The Yellow Wallpaper to Heart of Darkness, each one under three hours at 250 words a minute.","content_html":"<p>Each of the twelve books below takes less than three hours at 250 words a minute, so any one of them fits into a free afternoon with room for a break. The shortest, The Yellow Wallpaper, is about 6,100 words and takes under half an hour. The longest, Heart of Darkness, is about 38,500 words and takes 2 hours 34 minutes. All twelve have a free text at Project Gutenberg, linked from the table.</p>\n<table><thead><tr><th>Book</th><th>Author</th><th>Words</th><th>At 250 wpm</th><th>At 300 wpm</th><th>At 350 wpm</th></tr></thead><tbody>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1952\">The Yellow Wallpaper</a></td><td>Charlotte Perkins Gilman</td><td>6,100</td><td>24 min</td><td>20 min</td><td>17 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/41\">The Legend of Sleepy Hollow</a></td><td>Washington Irving</td><td>12,200</td><td>49 min</td><td>41 min</td><td>35 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/11231\">Bartleby, the Scrivener</a></td><td>Herman Melville</td><td>14,500</td><td>58 min</td><td>48 min</td><td>41 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/844\">The Importance of Being Earnest</a></td><td>Oscar Wilde</td><td>20,700</td><td>1 h 23 min</td><td>1 h 9 min</td><td>59 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/5200\">Metamorphosis</a>, translated by David Wyllie</td><td>Franz Kafka</td><td>22,000</td><td>1 h 28 min</td><td>1 h 13 min</td><td>1 h 3 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/43\">Strange Case of Dr Jekyll and Mr Hyde</a></td><td>Robert Louis Stevenson</td><td>25,700</td><td>1 h 43 min</td><td>1 h 26 min</td><td>1 h 13 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/10007\">Carmilla</a></td><td>Joseph Sheridan Le Fanu</td><td>28,200</td><td>1 h 53 min</td><td>1 h 34 min</td><td>1 h 21 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/46\">A Christmas Carol</a></td><td>Charles Dickens</td><td>28,600</td><td>1 h 54 min</td><td>1 h 35 min</td><td>1 h 22 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/215\">The Call of the Wild</a></td><td>Jack London</td><td>31,900</td><td>2 h 8 min</td><td>1 h 46 min</td><td>1 h 31 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/35\">The Time Machine</a></td><td>H. G. Wells</td><td>32,600</td><td>2 h 10 min</td><td>1 h 49 min</td><td>1 h 33 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/4517\">Ethan Frome</a></td><td>Edith Wharton</td><td>34,800</td><td>2 h 19 min</td><td>1 h 56 min</td><td>1 h 39 min</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/219\">Heart of Darkness</a></td><td>Joseph Conrad</td><td>38,500</td><td>2 h 34 min</td><td>2 h 8 min</td><td>1 h 50 min</td></tr>\n</tbody></table>\n<p>The word counts come from Gutenberg's plain-text files, counted as <a href=\"/blog/how-many-words-in-a-novel/\">how many words are in a novel</a> describes and rounded to the nearest hundred. The three speeds are the ones that <a href=\"/blog/how-long-is-war-and-peace/\">How long is War and Peace?</a> uses: 250 is the starting figure of <a href=\"/blog/reading-speed-words-per-minute/\">reading speed in words per minute</a>, and 300 and 350 appear in its examples.</p>\n<h2 id=\"what-each-one-is\">What each one is</h2>\n<p><strong>The Yellow Wallpaper.</strong> A woman's physician husband confines her to a nursery for \"temporary nervous depression\". Forbidden to work or write, she keeps a secret journal, and the pattern of the wallpaper takes over her mind.</p>\n<p><strong>The Legend of Sleepy Hollow.</strong> Ichabod Crane, a superstitious schoolmaster, competes with Brom Bones for Katrina Van Tassel, hears ghost stories of the Headless Horseman at a harvest party, and meets a cloaked rider on his way home.</p>\n<p><strong>Bartleby, the Scrivener.</strong> A Wall Street lawyer hires a copying clerk who answers more and more of his requests with \"I would prefer not to.\"</p>\n<p><strong>The Importance of Being Earnest.</strong> The only play on the list: two young men each use the name Ernest to get out of their social duties, and the farce mocks Victorian society through characters such as Lady Bracknell.</p>\n<p><strong>Metamorphosis.</strong> Gregor Samsa, a travelling salesman, wakes up as a monstrous insect, and his family has to live with him. Gutenberg's page marks the David Wyllie translation as copyrighted and refers to a notice inside the book for the terms.</p>\n<p><strong>Strange Case of Dr Jekyll and Mr Hyde.</strong> The lawyer Utterson looks into the link between his friend Dr Jekyll and a violent man called Edward Hyde.</p>\n<p><strong>Carmilla.</strong> A mysterious guest arrives at Laura's remote castle in Austria after a carriage accident, and young women nearby start to fall ill. Gutenberg's page notes that it came out 25 years before Dracula.</p>\n<p><strong>A Christmas Carol.</strong> Four ghosts, Jacob Marley first, visit the miser Ebenezer Scrooge, who scorns Christmas.</p>\n<p><strong>The Call of the Wild.</strong> Buck, a dog stolen from a comfortable home in California, is sold as a sled dog in the Klondike and slowly answers an older instinct.</p>\n<p><strong>The Time Machine.</strong> A Victorian scientist travels to the year 802,701 and finds humanity split into the Eloi and the Morlocks.</p>\n<p><strong>Ethan Frome.</strong> A man in a loveless marriage in wintry Massachusetts falls for his wife's young cousin.</p>\n<p><strong>Heart of Darkness.</strong> Marlow, a seaman, tells of his trip into the African interior as the captain of a riverboat, sent to find the ivory trader Kurtz.</p>\n<h2 id=\"short-story-novelette-or-novella\">Short story, novelette, or novella</h2>\n<p>The Hugo Awards put exact numbers on these words. The <a rel=\"external\" href=\"https://www.wsfs.org/wp-content/uploads/2026/06/WSFS-Constitution-as-of-2025-10-07-Clean_20260606.pdf\">WSFS Constitution</a> (sections 3.3.1 to 3.3.4) defines a short story as under 7,500 words, a novelette as 7,500 to 17,500, a novella as 17,500 to 40,000, and a novel as 40,000 or more. The rules are written for science fiction and fantasy, but the lines are a handy scale for any fiction:</p>\n<ul>\n<li>The Yellow Wallpaper is a short story.</li>\n<li>The Legend of Sleepy Hollow and Bartleby are novelettes.</li>\n<li>The other eight works of fiction, from Metamorphosis to Heart of Darkness, are novellas.</li>\n</ul>\n<p>So the line for a day of reading falls close to the line for a novel. Heart of Darkness is 1,500 words short of 40,000. <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/64317\">The Great Gatsby</a>, at about 48,600 words in Gutenberg's text, is a short novel but already 3 hours 14 minutes at 250 words a minute: still a day's reading, but more than an afternoon.</p>\n<h2 id=\"all-twelve-in-a-row\">All twelve in a row</h2>\n<p>The twelve counts add up to 295,800 words, or 19 hours 43 minutes at 250 words a minute. At <a href=\"/blog/books-a-year-at-30-minutes-a-day/\">half an hour a day</a>, that is 40 days, or about two books a week. The total is also smaller than <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/145\">Middlemarch</a> alone, which comes to about 318,600 words in Gutenberg's text.</p>\n<p>A table of minutes is only a guide. If you read them in Readwyn, the book screen shows the time left in each one from its word count and the speed measured from your own page turns (<a href=\"/blog/how-long-to-read-a-book/\">how long will this book take to read</a>).</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/three-musketeers-books-in-order/","url":"https://readwyn.com/blog/three-musketeers-books-in-order/","title":"The Three Musketeers books in order","summary":"Dumas wrote three d'Artagnan novels: The Three Musketeers, Twenty Years After, and The Vicomte de Bragelonne, which English editions split into three to five books.","content_html":"<p>Alexandre Dumas wrote three novels about d'Artagnan and the musketeers, and the order of publication is also the order of the story: <em>The Three Musketeers</em> (1844), <em>Twenty Years After</em> (1845), and <em>The Vicomte de Bragelonne</em> (1847 to 1850). The confusion starts with the third. It's so long that English publishers cut it into three, four, or five books, and one of those books is <em>The Man in the Iron Mask</em>.</p>\n<h2 id=\"the-three-novels\">The three novels</h2>\n<table><thead><tr><th>Order</th><th>Novel</th><th>French title</th><th>Serialised</th><th>Story set in</th></tr></thead><tbody>\n<tr><td>1</td><td>The Three Musketeers</td><td>Les Trois Mousquetaires</td><td>March to July 1844</td><td>1625 to 1628</td></tr>\n<tr><td>2</td><td>Twenty Years After</td><td>Vingt ans après</td><td>January to August 1845</td><td>1648 to 1649</td></tr>\n<tr><td>3</td><td>The Vicomte de Bragelonne</td><td>Le Vicomte de Bragelonne ou Dix ans plus tard</td><td>October 1847 to January 1850</td><td>1660 to 1673</td></tr>\n</tbody></table>\n<p><em>The Three Musketeers</em> first ran in the newspaper <em>Le Siècle</em>, according to <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Three_Musketeers\">Wikipedia's article on the novel</a>, which also names Auguste Maquet as the collaborator who researched and outlined plots that Dumas then wrote up. The serial dates of the other two come from the article on <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_d%27Artagnan_Romances\">the d'Artagnan Romances</a>, the full French title of the third from <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Vicomte_of_Bragelonne:_Ten_Years_Later\">its own article</a>, and the years of the story from the index that Project Gutenberg puts at the top of its <em>Twenty Years After</em>.</p>\n<p>The titles count from different points. <em>Twenty Years After</em> is twenty years after the first book. The subtitle of the third, “Ten Years Later”, is ten years after the second, not after the first.</p>\n<h2 id=\"how-the-third-novel-gets-split\">How the third novel gets split</h2>\n<p><a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_d%27Artagnan_Romances\">Wikipedia lists the usual cuts</a>. A three-volume edition has <em>The Vicomte de Bragelonne</em>, <em>Louise de la Vallière</em>, and <em>The Man in the Iron Mask</em>. A four-volume edition puts <em>Ten Years Later</em> second and pushes the other two to third and fourth. A five-volume edition usually leaves the smaller parts without titles of their own.</p>\n<p>So <em>The Man in the Iron Mask</em> isn't a separate sequel. It's the last part of <em>The Vicomte de Bragelonne</em>, and reading it on its own means starting a story near its end. The title <em>Ten Years Later</em> is the least reliable: it names the second of four parts, but it is also the subtitle of the whole novel.</p>\n<h2 id=\"the-free-editions-and-their-word-counts\">The free editions and their word counts</h2>\n<p>Project Gutenberg follows the four-volume split. The six files below make up the series, and its <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1259\"><em>Twenty Years After</em></a> opens with a table of them that numbers the chapters of the third novel straight through, from 1 to 269. None of the six files names its translator.</p>\n<table><thead><tr><th>Order</th><th>Book</th><th>Gutenberg ebook</th><th>Words</th></tr></thead><tbody>\n<tr><td>1</td><td>The Three Musketeers</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1257\">1257</a></td><td>230,300</td></tr>\n<tr><td>2</td><td>Twenty Years After</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1259\">1259</a></td><td>243,000</td></tr>\n<tr><td>3</td><td>The Vicomte de Bragelonne</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2609\">2609</a></td><td>190,500</td></tr>\n<tr><td>4</td><td>Ten Years Later</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2681\">2681</a></td><td>187,900</td></tr>\n<tr><td>5</td><td>Louise de la Vallière</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2710\">2710</a></td><td>168,600</td></tr>\n<tr><td>6</td><td>The Man in the Iron Mask</td><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2759\">2759</a></td><td>174,600</td></tr>\n</tbody></table>\n<p>The counts cover everything between Gutenberg's start and end lines, chapter headings and the transcriber's note at the top of each part included, rounded to the nearest hundred (the method is in <a href=\"/blog/how-many-words-in-a-novel/\">How many words are in a novel?</a>). The four parts of the third novel come to 721,500 words, and the six files to 1,194,800.</p>\n<p>That makes <em>The Vicomte de Bragelonne</em> longer on its own than Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2600\"><em>War and Peace</em></a> in the translation of Louise and Aylmer Maude, which counts 565,400 words by the same method. At 250 words a minute, the <a href=\"/blog/reading-speed-words-per-minute/\">starting speed that Readwyn assumes</a> before it has measured yours, <em>The Three Musketeers</em> takes about 15 hours, <em>Twenty Years After</em> about 16, and <em>The Vicomte de Bragelonne</em> about 48. The whole series is about 80 hours.</p>\n<p>Watch out for one more file. Gutenberg also has an older <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1258\"><em>Ten Years Later</em></a>, ebook 1258, released in 1998. Gutenberg's index says that it holds all of ebook 2609 and the first 28 chapters of ebook 2681, so a reader who takes it and then 2681 reads those 28 chapters twice.</p>\n<p><a href=\"/blog/standard-ebooks/\">Standard Ebooks</a> has <a rel=\"external\" href=\"https://standardebooks.org/collections/dartagnan-romances\">all three novels</a> in its d'Artagnan Romances collection, and it keeps the third novel whole: its page for <em>The Vicomte de Bragelonne</em> says, “This edition is presented in a single ebook.” It credits the translations of its three editions to William Robson, to Estes and Lauriat, and to George Routledge and Co.</p>\n<h2 id=\"keeping-the-six-files-in-order\">Keeping the six files in order</h2>\n<p>On a shelf of downloaded files, the four parts of the third novel sort by title, not by story. In Readwyn, the series field and its number in Edit details fix that: give all six files the same series and the numbers 1 to 6. With “Group series” on in the Sort menu, the six show as one tile, and a click on it opens the page of the series with the books in order. A number that you haven't added yet shows as “Not in this library”. If you'd rather number the three novels than the six files, <a href=\"/blog/find-the-reading-order-of-a-series/\">How to find the reading order of a book series</a> gives what an EPUB file can record about a book's place in a series.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-open-library/","url":"https://readwyn.com/blog/what-is-open-library/","title":"What is Open Library?","summary":"Open Library is the Internet Archive's free, editable catalogue of books, with a page for each edition. It also lends scans, and that part is the one the courts limited.","content_html":"<p>Open Library is a free catalogue of books run by the Internet Archive, a non-profit library. Its <a rel=\"external\" href=\"https://openlibrary.org/help/faq/about\">FAQ</a> calls it \"a universal catalog for book metadata\" that anyone can edit, and its aim is a web page for every book ever published. On top of the catalogue sit two other services: free downloads of public-domain books, and loans of scanned books that are still in copyright. A US court case did not touch the catalogue or the public-domain downloads. It was about the loans, and since the case more than 500,000 books have been taken out of lending.</p>\n<h2 id=\"a-catalogue-first\">A catalogue first</h2>\n<p>A book on Open Library is two kinds of record, not one. The <a rel=\"external\" href=\"https://openlibrary.org/dev/docs/api/books\">Books API documentation</a> separates a <strong>work</strong>, \"a logical collection of similar Editions\", from an <strong>edition</strong>, the particular printing with its own ISBN. A work has an ID ending in W (OL45804W), an edition one ending in M (OL7353617M), and an address such as <code>openlibrary.org/isbn/9780140328721</code> jumps from an ISBN straight to its edition.</p>\n<p>The scale is large. In November 2026 the <a rel=\"external\" href=\"https://openlibrary.org/stats\">stats page</a> counted nearly 57 million editions, nearly 42 million works and nearly 15.5 million authors. The <a rel=\"external\" href=\"https://openlibrary.org/about\">about page</a> still says \"over 20 million records\", less than half the number of editions on the stats page.</p>\n<p>Every record has an “Edit” button. The early catalogue came from libraries that gave their records to the public domain, according to the <a rel=\"external\" href=\"https://openlibrary.org/help/faq/using\">FAQ on using the data</a>, and contributors are asked to license their edits under CC0, so others can reuse them freely. The records are published as <a rel=\"external\" href=\"https://openlibrary.org/developers/dumps\">monthly data dumps</a>, and the code of the site is on <a rel=\"external\" href=\"https://github.com/internetarchive/openlibrary\">GitHub</a> under the GNU Affero General Public License, version 3.</p>\n<h2 id=\"what-you-can-read\">What you can read</h2>\n<p>The FAQ says Open Library offers \"over 1.7 million public domain books\" in PDF, ePub, DAISY, DjVu, and plain text. <a href=\"/blog/what-is-project-gutenberg/\">Project Gutenberg</a> and <a href=\"/blog/standard-ebooks/\">Standard Ebooks</a> also give away classics.</p>\n<p>Borrowing works like a library with one copy on the shelf. The <a rel=\"external\" href=\"https://openlibrary.org/help/faq/borrow\">borrowing FAQ</a> says a book can be lent \"by one patron at a time from anywhere in the world for free\". A loan lasts one hour, for browsing, or 14 days if the book is fully borrowable, and you can have ten books out at once. If someone else has the book, you can join a waiting list; when it comes back you get an email and 24 hours to take it. Borrowed books open in the reader in your browser, and are usually also offered as PDF and EPUB.</p>\n<h2 id=\"is-open-library-legal\">Is Open Library legal?</h2>\n<p>For the catalogue, yes: its records are in the public domain or under CC0. The free downloads are books that Open Library lists as public domain, but a book can be <a href=\"/blog/public-domain-differs-by-country/\">public domain in one country and not another</a>, so a reader outside the US should check the rules where they live.</p>\n<p>Lending in-copyright books is the part that went to court. The Internet Archive added lending to Open Library in June 2010, <a rel=\"external\" href=\"https://blog.archive.org/2010/06/28/digital-lending-library/\">with two-week loans</a>. It later defended the lending of its scans with a theory called controlled digital lending, which ties the number of scans on loan to the print copies it owns. In 2020 four publishers sued. The Archive lost <em>Hachette v. Internet Archive</em>, and on 4 December 2024 it <a rel=\"external\" href=\"https://blog.archive.org/2024/12/04/end-of-hachette-v-internet-archive/\">said</a> it would not ask the Supreme Court to review the case. The post on <a href=\"/blog/controlled-digital-lending/\">controlled digital lending</a> gives the history and what the courts held.</p>\n<p>The result shows on the Internet Archive as books marked “Borrow Unavailable”. Its <a rel=\"external\" href=\"https://help.archive.org/help/why-are-so-many-books-listed-as-borrow-unavailable-at-the-internet-archive/\">help page</a> says more than 500,000 books have been taken out of lending because of the suit and the publishers' requests that followed, though they stay available to readers with print disabilities.</p>\n<h2 id=\"open-library-behind-other-apps\">Open Library behind other apps</h2>\n<p>Other apps can read the catalogue through Open Library's <a rel=\"external\" href=\"https://openlibrary.org/developers/api\">API</a>. It allows one request a second from an anonymous client, and three from one that names itself in the User-Agent header with the app's name and a contact email or phone number. It also asks that nobody bulk download through the API, since the dumps exist for that.</p>\n<p>Readwyn is one of those apps. “Fetch metadata” asks Open Library first and Google Books second for a book's cover, description, page count and genres, and each request names the app as <code>Readwyn/&lt;version&gt; (https://readwyn.com)</code>. The post on <a href=\"/blog/fetch-covers-and-descriptions/\">fetching covers and descriptions</a> shows what the fetch fills in.</p>\n","date_published":"2026-11-17T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/before-you-sell-a-kindle-or-kobo/","url":"https://readwyn.com/blog/before-you-sell-a-kindle-or-kobo/","title":"What to do before you sell or give away a Kindle or a Kobo","summary":"Copy off the files you added by cable, sync once more, then factory reset it. Don't reset or deregister a Kindle released in 2012 or earlier: it can't be used again.","content_html":"<p>Three things, in this order. Copy off everything that exists only on the device: the books and documents you put on it with a cable, and the highlights you made in them. Sync it over Wi-Fi one last time, so the store has your latest notes and page. Then do a factory reset, which wipes the device and takes it off your account. The one exception is an old Kindle: Amazon says not to reset or deregister a Kindle released in 2012 or earlier, because it can never be registered again. Readwyn, the ebook library app for Mac and Linux, can copy the highlights and reading progress of its books off a Kobo before the wipe, and off any Kindle that opens on a Mac as a drive: every Kindle released before 2024 except the Scribe, <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon says</a>.</p>\n<h2 id=\"what-a-reset-takes-with-it\">What a reset takes with it</h2>\n<p>A reset removes the device's copy of everything on it. What comes back afterwards is only what an account holds.</p>\n<p>On a Kindle, the <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">Kindle Colorsoft user guide</a> says Factory Reset \"will remove all downloaded or transferred content\", and it adds that you \"must back up any personal documents that you'd like to keep\" first. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605114-Reset-your-Kobo-eReader\">resetting an eReader</a> is just as plain: books you bought from Kobo can be downloaded again, but PDF and EPUB files that you added from your computer will be deleted, and so will bookmarks and annotations made since your last sync.</p>\n<p>So connect the device to your computer and copy its files to a folder there. On a Mac, a Kindle Scribe or a Kindle released in 2024 or later doesn't open as a drive: <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's page on USB transfers</a> has you install Send to Kindle for Mac and copy the files with its USB File Manager. On a Kindle, the books you added are in the <code>documents</code> folder, along with <code>My Clippings.txt</code>, the running list of every highlight and note you made (<a href=\"/blog/export-kindle-highlights/\">where Kindle highlights are kept</a>). Those books need the copy most. The same guide says files moved by cable \"are not stored in the Cloud\", and that highlights sync only for personal documents \"archived in Kindle format\", so the highlights in them live on that one Kindle (<a href=\"/blog/does-kindle-sync-sideloaded-books/\">what a Kindle syncs for such books</a>).</p>\n<h2 id=\"sync-then-reset-a-kindle\">Sync, then reset a Kindle</h2>\n<p>Make sure the Kindle is online, swipe down from the home screen to open the Quick Actions (or select Menu), and choose the Sync icon or “Sync and Check for Items”, as <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GDCAMDFMC2LZP6BR\">Amazon's page on syncing</a> describes. Amazon says a sync backs up your furthest page read. Notes and highlights go to your account too if Whispersync for Books is on (All Settings, Device Options, Whispersync), as the Colorsoft user guide says.</p>\n<p>Then reset. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G4EWAYLEKM2UGFAC\">Amazon's steps</a> go through Quick Actions or Menu, then Settings or All Settings, then Device Options, then Factory Reset. Older Kindles call the option Reset Device. Select Yes to confirm. The same page says the reset also deregisters the Kindle from your Amazon account, so the next owner has to register it to their own.</p>\n<p>Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GP383NNSU4NSHP7C\">page on deregistering</a> gives the other route: on the Kindle, Settings, then Your Account (My Account on some models), then Deregister Device; or on the website, Manage Your Devices, the Kindle, and “Deregister”. The website is the way if the Kindle has already left your hands, unless it is a Kindle from 2012 or earlier (below). Deregistering removes your account from the device, and the same page recommends a factory reset on top of it before you give a Kindle away or sell it, to clear the content and the settings as well.</p>\n<h2 id=\"a-kindle-from-2012-or-earlier\">A Kindle from 2012 or earlier</h2>\n<p>Leave its registration alone. Amazon ended support for these Kindles on 20 May 2026, and its <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TRXsYxKJr4WTdsVs2P\">page on them</a> is blunt: after a deregistration or a factory reset you won't be able to register it again \"or use the device in any way\". The books already on it still open \"as long as you remain logged in\". So such a Kindle can only change hands still signed in to your account: copy off what you want to keep, delete from it what you don't want to hand over, and don't reset it. <a href=\"/blog/old-kindle-end-of-support/\">Why an old Kindle can't reach the store</a> has more on these models.</p>\n<h2 id=\"sync-then-reset-a-kobo\">Sync, then reset a Kobo</h2>\n<p>On the Home screen, tap the Sync icon at the top and then “Sync now”. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/articles/360017513334\">syncing over Wi-Fi</a> says this \"updates any bookmarks or annotations that you've made to your eBooks\".</p>\n<p>Then go to More at the bottom of the Home screen, Settings, Device information, and under Advanced tap Factory reset, then Reset now. The reset signs you out of your Kobo account and returns every setting to how it came.</p>\n<p>Don't sign out by hand first as a precaution. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605094-Sign-out-of-your-Kobo-eReader\">signing out</a> warns that signing out keeps your books and bookmarks, but you lose the highlights, notes, and annotations you added.</p>\n<p>Two extras are linked to the account on the other side, not only on the Kobo. If you used Dropbox to send books, sign in at Dropbox's connected apps page, tap Rakuten Kobo, and tap Disconnect, as Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360033882694-Turn-off-Dropbox-sharing-on-your-eReader\">turning off Dropbox</a> says. For Google Drive, Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/15335985512983-Add-books-to-your-eReader-using-Google-Drive\">Google Drive page</a> has you open Drive's settings, click Manage Apps, find Kobo, click Options, and choose Disconnect from Drive.</p>\n<h2 id=\"readwyn-before-and-after\">Readwyn before and after</h2>\n<p>If the books are in a Readwyn library, connect the e-reader before the reset, open “Sync”, and apply “Device → Readwyn”. The highlights, notes, and reading progress of each book that is both on the device and in the library come across and stay after the device has gone (<a href=\"/blog/kindle-highlights-in-ebook-library/\">Kindle highlights</a>, <a href=\"/blog/export-kobo-highlights-on-mac/\">Kobo highlights</a>). Afterwards, Settings › E-readers still lists the device. “Forget…” next to it removes the device from every library on this Mac, with its syncs and its options, and keeps the books and reading data that came from it. If you keep a reset Kobo instead, <a href=\"/blog/kobo-sync-after-factory-reset/\">what Readwyn does with it</a> covers that case.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/does-kobo-sync-sideloaded-books/","url":"https://readwyn.com/blog/does-kobo-sync-sideloaded-books/","title":"Does a Kobo sync sideloaded books?","summary":"No. A book you copy to a Kobo stays on that Kobo, and Kobo's sync carries neither the book nor your place in it to another eReader or the Kobo Books app.","content_html":"<p>No. Kobo's sync is for the books in your Kobo account, and a book you copy onto the eReader over USB never joins that account. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">adding EPUB and PDF files from a computer</a> says the content “will only appear on the device that you used to import the content”, and that even after you sync, “you won't see the item on your other eReaders or apps.” Copying the same file to a second Kobo doesn't help either: your position, highlights and notes stay on the Kobo where you made them.</p>\n<h2 id=\"what-a-kobo-sync-does-carry\">What a Kobo sync does carry</h2>\n<p>Kobo describes the sync in two places. Over Wi-Fi, you tap the Sync icon on the Home screen and then “Sync now”; <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017513334\">that page</a> says a sync brings in recently purchased books and previews, checks for software updates, and updates your bookmarks and annotations. Through the Kobo Desktop app with a cable, <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017478174\">Kobo's list</a> is almost the same: it “downloads books you've recently purchased”, checks for updates, and “updates any bookmarks or annotations that you've made to your eBooks.” Neither page offers to upload a file that came from somewhere else.</p>\n<p>The phone and tablet apps follow the same rule. Kobo's import pages for <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017864893-Import-items-from-your-Android-device-to-your-Kobo-Books-library\">Android</a> and for <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360060463753-Import-items-from-your-iOS-or-macOS-device-onto-your-Kobo-Books-app\">iPhone, iPad and Mac</a> repeat the sentence word for word: an imported EPUB appears only on the device that imported it. So a book you add to the Kobo Books app on your phone won't reach your eReader by sync either. A Kindle has its own rules: see <a href=\"/blog/does-kindle-sync-sideloaded-books/\">whether a Kindle syncs the progress of sideloaded books</a>.</p>\n<h2 id=\"the-2026-rebuild-didn-t-change-it\">The 2026 rebuild didn't change it</h2>\n<p>In September 2026 Kobo rebuilt the system that syncs bookmarks and reading progress. Nathan Groezinger <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2026/09/11/sync-issues-with-kobo-ereaders-and-apps-finally-fixed/\">wrote at The eBook Reader</a> that the fix is “strictly for syncing purchased Kobo ebooks”, and that each eReader has to be reconnected through Settings, Device information, Repair your Kobo account (an app needs you to sign out and back in). <a rel=\"external\" href=\"https://www.notebookcheck.net/Kobo-finally-fixes-syncing-issues-on-its-e-readers.1398521.0.html\">Notebookcheck</a> reported the same limit: the fix “does not extend to sideloaded content.”</p>\n<h2 id=\"dropbox-and-google-drive\">Dropbox and Google Drive</h2>\n<p>Some Kobos can fetch books from the cloud instead of a cable. Kobo's pages on <a rel=\"external\" href=\"https://help.kobo.com/hc/articles/360033830114-Add-books-to-your-eReader-using-Dropbox\">Dropbox</a> and <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/15335985512983-Add-books-to-your-eReader-using-Google-Drive\">Google Drive</a> name the models that have the feature: the Forma, Sage, Elipsa, Elipsa 2E and Libra Colour. The book lands in My Books on that eReader. Neither page says that your place in it syncs to another device, so don't count on it. <a href=\"/blog/send-books-to-kobo-with-dropbox-or-google-drive/\">Sending books to a Kobo with Dropbox or Google Drive</a> gives the setup.</p>\n<h2 id=\"carrying-your-place-by-hand\">Carrying your place by hand</h2>\n<p>If you read one sideloaded book on two devices, the plain answer is to copy it to both and note the page or percentage when you switch. A Kobo keeps your percent read, the date you last read, and the time you've spent in each book in its own database on the eReader, <code>.kobo/KoboReader.sqlite</code>.</p>\n<p>Readwyn, the ebook library app for Mac and Linux that this blog belongs to, reads that database over USB, for the books that are also in your Readwyn library (a copy you added by hand matches its library book by title and author). Plug in the first Kobo and run “Device → Readwyn” to bring its progress, highlights, notes, status and time read into your library; then plug in the second and run “Readwyn → Device” to write them to the same book there (<a href=\"/blog/sync-reading-progress-kobo-and-mac/\">why the order matters</a>). None of it reaches the Kobo Books app on your phone. If a copied book never turned up on the Kobo in the first place, start with <a href=\"/blog/epub-not-showing-on-kobo/\">why an EPUB doesn't show up on your Kobo</a>.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/epub-not-showing-on-kobo/","url":"https://readwyn.com/blog/epub-not-showing-on-kobo/","title":"Why an EPUB does not show up on your Kobo","summary":"A Kobo adds a copied EPUB only after you eject it and it shows “Importing content”. If the book is still missing, check the My Books filter, DRM, and which device you look on.","content_html":"<p>The Kobo may not have looked for it yet. A Kobo adds the books you copied only after it leaves the computer: you eject it, unplug the cable, and the screen shows “Importing content”. Until then the file sits on the drive and My Books doesn't know about it. If it's still missing after the import, check these in turn: whether the file reached the Kobo at all, a filter in My Books, DRM on the file, an EPUB 3 on a very old Kobo, and the device you're looking on.</p>\n<h2 id=\"eject-unplug-and-wait-for-the-import\">Eject, unplug, and wait for the import</h2>\n<p>On Kobo's help page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">adding PDF and ePub files with a computer</a>, ejecting is one of the steps, not a tidy-up for afterwards. You tap Connect on the Kobo, open the drive called KOBOeReader on the computer, and drag the files into it. Then you right-click KOBOeReader, click Eject, and unplug. Only then does the page say you'll see an “Importing content” message, and only after that does it send you to My Books.</p>\n<p>So a book that's on the drive but not in My Books may only be waiting. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017765933-Eject-your-Kobo-eReader-when-connected-to-a-computer\">eject article</a> gives the reason for the eject as “to avoid losing data”, and <a href=\"/blog/eject-kindle-or-kobo-from-mac/\">ejecting a Kobo from a Mac</a> has the three ways to do it in Finder.</p>\n<p>Readwyn depends on the same import. Its “Send to device” copies the book into the Kobo's <code>Books</code> folder, for example as <code>Books/&lt;first author&gt; - &lt;title&gt;.kepub.epub</code>, and leaves the Kobo to add it when it leaves USB mode. So a book sent from Readwyn also shows up only after you eject, and the message after a send says so: “Copied to the device. Eject the device so it adds the book.”</p>\n<h2 id=\"the-file-never-reached-the-kobo\">The file never reached the Kobo</h2>\n<p>If there was no KOBOeReader drive to drag the file into, the problem is the connection, not the book, and <a href=\"/blog/kobo-not-showing-up-on-mac/\">why a Kobo doesn't show up on a Mac</a> goes through it in order. On the Kobo, the “Computer detected” window has to appear and you have to tap Connect. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/1500005659942\">USB troubleshooting page</a> suggests another cable, because “some USB cables can only charge”, and a direct connection without a hub. The add page has a note for Macs: on macOS Sonoma 14.4, plugging straight into the computer's port without a hub “may solve connection issues”.</p>\n<h2 id=\"a-filter-in-my-books\">A filter in My Books</h2>\n<p>My Books can hide books on purpose. Kobo's article on the <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017676993\">search filter</a> lists what it can narrow the list to: purchased books, previews, downloaded books, eBooks or audiobooks, and read or unread books. Tap Filter near the top of My Books, clear the filter, and look again. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017990073-See-the-books-that-are-on-your-Kobo-eReader\">seeing the books on your eReader</a> also suggests a search when a book seems to be missing.</p>\n<h2 id=\"drm-on-the-file\">DRM on the file</h2>\n<p>An EPUB from a shop or a library may be locked. The add page is plain about it: if your file has DRM “or your file is from a third party, you won't be able to read it on your eReader with this method.” Dragging it across isn't enough.</p>\n<p>Kobo's route for those books is Adobe Digital Editions. Its article on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017814074\">adding eBooks with Adobe Digital Editions</a> says that for a book with DRM “you'll usually download a file that ends in .acsm”. That file isn't the book. You open it in Adobe Digital Editions, authorise both the computer and the Kobo, and drag the book to the Kobo under Devices in that app. <a href=\"/blog/what-is-an-acsm-file/\">What an ACSM file is</a> explains why it won't open as a book on its own. If you only ever copied the <code>.acsm</code> file to the Kobo, there's no book on it.</p>\n<p>For a library loan, <a href=\"/blog/library-ebooks-on-kobo-and-kindle/\">reading library ebooks on a Kobo</a> gives the other routes.</p>\n<h2 id=\"an-epub-3-on-a-very-old-kobo\">An EPUB 3 on a very old Kobo</h2>\n<p>Kobo's list of <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">file formats its eReaders support</a> has EPUB, EPUB2 and EPUB3, with one exception: the Kobo Original (N416) and the <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360035135394\">Kobo Wi Fi</a> (N647, which <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019676973-Identify-your-Kobo-eReader-or-Kobo-tablet\">Kobo's model list</a> calls the Kobo Wireless) don't support EPUB3. If you're not sure which Kobo you have, <a href=\"/blog/which-kobo-do-i-have/\">its model number tells you</a>.</p>\n<p>A <code>.zip</code> download isn't a book to the Kobo either: the format list doesn't name ZIP. Unpack it on the computer and copy the <code>.epub</code> that was inside.</p>\n<p>If a plain EPUB with no DRM still won't show, the file itself may be faulty. <a href=\"/blog/check-epub-with-epubcheck/\">EPUBCheck</a> reads an EPUB and lists what's wrong with it, so you can tell a broken file from a problem with the Kobo.</p>\n<h2 id=\"you-re-looking-on-another-device\">You're looking on another device</h2>\n<p>A book copied by cable lives on that one Kobo. Kobo's add page says your content “will only appear on the device that you used to import the content”, and adds that even after you sync, “you won't see the item on your other eReaders or apps.” It won't come up in the Kobo Books app on your phone, or on a second Kobo, however many times you sync. To have it on a second Kobo too, copy it to that Kobo as well. The app has its own import, which takes an EPUB without DRM: Kobo explains it for <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017864893-Import-items-from-your-Android-device-to-your-Kobo-Books-library\">Android</a> and for <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360060463753\">iPhone and iPad</a>.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/instapaper-on-kobo/","url":"https://readwyn.com/blog/instapaper-on-kobo/","title":"How to set up Instapaper on a Kobo now that Pocket has closed","summary":"Make an Instapaper account, then on the Kobo tap More › My Articles › Link with Instapaper and type the code it shows at kobo.com/instapaper.","content_html":"<p>Instapaper took Pocket's place on the Kobo, and it links in much the same way: the Kobo shows a code, you type that code into a page on kobo.com, and the articles that you save to Instapaper download to the Kobo over Wi-Fi. You need an Instapaper account before you start (you can make one at instapaper.com) and a Kobo with the software update that added the feature in August 2025. Kobo's help page <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/33359968957463-Use-Instapaper-with-your-Kobo-eReader\">Use Instapaper with your Kobo eReader</a> has the steps below.</p>\n<h2 id=\"what-happened-to-pocket\">What happened to Pocket</h2>\n<p>Mozilla announced the end of Pocket on 22 May 2025. Its post <a rel=\"external\" href=\"https://blog.mozilla.org/en/mozilla/building-whats-next/\">Building what's next</a> said \"Pocket shuts down July 8, 2025\", and gave as the reason that \"the way people save and consume content on the web has evolved\". Saves could be exported until 8 October 2025, after which, the post said, the data \"will be permanently deleted\". So if you didn't export your list by that date, it can't be brought back.</p>\n<p>On the Kobo, the Pocket feature stopped with the service. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763753-Use-the-Pocket-App-with-your-Kobo-eReader\">old Pocket page</a> now has a notice headed “Pocket has shut down”, says \"As of July 8, 2025, Pocket is no longer available\", and sends you to the Instapaper page instead. Kobo announced the partnership with Instapaper in July 2025, aiming to launch \"by the end of summer 2025\", as <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2025/07/22/kobos-getting-instapaper-integration-pocket-articles-import/\">The eBook Reader reported</a>. If you did save a Pocket export in time, the same site's <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2025/08/26/latest-kobo-software-update-adds-instapaper-support/\">report on the update</a> says that you import it into Instapaper from your account page at instapaper.com/user.</p>\n<h2 id=\"check-the-software-first\">Check the software first</h2>\n<p>The Instapaper link arrived in two builds, according to The eBook Reader: 4.38.23429 for the older models, as far back as the Kobo Touch of 2011, and 4.43.23418 for the Clara BW, Clara Colour and Libra Colour. The same report says that Kobo rolls an update out slowly across regions, and that the Kobo installs it during a sync. Engadget wrote that the feature is <a rel=\"external\" href=\"https://www.engadget.com/mobile/tablets/kobo-ereaders-are-swapping-out-pocket-for-instapaper-190615508.html\">\"available globally on all currently supported Kobo ereaders\"</a>.</p>\n<p>To see what your Kobo runs, tap More on the Home screen, then Settings › Device information, and read the number beside “Software version” (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024773213-Find-your-eReader-s-software-version\">Kobo's page</a>). A number lower than the one for your model means the update hasn't reached it yet. If you aren't sure which model you have, <a href=\"/blog/which-kobo-do-i-have/\">which Kobo do I have</a> goes by the model number.</p>\n<h2 id=\"link-the-account\">Link the account</h2>\n<ol>\n<li>On the Kobo, go to the Home screen, tap More, then “My Articles”, then “Link with Instapaper”. The Kobo shows a code.</li>\n<li>On a computer, open <code>kobo.com/instapaper</code>. On a phone, scan the QR code that the Kobo shows instead.</li>\n<li>Type the code, click “Link with Instapaper”, and click “Authorize” on the Application Access page.</li>\n</ol>\n<h2 id=\"saving-and-reading\">Saving and reading</h2>\n<p>Kobo's page names several ways to save an article: the page <code>instapaper.com/save</code>, the browser extension, email, IFTTT, or the Instapaper app for iOS or Android. On the Kobo, the saved articles are listed under More › My Articles, and the Kobo needs Wi-Fi to download them. Tap one to open it. If an article that you saved isn't in the list, tap the Menu icon in the grey bar and then “Refresh”.</p>\n<p>While you read, five icons act on the article. The Menu icon opens more options, such as adding it to favourites. Archive closes it and moves it to Instapaper's archive list, Like adds it to the Liked list, Delete removes it from your reading list, and Return to Articles goes back to the list.</p>\n<h2 id=\"what-stays-behind\">What stays behind</h2>\n<p>Kobo's page lists the limits. Only the main Instapaper list syncs: an article in an Instapaper folder doesn't appear on the Kobo until you move it back to the main list. Tags, highlights, notes, videos and PDFs aren't supported.</p>\n<p>Engadget noted two more gaps when the feature launched: there's no handwritten annotation on the Kobos with a stylus, such as the Libra Colour and the Elipsa 2E, and Instapaper \"seems much less concerned with preserving images\" than Pocket was.</p>\n<p>Since Instapaper doesn't bring PDFs to the Kobo, a PDF has to take another route. On a Forma, Sage, Elipsa, Elipsa 2E or Libra Colour, a <a href=\"/blog/send-books-to-kobo-with-dropbox-or-google-drive/\">Dropbox or Google Drive folder</a> puts it on the Kobo over Wi-Fi, and every Kobo <a href=\"/blog/kobo-file-formats/\">reads PDF</a> from a USB cable.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-book-has-no-cover/","url":"https://readwyn.com/blog/kobo-book-has-no-cover/","title":"Why a book has no cover on a Kobo","summary":"A Kobo store book can lose its cover until you repair the account. A book you copied over needs a cover that its EPUB marks in the way the Kobo reads.","content_html":"<p>It depends on where the book came from. A book bought from Kobo can show a blank tile because the cover image never stayed on the device, and repairing your Kobo account brings it back. A book that you copied over yourself has to name its cover inside the file. Kobo's EPUB guidelines for the people who make books for it say that Kobo picks the library thumbnail from one line in the book's package file, the EPUB 2 cover <code>meta</code> tag, so check first whether your book has that line or marks its cover only in the EPUB 3 form. A picture on the first page may not be enough.</p>\n<h2 id=\"a-book-from-the-kobo-store\">A book from the Kobo store</h2>\n<p>Kobo has a help page called <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360037971714-Missing-book-cover-images\">Missing book cover images</a>. It says that an issue which hid some covers on eReaders and in Kobo Desktop has been resolved, but that a cover which wasn't downloaded for good is lost again when the device restarts or goes offline. To download the covers permanently, it tells you to repair your account.</p>\n<p>On the eReader, Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605394-Repair-your-Kobo-account-on-your-Kobo-eReader\">repair steps</a> are: on the Home screen, tap More at the bottom, then Settings, then Device information, then Repair next to “Repair your Kobo account”, and Repair now. You need Wi-Fi, and with a large library it takes a while.</p>\n<p>Both of Kobo's pages are about the books and covers of your Kobo account. Neither says that a repair helps a book that you copied from your computer.</p>\n<h2 id=\"a-book-you-copied-over\">A book you copied over</h2>\n<p>Here the cover has to come from inside the file. An EPUB keeps a list of all its files in its package document (the <code>.opf</code> file; <a href=\"/blog/what-is-inside-an-epub-file/\">what is inside an EPUB file</a> walks through it), and a reading system needs a mark in that list to know which image is the cover. There are two such marks:</p>\n<ul>\n<li><strong>EPUB 2:</strong> a <code>meta</code> element in the metadata, <code>&lt;meta name=\"cover\" content=\"…\"/&gt;</code>, whose <code>content</code> is the <code>id</code> of the image in the manifest.</li>\n<li><strong>EPUB 3:</strong> the property <code>cover-image</code> on the manifest item of the image. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-cover-image\">EPUB 3.3 specification</a> allows it on at most one item, and it <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-item-resource-properties\">recommends setting it</a> when the book has a cover, without making it compulsory.</li>\n</ul>\n<p>A book with neither mark has a cover only by accident: it may open on a picture, but nothing in the book marks that picture as the cover.</p>\n<p>Kobo's own <a rel=\"external\" href=\"https://github.com/kobolabs/epub-spec\">EPUB guidelines</a>, as they read in November 2026, show only the EPUB 2 mark as the cover in their example. They say that Kobo uses the <code>meta</code> tag “to identify which image to display as a thumbnail” in readers' libraries. They also advise tags for manifest items, “specifically the cover tag”, but say only that some of these tags are read across Kobo's platforms. The EPUB 3.3 specification files that tag under <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-opf2-meta\">legacy features</a>, kept for EPUB 2 reading systems, and allows an EPUB 3 book to carry it alongside <code>cover-image</code>. A book that has both marks works whichever one a reading system looks for.</p>\n<h2 id=\"other-things-in-the-file-that-hide-a-cover\">Other things in the file that hide a cover</h2>\n<p>The same guidelines, written for books that go to Kobo, give three more rules that are worth following in a copied book too:</p>\n<ul>\n<li><strong>GIF.</strong> “GIF files are not supported for covers.” A cover in JPEG or PNG avoids the question.</li>\n<li><strong>A cover drawn with CSS.</strong> Kobo's automatic cover extraction doesn't support a cover set as a CSS <code>background-image</code>. The guidelines ask for an <code>img</code> element in the cover page instead.</li>\n<li><strong>KEPUB files you add yourself.</strong> In the section on sideloading for testing, they warn that for a sideloaded <code>.kepub.epub</code> file, “Thumbnails for covers may not display.”</li>\n</ul>\n<p>The last one is worth knowing if your books go over as KEPUB. Readwyn sends an EPUB to a Kobo as a <code>.kepub.epub</code> by default; “Convert EPUB to KEPUB when sending” on the Sync screen turns that off for one Kobo, and the book then goes as a plain <code>.epub</code> (<a href=\"/blog/kepub-or-epub-for-kobo/\">KEPUB or EPUB</a> gives what each form does for highlights).</p>\n<h2 id=\"how-to-look-at-the-file\">How to look at the file</h2>\n<p>You can open the <code>.opf</code> file of a copy of the book in any text editor and look for <code>name=\"cover\"</code> and <code>cover-image</code>. On a Mac with Readwyn installed, select the EPUB in Finder and press Space. If the Quick Look preview shows a drawn cover, the file has no cover image that the preview could find and open. A real cover there doesn't prove that the Kobo mark is present, because the preview also takes an image named like a cover, so the <code>.opf</code> file is still the place to check.</p>\n<p>To fix the file, add the missing mark (or both), point it at a JPEG or PNG image, and copy the book to the Kobo again. (<a href=\"/blog/ebook-cover-size/\">What size should an ebook cover be?</a> gives the shape that Kobo asks for.)</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-file-formats/","url":"https://readwyn.com/blog/kobo-file-formats/","title":"Which file formats can a Kobo read?","summary":"A Kobo e-reader reads EPUB, PDF, and MOBI books, CBZ and CBR comics, TXT, HTML, and RTF text, and five image formats. AZW3 isn't on the list, and a book with Kindle DRM won't open.","content_html":"<p>A Kobo e-reader reads EPUB, PDF, and MOBI books, CBZ and CBR comics, TXT, HTML, and RTF text, and five image formats. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">list of supported file formats</a> sorts them like this:</p>\n<table><thead><tr><th>Kind</th><th>Formats</th></tr></thead><tbody>\n<tr><td>Books</td><td>EPUB, EPUB2, EPUB3, PDF, MOBI, and one that the page spells \"FlePub\" and doesn't explain</td></tr>\n<tr><td>Documents</td><td>PDF</td></tr>\n<tr><td>Comics</td><td>CBZ and CBR</td></tr>\n<tr><td>Text</td><td>TXT, HTML, and RTF</td></tr>\n<tr><td>Images</td><td>JPEG, GIF, PNG, BMP, and TIFF</td></tr>\n</tbody></table>\n<p>There's one exception in the list: the Kobo Original and the Kobo Wi Fi don't support EPUB3. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019676973-Identify-your-Kobo-eReader-or-Kobo-tablet\">identification page</a> gives the Original the model number N416, and its <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360035135394\">page for the Wi Fi</a> gives N647 or N647B, the model that <a href=\"/blog/which-kobo-do-i-have/\">which Kobo do I have</a> calls the Kobo Wireless. Both numbers are on the bottom of the device. What the list leaves out matters as much. AZW3, the Kindle format, isn't on it, and neither is ZIP.</p>\n<h2 id=\"mobi-as-long-as-it-has-no-kindle-drm\">MOBI, as long as it has no Kindle DRM</h2>\n<p>A MOBI that you made yourself, or downloaded without DRM, goes onto a Kobo as it is. A book bought in the Kindle Store is another matter. Amazon tells authors on its <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GDDXGH9VR22ACM8U\">publishing help page</a> that \"applying DRM limits reader access to Kindle apps and devices only\". The same page says that since 20 January 2026, a buyer of a book that its author has confirmed as DRM-free can download it as an EPUB and a PDF. That EPUB is an ordinary file, and the Kobo reads it. <a href=\"/blog/moving-from-kindle-to-kobo/\">Moving from a Kindle to a Kobo</a> goes through the rest of a Kindle library, from store books to library loans.</p>\n<p>The DRM that a Kobo does handle is of two kinds, in the words of Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">format page</a>: \"Your eReader supports Adobe DRM and Social DRM.\" A book with <a href=\"/blog/social-drm-watermarks/\">social DRM</a> opens without extra software. Kobo also says that an EPUB or PDF book with an <code>.acsm</code> file extension is protected by DRM. That file isn't the book itself (<a href=\"/blog/what-is-an-acsm-file/\">what an ACSM file is</a>): a book with Adobe DRM has to go through Adobe Digital Editions first. The format page adds that Adobe Digital Editions works with Kobo e-readers only, not with the Kobo app.</p>\n<h2 id=\"pdf-works-with-limits\">PDF works, with limits</h2>\n<p>A PDF opens, but Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017513414-Turn-pages-magnify-and-rotate-PDFs-using-your-Kobo-eReader\">reading PDFs</a> lists four things you can't do in one: select text, add highlights, look up words in the dictionary, and change the text size or style. What you get instead is zoom (a slider under the Navigation icon, or a pinch) and a Rotate icon in the same menu that turns the page 90°, which Kobo says regular ebooks can't do: they read in portrait only.</p>\n<p>If you'd rather read the text at your own size, the book has to become reflowable first. Readwyn, the library app that this blog belongs to, converts a PDF to an EPUB 3 when you add it, and keeps the PDF unchanged next to it. A scanned PDF without a text layer, or an encrypted one, doesn't convert. The <a href=\"/blog/convert-pdf-to-epub-without-calibre/\">post on PDF to EPUB</a> shows what the conversion keeps.</p>\n<h2 id=\"kepub-the-format-that-isn-t-on-the-list\">KEPUB, the format that isn't on the list</h2>\n<p>Kobo's list doesn't name KEPUB, Kobo's own flavour of EPUB, but Kobos open it. <a rel=\"external\" href=\"https://standardebooks.org/help/how-to-use-our-ebooks\">Standard Ebooks</a> offers a kepub download that it calls \"specially prepared for the best reading experience on Kobos\". Its instructions have one detail that's easy to miss: when you drag the file to the Kobo, the name must end in <code>.kepub.epub</code>.</p>\n<p>Readwyn sends EPUBs that way by default. Its “Send to device” converts the EPUB to <code>Books/&lt;first author&gt; - &lt;title&gt;.kepub.epub</code> on the Kobo, and with the option “Convert EPUB to KEPUB when sending” turned off, it copies the plain EPUB. A MOBI or AZW3 in your library goes as the EPUB 3 that Readwyn converted from it. <a href=\"/blog/kepub-or-epub-for-kobo/\">KEPUB or EPUB</a> explains the difference for highlights.</p>\n<h2 id=\"comics\">Comics</h2>\n<p>Both comic archives are on the list, but the list says nothing about the images inside one. WebP isn't among the image formats Kobo names, so a comic with WebP pages is the one to test before you count on it, and <a href=\"/blog/put-cbz-comics-on-kobo/\">putting CBZ comics on a Kobo</a> has the details.</p>\n<h2 id=\"the-kobo-books-app-reads-less\">The Kobo Books app reads less</h2>\n<p>In the app for phones and tablets, the list is shorter. Kobo gives the iOS app EPUB, EPUB2, EPUB3, PDF, FlePub, and MOBI. The Android app gets the same minus PDF. Neither list has comics, text files, or images. Both add ePub3AV, which Kobo describes as books with audio or video content.</p>\n<p>A book that you copied to a Kobo by cable doesn't reach the app either. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">page on adding files from a computer</a> says it will \"only appear on the device that you used to import the content\". If a file of a supported format still doesn't turn up in My Books, <a href=\"/blog/epub-not-showing-on-kobo/\">why an EPUB doesn't show up on a Kobo</a> goes through the causes in order.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-folder-and-koboreader-sqlite/","url":"https://readwyn.com/blog/kobo-folder-and-koboreader-sqlite/","title":"What is in the hidden .kobo folder on a Kobo, and what is KoboReader.sqlite?","summary":"The .kobo folder is the Kobo's own system folder. KoboReader.sqlite inside it is the database of your books, highlights, progress, collections and saved words.","content_html":"<p>The <code>.kobo</code> folder is where a Kobo keeps its own working files: its database, the encrypted copies of books you bought from Kobo, the dictionaries, a settings file, and a small file that names the model and the software version. <code>KoboReader.sqlite</code> is the database. It's an ordinary SQLite file, and it holds the list of books on the device and almost everything you do with them: highlights, notes, bookmarks, how far you've read, ratings, time spent reading, collections, and the words you saved to My Words. If you can't see the folder at all, that's by design on a Mac: its name starts with a full stop, and macOS hides such folders unless you ask to see them.</p>\n<h2 id=\"showing-the-folder\">Showing the folder</h2>\n<p>Connect the Kobo with its cable and tap Connect on the “Computer detected” window. The Mac then gets a drive named KOBOeReader, as Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">guide to adding files</a> says. If the drive itself doesn't appear, that's a different problem, covered in <a href=\"/blog/kobo-not-showing-up-on-mac/\">why your Kobo does not show up on a Mac</a>.</p>\n<p>Open the drive in Finder and press Command-Shift-full stop. <a rel=\"external\" href=\"https://www.macworld.com/article/342749/how-to-reveal-hidden-files-in-macos-and-hide-visible-ones.html\">Macworld</a> describes the hidden items coming back “in a grayed-out state”, and the same keys hide them again. <a rel=\"external\" href=\"https://tidbits.com/2018/05/16/tipbits-reveal-invisible-files-on-the-mac-with-a-keystroke/\">TidBITS</a> gives the reason they were hidden: some Unix files and folders start with a period, so they're hidden by default. On Windows 11, Microsoft's <a rel=\"external\" href=\"https://support.microsoft.com/en-us/windows/view-hidden-files-and-folders-in-windows-97fbc472-c603-9d90-91d0-1166d1d9f4b5\">help page</a> says to choose View, Show, Hidden items in File Explorer. On Windows 10 it's View, Options, “Change folder and search options”, then “Show hidden files, folders, and drives” on the View tab.</p>\n<h2 id=\"what-the-folder-holds\">What the folder holds</h2>\n<p>The contents differ by model and software version, but these are the parts you're most likely to meet:</p>\n<ul>\n<li><strong><code>KoboReader.sqlite</code></strong>, the database. More on it below.</li>\n<li><strong><code>version</code></strong>, a text file of about 100 bytes. It holds the serial number, the software version, and, at the end, a product ID that identifies the model.</li>\n<li><strong><code>kepub</code></strong>, the books from the Kobo store. The <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Kepub\">MobileRead wiki</a> describes files with long numbered names and no extension, one per book, most of their contents encrypted.</li>\n<li><strong><code>dict</code></strong>, the dictionaries. On a Kobo Libra Colour we examined, it held <code>dicthtml.zip</code> for English and a <code>dicthtml-&lt;language&gt;-en.zip</code> file for German, Spanish, French, Italian, Dutch and Portuguese. A <code>custom-dict</code> folder next to it takes your own dictionaries, which <a href=\"/blog/how-e-reader-dictionaries-work/\">how e-reader dictionaries work</a> explains.</li>\n<li><strong><code>Kobo/Kobo eReader.conf</code></strong>, the settings file. The MobileRead wiki's page of <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Kobo_Configuration_Options\">configuration options</a> lists settings that you can add to this file by hand, for example <code>ExcludeSyncFolders</code>, which “controls which folders are not processed into the database”. Many entries there are marked as unknown or broken on recent software.</li>\n</ul>\n<p>One thing the folder isn't is a place for your own books. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/43753873356055-Missing-books-reading-stats-or-notes-after-Kobo-software-update\">article on the 6.0 update</a> says that books sideloaded into the hidden <code>.kobo</code> folder weren't kept through the update to 6.0, and that books put there after it aren't imported into My Books. The article is about the Libra Colour, Clara BW and Clara Colour. Kobo's advice is to put them at the top level of the drive instead.</p>\n<h2 id=\"inside-koboreader-sqlite\">Inside KoboReader.sqlite</h2>\n<p>What follows comes from Readwyn's study of a Kobo Libra Colour on software 4.45, whose database reported <code>DbVersion</code> 176. Other versions can lack a table or a column.</p>\n<table><thead><tr><th>Table</th><th>What it holds</th></tr></thead><tbody>\n<tr><td><code>content</code></td><td>A row for each book, with <code>ContentType</code> 6, and rows for its chapters. A book row carries <code>ReadStatus</code> (0 unread, 1 reading, 2 finished), <code>___PercentRead</code>, <code>DateLastRead</code>, and <code>TimeSpentReading</code> in seconds.</td></tr>\n<tr><td><code>Bookmark</code></td><td>Highlights, notes, dog-ears and stylus drawings, one row each. A highlight's <code>Color</code> is 0 to 3: yellow, pink, blue, green.</td></tr>\n<tr><td><code>ratings</code></td><td>Your star rating of each book, 1 to 5.</td></tr>\n<tr><td><code>Shelf</code> and <code>ShelfContent</code></td><td>Your collections, and which books are in each.</td></tr>\n<tr><td><code>WordList</code></td><td>My Words: the word, the book, the dictionary, and the date you saved it.</td></tr>\n</tbody></table>\n<p>A few details explain what you see when you open it. Books are named by where the Kobo sees them, so a sideloaded KEPUB's ID looks like <code>file:///mnt/onboard/&lt;path&gt;.kepub.epub</code>. Yes or no values are often the text <code>true</code> or <code>false</code>, not a number. A collection's books are marked rather than removed: a <code>ShelfContent</code> row has an <code>_IsDeleted</code> flag, and a row with it set to <code>true</code> counts as gone. A <code>Bookmark</code> row has a <code>Hidden</code> column too, normally the text <code>false</code>; Readwyn treats a row with it set as deleted. Dates come in more than one form, sometimes <code>2026-10-03T20:19:47.911</code> and sometimes <code>2026-10-03T20:19:47Z</code>. And <code>WordList</code> has no sentence and no position: only the word you saved, its book, the dictionary you looked it up in, and the time.</p>\n<h2 id=\"opening-it-without-harming-it\">Opening it without harming it</h2>\n<p>The database uses SQLite's write-ahead log, so changes can sit in a <code>KoboReader.sqlite-wal</code> file beside it before they reach the main file. The <a rel=\"external\" href=\"https://www.sqlite.org/wal.html\">SQLite documentation</a> warns that the WAL file is part of the database: copy the database without it and committed changes can be lost, or the file can be corrupted.</p>\n<p>So the safe way to look is to copy <code>KoboReader.sqlite</code>, and a <code>-wal</code> file if there is one, to your Mac, eject the Kobo, and open the copy. <a rel=\"external\" href=\"https://sqlitebrowser.org/\">DB Browser for SQLite</a>, an open-source viewer for macOS, Windows and Linux, shows each table like a spreadsheet. Editing the original on the Kobo is riskier, and it's where the WAL bites: the Kobo wants no <code>-wal</code> or <code>-shm</code> file when you eject, yet you must never delete a <code>-wal</code> file by hand. SQLite says the only safe way to remove one is to open the database and close it again, and the <code>sqlite3</code> that comes with macOS leaves it in place even then. Explore the copy, and leave the original alone.</p>\n<p>Readwyn reads this database over USB to bring a Kobo's highlights, notes, progress and saved words into your library. When a sync writes back, it first copies <code>KoboReader.sqlite</code> into its app folder on your Mac, and <a href=\"/blog/how-readwyn-writes-to-kobo-safely/\">how Readwyn writes to a Kobo safely</a> gives the rest of the steps.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-not-showing-up-on-mac/","url":"https://readwyn.com/blog/kobo-not-showing-up-on-mac/","title":"Why your Kobo does not show up on a Mac","summary":"A Kobo joins the Mac only after you tap Connect. If “Computer detected” never appears, suspect the cable, a hub, or the port. If it does, check a macOS alert and Finder.","content_html":"<p>A Kobo doesn't appear on a Mac the moment you plug it in. It first shows a window called “Computer detected” on its own screen, and only after you tap Connect there does the Mac get a drive named KOBOeReader, as Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">guide to adding files</a> describes. So the first question is what the Kobo's screen says. If that window never appears, the Kobo and the Mac haven't made a data connection, and the first things to check are the cable, a hub, and the port. If the window appears, you tap Connect, and Finder still shows nothing, the cause is on the Mac: an accessory alert you dismissed, or a Finder setting that hides external disks.</p>\n<h2 id=\"no-computer-detected-window\">No “Computer detected” window</h2>\n<p><a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/1500005659942-Troubleshoot-issues-connecting-your-eReader-to-your-computer-with-a-USB-cable\">Kobo's troubleshooting page</a> treats the window as the sign that all is well: it means the eReader is charging and ready to connect. When it doesn't come up, work through these in order: the first ones are the cheapest to try.</p>\n<p><strong>The cable.</strong> Kobo's page warns that “some USB cables can only charge”. A charge-only cable fills the battery and carries no data, so the Kobo has nothing to detect. Try another cable, ideally one you know has moved files before. On an older model such as the Kobo Glo, the page suggests a micro USB-A cable.</p>\n<p><strong>A hub.</strong> Plug the Kobo straight into the Mac, with no hub or dock between them. Kobo adds a note for Mac owners: on macOS Sonoma 14.4, connecting directly “may solve connection issues”. That version did have a hub bug: Apple released macOS 14.4.1 on 25 March 2024 with a fix for “USB hubs connected to external displays may not be recognized”, as <a rel=\"external\" href=\"https://www.macrumors.com/2024/03/25/apple-releases-macos-sonoma-14-4-1/\">MacRumors reported</a>. If your Mac is still on 14.4, update it.</p>\n<p><strong>The port, on both ends.</strong> Try another USB port on the Mac. Then look at the port on the Kobo itself: Kobo's page lists dust, a loose fit, and damage as things to check.</p>\n<p><strong>The battery.</strong> An eReader that hasn't been used for a while “may not detect the computer right away”, Kobo says. Leave it charging for at least an hour before you decide that the connection is broken.</p>\n<p><strong>The Kobo's software.</strong> Kobo's page says that “software updates may fix issues related to the eReader's USB port”, and the way to update is a sync over Wi-Fi, which needs no cable.</p>\n<p><strong>Another computer.</strong> Restart the Mac and try again. If you have another computer, try the Kobo there, plugged straight into a port. A Kobo that no computer detects points to the Kobo or its cable rather than to your Mac.</p>\n<p>The last step on Kobo's list, before you contact its support, is a factory reset. Leave it for last, because it costs you something. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605114-Reset-your-Kobo-eReader\">page on resetting</a> says a reset will “remove books you've added to your eReader using your computer”, and you'll lose bookmarks and annotations made since the last sync. Books bought from Kobo can be downloaded again.</p>\n<h2 id=\"the-window-appears-but-finder-shows-nothing\">The window appears, but Finder shows nothing</h2>\n<p>Once you've tapped Connect, the Kobo is in USB mode and the Mac should mount it. If it doesn't, the Kobo has done its part and the rest is on the Mac.</p>\n<p><strong>An accessory alert.</strong> A Mac laptop with Apple silicon asks before a new USB accessory may connect, <a rel=\"external\" href=\"https://support.apple.com/en-us/102282\">Apple's support page says</a>. If you clicked “Don't Allow”, the Mac “won't recognize the accessory”, and yet the same page notes that an accessory can still charge, so the Kobo's battery icon tells you nothing. The Mac asks again the next time you connect it: unplug the Kobo, plug it back in, and click “Allow”. The Mac has to be past its lock screen before it can allow an accessory. The setting lives in System Settings &gt; Privacy &amp; Security &gt; Accessories, under “Allow accessories to connect”, and its default is “Ask for new accessories”.</p>\n<p><strong>Finder's settings.</strong> The Kobo can be mounted and still hidden. Choose Finder &gt; Settings: the General pane sets which items show on the desktop and the Sidebar pane which show in the sidebar, and <a rel=\"external\" href=\"https://support.apple.com/guide/mac-help/mchlp1039/mac\">external disks are among those items</a>.</p>\n<p><strong>An ejected volume.</strong> If KOBOeReader was ejected without the cable coming out, open Disk Utility, select the volume, and click Mount. <a rel=\"external\" href=\"https://support.apple.com/guide/disk-utility/mount-a-volume-thats-been-ejected-dskutl1027/mac\">Apple's guide</a> also says that disconnecting the device and connecting it again makes the volume available.</p>\n<p><strong>Whether the Mac sees it at all.</strong> <a rel=\"external\" href=\"https://support.apple.com/guide/mac-help/if-a-usb-device-doesnt-work-mchlp1641/14.0/mac/14.0\">Apple's guide to a USB device that doesn't work</a>, in its edition for macOS 14, gives a test: hold the Option key, choose Apple menu &gt; System Information, and look for the device under Hardware in the list on the left. If the Kobo is listed there, the Mac sees it over USB, so the cable and the port carry data: look at Finder's settings and at an ejected volume. If it isn't, check the accessory alert above, then go back to the first section. The same guide suggests removing hubs and extension cables and disconnecting every other USB device except an Apple keyboard and mouse.</p>\n<h2 id=\"when-finder-shows-it-and-an-app-doesn-t\">When Finder shows it and an app doesn't</h2>\n<p>An app that reads a Kobo over USB reads that same KOBOeReader drive, so if the Mac hasn't mounted it, no app will find it either: Readwyn's Sync screen then says “No device connected” and reminds you to tap Connect on the Kobo. If Finder does show the drive, a sandboxed app may still need your permission. In the Mac App Store version of Readwyn, the Sync screen shows “Give Readwyn access to KOBOeReader”, and you <a href=\"/blog/kobo-access-mac-app-store/\">select the Kobo in a Finder dialog</a> once.</p>\n<p>When you've finished, <a href=\"/blog/eject-kindle-or-kobo-from-mac/\">eject the Kobo before you unplug it</a>. A Kindle has its own set of causes, starting with the models that never mount on a Mac at all: they're in <a href=\"/blog/kindle-not-showing-up-on-mac/\">why your Kindle does not show up on a Mac</a>.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-not-syncing-new-books/","url":"https://readwyn.com/blog/kobo-not-syncing-new-books/","title":"Why your Kobo does not sync new books","summary":"A Kobo fetches new store books only when a Wi-Fi sync runs on the right account. Check the network, the account, the archive, and free space, then repair the account.","content_html":"<p>Check two things first: that a sync really ran, and that the Kobo is signed in to the account that bought the book. A Kobo downloads new purchases only during a Wi-Fi sync, and only for the account it's signed in to. If both are right, the book may be in your archive, the Kobo may be out of space, or its copy of your account may need a repair. Most of these checks come from Kobo's own <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605054-Troubleshoot-eBook-and-audiobook-issues\">list for missing books</a>.</p>\n<p>One case is different: a book you copied over by cable, yourself or with an app on your computer, never comes through a Wi-Fi sync at all.</p>\n<h2 id=\"start-a-sync-by-hand\">Start a sync by hand</h2>\n<p>Kobo's article on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017513334\">syncing over Wi-Fi</a> has three steps: go to the Home screen, tap the Sync icon at the top, and tap Sync now. That sync downloads the books you've bought recently, previews you added to your account, your bookmarks and annotations, and any software update. Once it's done, new books show on the Home screen and in My Books.</p>\n<p>Leave the Kobo awake while it works. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017477434\">turning the eReader on and off</a> warns that putting it to sleep “can interrupt things like syncing”. A quick press of the power button is enough to send it to sleep.</p>\n<p>If Sync now finishes and the book still isn't there, turn the Kobo off and on, then sync again: that's the second step on Kobo's list.</p>\n<h2 id=\"the-kobo-isn-t-on-wi-fi\">The Kobo isn't on Wi-Fi</h2>\n<p>If Sync now fails, the network is the first suspect. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017478194-Fix-Wi-Fi-issues-on-your-Kobo-eReader\">fixing Wi-Fi issues</a> points to one trap: most Kobo eReaders join only 2.4 GHz networks. The Kobo Elipsa, released in June 2021, and the models after it can use 5 GHz as well. If your router offers only 5 GHz, an older Kobo can't see it. <a href=\"/blog/which-kobo-do-i-have/\">Which Kobo you have</a> settles which side of that line yours is on.</p>\n<p>The same page's other steps are the usual ones: check that the network works on a phone, turn Wi-Fi off and on, update the Kobo, restart it, restart the router, then remove the network and add it again. The list also has a factory reset. Keep that for last: the page says it erases your bookmarks, annotations, and settings.</p>\n<h2 id=\"the-kobo-is-signed-in-to-another-account\">The Kobo is signed in to another account</h2>\n<p>Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017765893\">checking the account</a> tells you to open More, then Settings, then Accounts, and read the email address under Kobo. Compare it with the address on the receipt for the book. If they differ, the Kobo is syncing a library that doesn't have the book.</p>\n<h2 id=\"the-book-is-in-your-archive\">The book is in your archive</h2>\n<p>A book in the archive of your Kobo account doesn't go to your devices: Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017640033-Delete-a-book-from-your-eReader\">deleting a book</a> says a book removed from My Books stays out of your library “even after you sync”. To restore it, follow Kobo's article on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017765873\">moving items out of the archive</a> on kobo.com: open My Books from your profile, click Archive, click the three dots next to the book, choose Restore to Library, and confirm with Yes, Move to Library. Then sync the Kobo again, and the book appears in My Books.</p>\n<h2 id=\"the-book-is-there-but-not-downloaded\">The book is there but not downloaded</h2>\n<p>A synced book can sit in My Books without its file. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017513474\">searching your eReader</a> has you tap the magnifying glass, switch the search from the Kobo Store to My Books, and type the title or author. If the result has “Download” beside it, tap the cover and wait for the download to finish.</p>\n<p>If the download itself won't complete, the Kobo may be short of space. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/42850160943511-Kobo-Cloud-upload-eBooks-and-documents-with-Kobo-Stack\">Kobo Cloud page</a> says that if your device is low on storage, “new uploads and synced content may not download successfully.” Remove books you've finished and sync again.</p>\n<h2 id=\"repair-the-account\">Repair the account</h2>\n<p>When none of that helps, Kobo suggests a <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605394\">repair of the account</a> on the eReader: More, Settings, Device information, then Repair next to your Kobo account, and Repair now. It needs Wi-Fi, and with a large library it “may take a while”.</p>\n<p>The last step on Kobo's list is to sign out of the Kobo and sign in again. Think twice before that one if your Kobo holds books you added yourself. Kobo's article on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/43753873356055-Missing-books-reading-stats-or-notes-after-Kobo-software-update\">missing books after a software update</a> warns that signing out or a reset can permanently delete what lives only on the eReader: sideloaded books and their annotations, My Words, reading stats, and Adobe authorisation.</p>\n<h2 id=\"books-you-copied-over-usb-never-sync\">Books you copied over USB never sync</h2>\n<p>Wi-Fi sync brings what your Kobo account holds. An EPUB you copied over USB isn't in your account, so no sync fetches it, and no sync will copy it to your second Kobo either (<a href=\"/blog/does-kobo-sync-sideloaded-books/\">what a Kobo sync carries for a sideloaded book</a>). A book you uploaded to Kobo Cloud does come down with a sync, like a purchase, and on some models so does one from <a href=\"/blog/send-books-to-kobo-with-dropbox-or-google-drive/\">Dropbox or Google Drive</a>. A Kobo adds a copied book when you eject it and it shows “Importing content”, and <a href=\"/blog/epub-not-showing-on-kobo/\">why an EPUB does not show up on your Kobo</a> goes through that route in order, from the eject to DRM.</p>\n<p>The same goes for Readwyn. Its sync carries highlights, progress, and reading data between your Mac and the Kobo, and never adds a book: a book goes to the Kobo only with “Send to device”, which writes it to the Kobo over USB. The message afterwards says what to do next: “Copied to the device. Eject the device so it adds the book.” <a href=\"/blog/send-books-to-kobo-from-mac/\">Sending books to a Kobo from a Mac</a> has the steps.</p>\n<h2 id=\"after-a-kobo-software-update\">After a Kobo software update</h2>\n<p>One fault is Kobo's own, and it hits only the Kobo Libra Colour, Clara BW, and Clara Colour. After the beta software 5.18.270971 or 6.0.274403 (you can see your version under More, Settings, Device information), the <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/43753873356055-Missing-books-reading-stats-or-notes-after-Kobo-software-update\">update article</a> gives two more reasons. Library loans stop arriving until you link your library card again under Discover, then Libby. And a book that you put inside the hidden <code>.kobo</code> folder after the 6.0 update isn't imported: copy it to the top level of the Kobo instead.</p>\n<p>If sideloaded books vanished after that beta, the same article names the software versions that bring them back, and asks you not to roll back the software, reset the Kobo, or sign out while you wait.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/manage-kobo-collections-from-mac/","url":"https://readwyn.com/blog/manage-kobo-collections-from-mac/","title":"How to manage Kobo collections from a Mac","summary":"In Readwyn, a library shelf is a Kobo collection of the same name. Put books on a shelf, sync Readwyn → Device, and the collection appears on the Kobo.","content_html":"<p>On a Mac, you manage Kobo collections as shelves. Readwyn, an ebook library app for Mac and Linux, treats each <a href=\"/blog/reading-status-and-shelves/\">shelf</a> in your library as a Kobo collection of the same name. Put books on a shelf called “Holiday”, connect the Kobo, run a Readwyn → Device sync, and a collection called “Holiday” holds those books on the Kobo. The other direction works too: a collection you made on the Kobo comes into the library as a shelf when you sync Device → Readwyn.</p>\n<h2 id=\"putting-books-on-a-shelf\">Putting books on a shelf</h2>\n<p>For one book, open it and look at the Overview tab, which lists its shelves with “Add to shelf”. For a batch, select the books in the grid or the list (Cmd+click adds one, Shift+click takes a run) and click “Add to shelf” in the selection bar. That opens a sheet with the number of books in its title (“Add 3 books to a shelf” for three) and a checkbox for each shelf. A shelf that holds only some of the selected books shows a bar and a count such as “1 of the 3 books”. Clicking a shelf that holds them all takes them off it, and “New shelf” with “Add” starts a new one.</p>\n<p>The Filter menu counts the books of each shelf. The collection on the Kobo gets only those of them that are already on the Kobo.</p>\n<h2 id=\"sending-the-shelves-to-the-kobo\">Sending the shelves to the Kobo</h2>\n<p>Connect the Kobo with a USB cable and tap Connect on its screen. On the Sync screen, the Readwyn → Device card says what it carries: “Send progress, highlights, notes, shelves and My Words from the library to the Kobo.”</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-sync.webp?h=3748d9769c62e82b8755\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-sync.webp?h=25b1a7e75f31a1231622\" alt=\"The Sync screen of a Kobo Clara 2E. The card Readwyn → Device reads Send progress, highlights, notes, shelves and My Words from the library to the Kobo, with 5 changes in 4 books and the button Review 5 changes\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Shelves travel with Readwyn → Device</figcaption>\n</figure>\n<p>Its button opens the review. Each new membership is its own row, “Add · Shelf”, with the shelf's name and “a collection on the Kobo”. Nothing is written until you apply the rows you've ticked, and the message afterwards confirms that the device database was backed up first. The write goes into the <code>Shelf</code> and <code>ShelfContent</code> tables of <code>.kobo/KoboReader.sqlite</code>, in the same guarded transaction as everything else the sync changes (<a href=\"/blog/how-readwyn-writes-to-kobo-safely/\">how those writes are kept safe</a>). A new collection gets the type <code>UserTag</code>.</p>\n<p><a href=\"/blog/eject-kindle-or-kobo-from-mac/\">Eject the Kobo</a>, and look under My Books, Collections. One caveat from Readwyn's own design notes: the table layout comes from the database of a Kobo Libra Colour that had no collections, so the rows Readwyn writes haven't yet been compared with rows a Kobo made itself.</p>\n<h2 id=\"taking-a-book-out-of-a-collection\">Taking a book out of a collection</h2>\n<p>Take a book off a shelf in Readwyn and the next Readwyn → Device review shows the collection it's still in on the Kobo. What happens to it depends on the merge option, “Merge highlights, notes and words”, which each device has and each review can change.</p>\n<p>With merging on, nothing is removed. The row reads “The book stays in “Holiday” on the Kobo; merging deletes nothing.” With merging off, the row says the book leaves this collection on the Kobo, and applying it takes the book out. Readwyn marks the old row as deleted rather than erasing it, which is how the Kobo records its own removals.</p>\n<p>A collection that loses its last book stays on the Kobo, empty. Remove it on the device: <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360033887953\">Kobo's help page</a> has you open My Books, tap Collections, tap and hold the collection's name, and tap Delete twice.</p>\n<h2 id=\"collections-you-made-on-the-kobo\">Collections you made on the Kobo</h2>\n<p>The same help page gives the steps on the device: My Books, Collections, the Menu icon, then New collection. Kobo also says a collection appears in the Kobo Books app, possibly after a delay.</p>\n<p>To bring those collections into the library, run Device → Readwyn. Its card doesn't name shelves, but its review has a row for each collection a book is in that the library doesn't have. Each one becomes a shelf of the same name for the books that are in both places.</p>\n<p>The merge option counts here too. With it off, a shelf that a book has in the library but not on the Kobo shows as a row in which the book leaves that shelf in Readwyn; untick it, or turn merging on for that review, to keep the shelf.</p>\n<p>The sync takes only collections of the type <code>UserTag</code>, or with no type at all; a collection of any other type isn't one the user made and stays out.</p>\n<p>Names are what link a shelf to a collection: a shelf and a collection with different names are two things to the sync.</p>\n<h2 id=\"what-a-sync-leaves-alone\">What a sync leaves alone</h2>\n<ul>\n<li><strong>Books that aren't on the Kobo yet.</strong> A sync covers books that are in the library, on the device, and in the Kobo's own database. It never adds or deletes a book. Send one with “Send to device” on its book page, then eject and unplug the Kobo so that it <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">imports the new book</a>. Connect it again and sync its shelves.</li>\n<li><strong>Smart shelves.</strong> A shelf that fills itself from rules stays in the library. No Kobo, Kindle or reading service gets it.</li>\n<li><strong>Books only on the Kobo.</strong> The Sync screen counts them in one line, such as “3 books on Kobo Clara 2E are not in this library”, and the sync skips them, so their collections stay as they are.</li>\n<li><strong>A Kobo without collection tables.</strong> If its database has no <code>Shelf</code> and <code>ShelfContent</code> tables, it has no collections, and the sync carries no shelves to it.</li>\n</ul>\n<h2 id=\"if-calibre-built-your-collections\">If calibre built your collections</h2>\n<p>calibre's KoboTouch driver makes collections from the library columns you pick in its settings, according to the <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Kobo_Shelves_and_Collections\">MobileRead wiki's page on Kobo shelves and collections</a>, for example a custom column of shelf names or the series column. Those collections are already on the Kobo. Import your calibre library into Readwyn, then run Device → Readwyn: each of those collections that has the type <code>UserTag</code> or no type arrives as a shelf for the books you imported (<a href=\"/blog/leave-calibre-keep-kobo-reading-history/\">leaving calibre without losing your Kobo history</a> covers the whole move). The review lists every change before anything is applied, so you can see which shelves came across. From then on, you change the shelves in Readwyn and send them back.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/metadata-calibre-and-driveinfo-calibre/","url":"https://readwyn.com/blog/metadata-calibre-and-driveinfo-calibre/","title":"What are metadata.calibre and driveinfo.calibre on my e-reader?","summary":"calibre writes both files on an e-reader: a cached list of the books it found there, and an ID for that storage. You can delete them: your books stay, and calibre rebuilds its list.","content_html":"<p>Both files come from calibre, the ebook manager, and they're there because at some point you plugged the reader into a computer running it. <code>metadata.calibre</code> is calibre's list of the books it found on the device, kept so that it doesn't have to scan every book again at the next connection. <code>driveinfo.calibre</code> holds the ID that calibre gives that storage, and says whether it's the reader's own memory or a memory card. Neither is for the reader: a <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?t=89300\">2011 forum answer</a> called the list “only a cache for Calibre”. Delete them and your books stay where they are. calibre makes a new <code>metadata.calibre</code> when it scans the device again (<a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=4179013\">Goyal, 2021</a>: “let calibre rebuild it”).</p>\n<h2 id=\"metadata-calibre\">metadata.calibre</h2>\n<p>Kovid Goyal, who created calibre, described it on MobileRead as the file “which caches the results of previous scans” (<a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?t=239311\">2014</a>). It sits at the root of the storage, next to the folders of books, and a reader with a memory card can have one on each. It has one entry for each book: the path of the file on the device, calibre's uuid and ID for the book, the title, the authors, and more. A user who opened his in 2021, after calibre showed none of his books as on the device, <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=4179013\">found entries</a> with the authors and the timestamp left blank. Goyal's advice in that thread was to “Delete the metadata.calibre file and let calibre rebuild it.”</p>\n<p>That's the usual cure when calibre's view of the device has gone wrong: books that it lists as missing, or files it shouldn't list at all. In the 2014 thread the trouble was a phone whose scanned folders had changed. The deletion worked once the owner had also taken the unwanted folder out of the scan settings. The cost of deleting the file is a full scan: calibre has to read the device again before the list is back, and a forum member in the <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=4179013\">2021 thread</a> warned against touching the files that calibre writes to the root unless you want “to do a full scan again”.</p>\n<p>On at least one reader the file once showed up as a book. In 2010 an owner of an Onyx Boox 60 <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?t=89300\">reported</a> that <code>metadata.calibre</code> appeared in the reader's book list, and that a full stop in front of the name hid it. Goyal offered to change the name that the driver writes, said it would be in the next release, and told users to delete the old file by hand after upgrading.</p>\n<h2 id=\"driveinfo-calibre\">driveinfo.calibre</h2>\n<p>Each storage of a device gets its own <code>driveinfo.calibre</code>, in its root folder. In a <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?t=333204&amp;page=2\">2020 thread</a> on calibre's MobileRead forum, the forum member chaley explained that the unique ID calibre gives a storage is kept in this file. calibre's <a rel=\"external\" href=\"https://manual.calibre-ebook.com/generated/en/template_ref.html\">template reference</a> confirms that each storage location “has its own device name” and a different uuid, under the keys <code>main</code>, <code>carda</code> and <code>cardb</code>. The file also records a location code, which on a Kobo is <code>main</code> for the device's memory and <code>A</code> for the card.</p>\n<h2 id=\"when-to-delete-them\">When to delete them</h2>\n<p>When calibre reports an error communicating with the device, Goyal's steps start with a restart of both the reader and the computer. If that fails, the next step is to find <code>driveinfo.calibre</code> and <code>metadata.calibre</code> on the device and delete them, and after that to reset the device (<a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?t=351735\">2023</a>). He gave the same steps to an owner of a <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?t=320830\">Kindle Fire HD 10</a> in 2019. Finder, Windows Explorer, or a file manager on the device can delete them.</p>\n<p>If you've stopped using calibre, there's no need to delete them. If you sync a Kobo with Readwyn, there's a reason to keep them. Readwyn, the Mac and Linux ebook library app behind this blog, reads <code>metadata.calibre</code> on a Kobo to tell which book on the device is which book in your library. When you <a href=\"/blog/import-calibre-library/\">import a calibre library</a>, Readwyn keeps each book's calibre uuid, and the list on the Kobo links that uuid to a file path, so books that calibre sent match exactly. Without the files, a book that Readwyn hasn't matched in an earlier sync is matched by its title, and also by its first author if your library has more than one book with that title (<a href=\"/blog/sync-kobo-with-mac-without-calibre/\">how that sync works</a>). The Kobo's own data lives somewhere else, in the <a href=\"/blog/kobo-folder-and-koboreader-sqlite/\">hidden .kobo folder</a>.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/move-books-from-old-kobo-to-new-kobo/","url":"https://readwyn.com/blog/move-books-from-old-kobo-to-new-kobo/","title":"How to move your books from an old Kobo to a new Kobo","summary":"Sign the new Kobo in to the same Kobo account and your Kobo Store books come with it. Files you copied to the old Kobo by cable have to be copied across by hand.","content_html":"<p>Your Kobo Store books belong to your Kobo account, not to the eReader, so the move is mostly a sign-in. Sync the old Kobo one last time, then sign the new one in to the same account, and your purchases appear in its library. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019527954-Download-books-from-your-Kobo-account-to-export-to-another-device-or-app\">exporting books</a> says store purchases sync to Kobo eReaders and apps on their own, so there is nothing to export. What the account doesn't hold is anything you put on the old Kobo with a cable: EPUBs and PDFs from your computer, fonts, and the highlights you made in those books. Those you copy across yourself.</p>\n<h2 id=\"sync-the-old-kobo-first\">Sync the old Kobo first</h2>\n<p>On the old Kobo's Home screen, tap the Sync icon at the top, then “Sync now”. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017513334-Sync-your-Kobo-eReader-over-Wi-Fi\">syncing over Wi-Fi</a> says a sync “updates any bookmarks or annotations that you've made to your eBooks”, so this is the step that puts your latest place and notes in store books where the new Kobo can find them.</p>\n<p>In September 2026 Kobo said it had rebuilt its bookmark and progress sync, and the fix wasn't automatic. Nathan Groezinger <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2026/09/11/sync-issues-with-kobo-ereaders-and-apps-finally-fixed/\">wrote at The eBook Reader</a> that each eReader needs Settings, Device information, Repair your Kobo account, and that you have to do it on every Kobo device or app you want to keep in step. If the old Kobo hasn't had that since, run it before the last sync.</p>\n<h2 id=\"sign-in-on-the-new-kobo\">Sign in on the new Kobo</h2>\n<p>Setup asks for a language and a Wi-Fi network, installs any update (Kobo says that can take up to 20 minutes), and then asks you to sign in. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017439594-Set-up-your-Kobo-eReader-over-Wi-Fi\">setup page</a> says that if you already have Kobo books, you'll see them being added to the eReader at this point.</p>\n<p>If the library comes up empty, the usual cause is a different account. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017765893-Verify-that-you-sign-in-to-the-correct-Kobo-account\">checking the account</a> has you tap More, then Settings, then Accounts, where the address under “Kobo” is the one signed in. It should match the address on your Kobo purchase receipts. A book you once archived on kobo.com won't show either: <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017765873-Move-items-out-of-your-account-archive\">restore it</a> from the Archive tab of My Books on the website with “Restore to Library”, and it appears after the next sync.</p>\n<p>Not every book is downloaded at once. Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605034-Download-all-your-books-onto-your-Kobo-eReader\">downloading all your books</a> says the eReader downloads your five most recent purchased or opened books. The rest wait online: Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017513294\">the download icon</a> says they show a download icon on the cover, or the word “Download” beside the title in list view. Tap one to fetch it. To fetch them all, open My Books, tap the More icon (the three dots in the grey bar), and tap “Download all”. Do that before a trip if you'll read without Wi-Fi.</p>\n<p>Collections you make show on both your eReader and the Kobo Books app, according to Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360033887953-Organize-your-eBooks-on-your-Kobo-eReader\">organising books</a>, though “there may be a delay” before a new one appears on the other.</p>\n<p>Notebooks, on the models that take a Kobo Stylus (the Sage, Elipsa, Elipsa 2E and Libra Colour), sync with your Kobo account by default, so a new eReader on the same account can open them (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360062226733-Use-your-Kobo-eReader-as-a-notebook\">Kobo's notebook page</a>). The same page also gives a cable route between a Sage and an Elipsa: copy the files ending in <code>.nebo</code> off the old eReader, then onto the new one.</p>\n<h2 id=\"what-you-copied-by-cable\">What you copied by cable</h2>\n<p>Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">adding EPUB and PDF files from a computer</a> is clear about these: “Your content will only appear on the device that you used to import the content.” No sync moves them, as <a href=\"/blog/does-kobo-sync-sideloaded-books/\">does a Kobo sync sideloaded books?</a> explains. If the originals are still on your computer, copy them to the new Kobo. If the old Kobo has the only copies, connect it with a USB cable, tap Connect when it shows “Computer detected”, and copy the files from the KOBOeReader drive to a folder on the computer. Then eject it, connect the new Kobo, and copy them on. If the drive doesn't appear on a Mac, see <a href=\"/blog/kobo-not-showing-up-on-mac/\">why a Kobo doesn't show up on a Mac</a>.</p>\n<p>Two more things live only on the device:</p>\n<ul>\n<li><strong>Fonts.</strong> Kobo reads your own .ttf and .otf files from a folder named <code>fonts</code> on the KOBOeReader drive (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/13009477876631-Load-fonts-onto-your-Kobo-eReader\">Kobo's page</a>). Copy the folder across, or <a href=\"/blog/add-fonts-to-kobo-and-kindle/\">add them again</a>.</li>\n<li><strong>Dropbox and Google Drive.</strong> You link these from the eReader itself, with a code that it shows, so link the new Kobo the same way. Kobo names a change of eReader as one reason to <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360033882694-Turn-off-Dropbox-sharing-on-your-eReader\">sign the old one out of Dropbox</a>. Only the Forma, Sage, Elipsa, Elipsa 2E and Libra Colour have the feature (<a href=\"/blog/send-books-to-kobo-with-dropbox-or-google-drive/\">setting up Dropbox or Google Drive on a Kobo</a>).</li>\n</ul>\n<p>In six countries (Australia, Canada, Ireland, Great Britain, New Zealand and Spain), Kobo sells a subscription called Kobo Stack. <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/38802863972503-About-Kobo-Stack-features-supported-devices-and-how-to-sign-up\">Its page</a> lists up to 50 GB of storage for your own PDF and EPUB files, 200 MB each at most, and names three eReaders, the Libra Colour, Clara Colour and Clara BW, on software 6.0 or later. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/42850160943511-Kobo-Cloud-upload-eBooks-and-documents-with-Kobo-Stack\">Kobo Cloud page</a> says uploaded titles are available on Kobo apps and supported Kobo devices after syncing. If both your Kobos are on that list, an upload is one way to move your own files without a cable.</p>\n<h2 id=\"highlights-in-your-own-books\">Highlights in your own books</h2>\n<p>Copying a book across doesn't copy what you marked in it. A Kobo keeps your place, highlights and notes for a sideloaded book in its own database on the eReader, and no sync carries them. Readwyn, the ebook library app for Mac and Linux that this blog belongs to, can move them for books that are also in your Readwyn library. Connect the old Kobo and apply “Device → Readwyn” on the Sync screen, which brings the progress, highlights and notes of each book on both sides into the library. Then send the books to the new Kobo with “Send to device”, eject it so it adds them, connect it again, and apply “Readwyn → Device” (<a href=\"/blog/sync-reading-progress-kobo-and-mac/\">how the two directions work</a>).</p>\n<h2 id=\"the-old-kobo-afterwards\">The old Kobo afterwards</h2>\n<p>It can stay signed in and keep working beside the new one. If it's going to someone else, a factory reset deletes the books you added from your computer and any bookmarks and annotations since the last sync (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605114-Reset-your-Kobo-eReader\">Kobo's reset page</a>), so copy those off first. <a href=\"/blog/before-you-sell-a-kindle-or-kobo/\">Before you sell a Kindle or a Kobo</a> has the steps and the rest of that list.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/put-cbz-comics-on-kobo/","url":"https://readwyn.com/blog/put-cbz-comics-on-kobo/","title":"How to put CBZ comics on a Kobo from a Mac","summary":"Plug the Kobo in, open the comic in Readwyn, and pick “Send to device” from the “…” menu. It lands in the Books folder as a CBZ, never converted.","content_html":"<p>Kobo e-readers open CBZ files themselves, so a comic needs no conversion to get there. Connect the Kobo to your Mac with a USB cable and tap Connect on its screen. Open the comic in Readwyn, click “…” (“More actions”) on its book screen, and choose “Send to device”. Eject the Kobo when the send ends, and the comic turns up in My Books once the Kobo has imported it.</p>\n<h2 id=\"what-the-kobo-reads\">What the Kobo reads</h2>\n<p>Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">list of supported file formats</a> gives two comic formats for its e-readers: CBZ and CBR. Of the two, Readwyn only ever sends CBZ. A CBZ goes over as the very file you added. A CB7 or CBT goes as its <a href=\"/blog/open-cb7-and-cbt-comics/\">reading copy</a>, a CBZ that Readwyn keeps in its cache with page images byte for byte the same as the original's.</p>\n<p>The setting “Convert EPUB to KEPUB for Kobo e-readers” makes no difference here. It applies to ebooks; comics always go as CBZ, never as KEPUB.</p>\n<h2 id=\"where-the-file-lands\">Where the file lands</h2>\n<p>Readwyn names the file after the comic's first author and title, in the Kobo's <code>Books</code> folder:</p>\n<p><code>Books/&lt;first author&gt; - &lt;title&gt;.cbz</code></p>\n<p>Those names come from the library, so for a comic with a <a href=\"/blog/comicinfo-xml/\"><code>ComicInfo.xml</code></a> inside, the author is the first <code>Writer</code>, else the first <code>Penciller</code>, else the first <code>CoverArtist</code>. A comic with no author at all goes as <code>Books/Unknown - &lt;title&gt;.cbz</code>. A colon, a slash, a question mark, or another character that a file name on the Kobo can't hold becomes an underscore. If the file name on the Kobo matters to you, fix the author and title in Readwyn before you send.</p>\n<p>While the Kobo is connected and has the file, the item in the “…” menu reads “Send to device again”, and the book screen's Details tab shows a row, “On the device”, with the copy on the Kobo and a “Remove from device” button.</p>\n<h2 id=\"why-it-isn-t-in-my-books-yet\">Why it isn't in My Books yet</h2>\n<p>A send copies the file and nothing more. Readwyn writes no row for the comic to the Kobo's database, <code>KoboReader.sqlite</code>; the Kobo adds that row itself when it leaves USB mode. Until then the file sits on the Kobo's storage, and Readwyn already counts it as on the device.</p>\n<p>So eject before you unplug. “Eject device” on the Sync screen does it for you. If Readwyn is still writing, it says so first (“Readwyn is still writing to the Kobo”), and “Stop and eject” waits up to 30 seconds for the write to end. If the write is still going after that, the Kobo stays connected and Readwyn asks you to try again. Kobo's own guide to <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">copying files from a computer</a> says what happens next on the device: an “Importing content” message, and then the book is under My Books on the Home screen. If the comic still doesn't show up, the causes are much the same as for an <a href=\"/blog/epub-not-showing-on-kobo/\">EPUB that doesn't show on a Kobo</a>, and <a href=\"/blog/eject-kindle-or-kobo-from-mac/\">ejecting a Kobo from a Mac</a> covers the eject itself.</p>\n<h2 id=\"webp-pages\">WebP pages</h2>\n<p>Readwyn reads comic pages in JPEG, PNG, GIF, and WebP, and it never re-encodes a page, so what's in your archive is what reaches the Kobo. Kobo's format list names JPEG, GIF, PNG, BMP, and TIFF as the image files its e-readers open. WebP isn't on it, and the page says nothing about the images inside a comic archive. If a comic of yours has WebP pages, open it on the Kobo before you count on reading it there.</p>\n<h2 id=\"cbr-files\">CBR files</h2>\n<p>Readwyn doesn't take CBR, the RAR flavour of comic archive, because the licence of the RAR decoder isn't one that the Open Source Initiative has approved. A CBR can't go through “Send to device”, even though the Kobo lists CBR as a format it reads.</p>\n<h2 id=\"more-than-one-comic\">More than one comic</h2>\n<p>“Send to device” goes one book at a time, from each book screen. The selection bar for many books doesn't offer it, so a run of twelve issues means twelve sends before you eject. <a href=\"/blog/send-books-to-kobo-from-mac/\">Sending books to a Kobo</a> goes through the whole send for ebooks too, and <a href=\"/blog/comics-and-ebooks-in-one-library/\">keeping comics and ebooks in one library</a> covers what a comic looks like in Readwyn before it ever reaches the device.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/remove-book-from-kobo-keep-highlights/","url":"https://readwyn.com/blog/remove-book-from-kobo-keep-highlights/","title":"How to remove a book from a Kobo and keep its highlights","summary":"Run a Device → Readwyn sync first, so the highlights are in your library. Then use “Remove from device” on the book's Details tab. The highlights stay on the Mac.","content_html":"<p>Copy the highlights off the Kobo before the book goes. In Readwyn, that's a Device → Readwyn sync with the Kobo plugged in. Then open the book's Details tab and click “Remove from device”. Readwyn deletes the book file from the Kobo, and the book, its highlights, its notes and its reading data stay in your library.</p>\n<p>The order matters because of where a Kobo keeps notes. For a book that didn't come from the Kobo store, Kobo's help page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017481314-Remove-notes-from-your-eBook\">removing notes</a> says the notes are saved on the eReader only, and that they're removed if you delete the book on the eReader. Nothing else holds a copy until you make one.</p>\n<h2 id=\"1-bring-the-highlights-in\">1. Bring the highlights in</h2>\n<p>Connect the Kobo with a USB cable, tap Connect on its screen, and open the Sync screen in Readwyn. The card “Device → Readwyn” counts the changes that would come from the Kobo, for example “8 changes in 3 books”. Click “Review 8 changes”.</p>\n<p>The review lists every change by book, each new highlight with its text, and each note under its highlight. Most rows start ticked. A row where the library's version is newer starts unticked, so look at those and tick the ones you want. Then click the button at the foot, which counts the ticked rows (“Apply 7 changes”). The library now holds each highlight with its note and its colour, and, where Readwyn can place it in the book, its page and the paragraphs around it (<a href=\"/blog/kobo-notes-in-your-library/\">what a note looks like in the library</a>).</p>\n<p>One case needs a step more. A sync only covers books that are both on the Kobo and in the library, so highlights in a book the library doesn't have are listed under “Left out” with “Not in this library.” and never come in. Add the book to the library first. A book that no other record matches is matched by its title (<a href=\"/blog/kobo-books-not-in-library/\">how matching works</a>).</p>\n<h2 id=\"2-remove-the-book\">2. Remove the book</h2>\n<p>Open the book in Readwyn and go to its Details tab. With the Kobo connected, the row “On the device” shows the copy on it: the format, the device, and the date it was sent.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book-details.webp?h=3f5c56cb7a1e9927112f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book-details.webp?h=647f1b629b6b477420d1\" alt=\"The Details tab of Middlemarch: the original MOBI, the EPUB 3 converted from it, and a KEPUB on a Kobo Clara 2E, sent 2 Oct 2026, with the button Remove from device\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>“Remove from device” sits in the row of the copy on the Kobo</figcaption>\n</figure>\n<p>Click “Remove from device”. It's the only action in Readwyn that deletes a book from a device. On the Kobo it deletes the book file and nothing else, and in the library it forgets where the book was on the Kobo. It writes nothing to the Kobo's database, so the Kobo's own record of the book stays until the Kobo next looks for books, when it leaves USB mode. While the Kobo is still adding books from an earlier send, “Remove from device” can delete only a book that the Kobo hasn't added yet.</p>\n<p>Then click “Eject device” on the Sync screen and unplug the cable. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">steps for books added from a computer</a> end the same way, and the Kobo then shows “Importing content” while it updates its list.</p>\n<h2 id=\"what-stays-on-the-mac\">What stays on the Mac</h2>\n<p>The library keeps the book file and every highlight that came in, with its note, and its page and context where it has them. Removing the book doesn't change that, and neither does a later sync. A sync covers only books that are on the Kobo at the time, and a book that has gone from the device keeps its highlights in the library even when you sync with the merge option off.</p>\n<p>The highlights show in the Notebook and on the book's “Highlights” tab. “Export” in the Notebook saves them as a Markdown or plain-text file, each highlight with its page and note (<a href=\"/blog/export-kobo-highlights-on-mac/\">exporting Kobo highlights</a>).</p>\n<h2 id=\"putting-the-book-back-later\">Putting the book back later</h2>\n<p>A removed book drops out of the sync: Readwyn no longer records it on the Kobo, so the reviews leave it out. To put it back, use “Send to device” in the “…” menu of the book screen, which copies it to the Kobo again (<a href=\"/blog/send-books-to-kobo-from-mac/\">sending books to a Kobo</a>).</p>\n<p>Once the Kobo has added it, a Readwyn → Device sync writes the library's highlights back into the Kobo's database, in the same form as the Kobo's own, with their colours. A highlight that has no place in the copy on the Kobo isn't written, and the review names it and says why. Before the write, Readwyn backs up <code>KoboReader.sqlite</code> on the Mac (<a href=\"/blog/how-readwyn-writes-to-kobo-safely/\">how those writes work</a>).</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/send-books-to-kobo-with-dropbox-or-google-drive/","url":"https://readwyn.com/blog/send-books-to-kobo-with-dropbox-or-google-drive/","title":"How to send books to a Kobo with Dropbox or Google Drive","summary":"Link the account from the Kobo with a code, put DRM-free files in the Rakuten Kobo folder, and sync. Only the Forma, Sage, Elipsa, Elipsa 2E and Libra Colour can do it.","content_html":"<p>You link the cloud account from the Kobo itself: the Kobo shows a code, you type that code into a page on kobo.com from a computer or a phone, and a folder called “Rakuten Kobo” appears in the account. Any DRM-free EPUB or PDF that you put in that folder shows up on the Kobo after a sync over Wi-Fi. Five models can do this, according to Kobo's help pages for <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360033830114-Add-books-to-your-eReader-using-Dropbox\">Dropbox</a> and <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/15335985512983-Add-books-to-your-eReader-using-Google-Drive\">Google Drive</a>: the Kobo Forma, Sage, Elipsa, Elipsa 2E and Libra Colour. A Clara or a Libra 2 isn't on either list.</p>\n<h2 id=\"what-the-kobo-needs\">What the Kobo needs</h2>\n<p>Each service has a minimum software version. Dropbox needs 4.18 or later and Google Drive needs 4.37 or later. The Kobo shows its version beside “Software version” under More › Settings › Device information (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024773213-Find-your-eReader-s-software-version\">Kobo's page</a>), and <a href=\"/blog/which-kobo-do-i-have/\">which Kobo do I have</a> covers the model number if you aren't sure which one you have.</p>\n<p>The Kobo has to be on Wi-Fi for the linking and for every transfer. A free Dropbox account links to three devices, and Kobo's page says the Kobo account doesn't count towards that limit.</p>\n<h2 id=\"dropbox\">Dropbox</h2>\n<ol>\n<li>On the Kobo, tap More › Settings › Accounts, and under Dropbox tap “Get Started”. The Kobo shows a code.</li>\n<li>On a computer or a phone, go to <code>kobo.com/dropbox</code>, type the code, click “Link with Dropbox”, sign in, and click “Allow”.</li>\n<li>In the Dropbox app or on the website, put your files in Apps › Rakuten Kobo.</li>\n<li>On the Kobo, tap Sync on the Home screen, then “Sync now”.</li>\n</ol>\n<p>The files are listed under More › My Dropbox. Tap “Refresh” if a new one isn't there yet, and tap a file to download it and open it.</p>\n<h2 id=\"google-drive\">Google Drive</h2>\n<ol>\n<li>On the Kobo, tap More on the Home screen, then “My Google Drive”, then “Get Started”. Note the code.</li>\n<li>On a computer or a phone, go to <code>kobo.com/googledrive</code>, type the code and click “Continue”. Click “Continue” again on the screen that asks you to approve the data sharing, sign in to Google, click “Allow”, and then “Go to Google Drive”.</li>\n<li>Upload your files to the “Rakuten Kobo” folder, which appears in your Drive on its own once the accounts are linked. A file elsewhere in the Drive doesn't reach the Kobo.</li>\n<li>Sync the Kobo from the Sync icon on the Home screen.</li>\n</ol>\n<p>More › My Google Drive lists the folder. Tapping a file copies it to My Books, and the Menu icon there has “Refresh” and “Download all”, so a whole folder can come across in one go.</p>\n<h2 id=\"what-can-go-through-the-folder\">What can go through the folder</h2>\n<p>Both pages name DRM-free EPUB and PDF files and point to Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">list of supported formats</a> for the rest: MOBI, CBZ and CBR comics, plain text, HTML, RTF, and images. A file with DRM can't take this route: for those, Kobo sends you to Adobe Digital Editions instead. A book bought in the Kobo Store needs neither, only a sync.</p>\n<p>If a file sits in the folder and never appears on the Kobo, look at its name first. Both pages ask you to avoid these characters in file names: <code>|</code>, <code>\"</code>, <code>:</code>, <code>?</code>, <code>&lt;</code> and <code>&gt;</code>. Then check that the file is in the “Rakuten Kobo” folder and not one level above it, and that the Kobo was on Wi-Fi when it synced. For a book that is on the Kobo but still missing from My Books, <a href=\"/blog/epub-not-showing-on-kobo/\">an EPUB that doesn't show up on your Kobo</a> goes through the other checks: the filters of My Books, DRM, and the device you look on.</p>\n<h2 id=\"unlinking\">Unlinking</h2>\n<p>For Dropbox, <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360033882694-Turn-off-Dropbox-sharing-on-your-eReader\">Kobo's sign-out page</a> lists two separate steps. On the Kobo, go back to More › Settings › Accounts, where you linked it, and tap “Sign out” under Dropbox, then “Sign out” again to confirm. In Dropbox, go to <code>dropbox.com/account/connected_apps</code>, choose Rakuten Kobo and click “Disconnect”. Kobo suggests the second step when you change e-readers or lose one. For Google Drive, the unlink happens in Google's <a rel=\"external\" href=\"https://drive.google.com/drive/settings\">Drive settings</a>: under “Manage Apps”, find Kobo, open “Options”, and click “Disconnect from Drive”.</p>\n<p>Either way, the books already on the Kobo stay there. Unlinking stops only new downloads.</p>\n<h2 id=\"the-other-kobos\">The other Kobos</h2>\n<p>Every other Kobo, a Clara or a Libra 2 among them, takes your own files over a USB cable: connect it, tap “Connect”, drag the files onto the drive called KOBOeReader, and eject, as Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">page on adding files with a computer</a> describes. On a Mac, Readwyn can do the copying: <a href=\"/blog/send-books-to-kobo-from-mac/\">“Send to device”</a> puts a book on a plugged-in Kobo, converted to KEPUB unless you turn that off.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/use-kindle-or-kobo-without-wifi/","url":"https://readwyn.com/blog/use-kindle-or-kobo-without-wifi/","title":"Can you use a Kindle or a Kobo without Wi-Fi?","summary":"Yes for reading: both read their books with Wi-Fi off and take files over USB. Setting one up is the catch, because Amazon and Kobo both expect an account and the internet.","content_html":"<p>You can read on either one without Wi-Fi, once it's set up. The books already on a Kindle or a Kobo open with the radio switched off, and both take new files from a computer over a USB cable. The first start is the hard part. Amazon's own article on <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/register-kindle-account\">setting up a Kindle</a> says \"Before you can register your Kindle, you need to be connected to the internet\", and Kobo's help page <a rel=\"external\" href=\"https://help.kobo.com/hc/articles/360045168633\">Set up your eReader with Wi Fi or your computer</a> says that to sign in \"you'll need a Kobo account\". Neither maker describes a setup that never touches the internet.</p>\n<h2 id=\"the-first-start-of-a-kindle\">The first start of a Kindle</h2>\n<p>The <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_11th/v1/en-US/html/kug.html\">user guide of the 11th-generation Kindle</a> lists registration to your Amazon account as a step of setup. With the Kindle app on your phone, or another Amazon device on your network, it can pick up a Wi-Fi password that you saved to Amazon. Without those, \"the device will ask you to connect to a wireless network.\" If you bought the Kindle online with your Amazon account, it arrives registered to you.</p>\n<p>The newest models make the point bluntly. According to <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2026/10/04/how-to-bypass-registration-on-new-kindles-for-use-without-amazon-account/\">The eBook Reader</a>, the product pages of the 2026 Kindles say \"Wi-Fi and an Amazon account are required to set up your Kindle\". The same site found a way past the sign-in screen in October 2026: hold the power button there and choose Restart, and the Kindle boots to its home screen unregistered. It isn't an offline setup, though. In that test the Kindle had already joined Wi-Fi and downloaded a software update before the sign-in screen appeared. An unregistered Kindle still takes sideloaded books, dictionaries and fonts, the article says, but a sign-in prompt pops up each time you open the library, and online features such as Wikipedia lookup don't work. Amazon doesn't document any of this.</p>\n<h2 id=\"the-first-start-of-a-kobo\">The first start of a Kobo</h2>\n<p>A Kobo gives you a second route. After you choose a language, tap \"Don't have a Wi Fi network?\" and connect the Kobo to a computer that runs the Kobo Desktop app. Kobo's page <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017598633\">Set up your Kobo eReader for the first time using a computer</a> has the steps, and three conditions:</p>\n<ul>\n<li>The computer must be connected to the internet, and you still sign in with a Kobo account.</li>\n<li>Kobo Desktop \"does not work for customers in Europe.\"</li>\n<li>Kobo Desktop is not compatible with the Kobo Libra Colour, the Kobo Clara Colour or the Kobo Clara BW.</li>\n</ul>\n<p>So the Kobo itself can be set up without Wi-Fi, but not without an internet connection somewhere, and with those three models (or in Europe) Wi-Fi is the only route on Kobo's pages.</p>\n<p>There's also a mode that Kobo's help centre doesn't describe. In 2022 The eBook Reader reported that firmware 4.31.19086 added a <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2022/02/10/how-to-use-kobos-new-sideloaded-mode/\">Sideloaded Mode</a>, which lets you use a Kobo without a Kobo account: you sign out of the account in Settings. The article says the mode is meant for use during the first setup by people who don't have an internet connection, but it doesn't give the steps for a new Kobo. In that mode there's no syncing and no link to the store. Be careful with the sign-out itself. Kobo's page <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605094-Sign-out-of-your-Kobo-eReader\">Sign out of your Kobo eReader</a> says your books and bookmarks are kept, but you \"will lose highlight, notes, and annotations\".</p>\n<h2 id=\"what-works-with-wi-fi-off\">What works with Wi-Fi off</h2>\n<p>On a Kindle, Airplane Mode \"disables wireless connectivity, including Wi-Fi and Bluetooth\", the user guide says. Before you go offline, it tells you to download to the Kindle anything that you want to have with you. The store needs a connection (\"Your Kindle must have an active Wi-Fi connection to use this feature\"), and so does Amazon's sync. Whispernet \"delivers and syncs content to your Kindle whenever you have a Wi-Fi connection\", and Sync is what keeps track of \"your last page read, notes, highlights, and bookmarks\", so those reach your other devices only after the Kindle is online again.</p>\n<p>New books can still arrive by cable. A Kindle takes files in its <code>documents</code> folder, and Amazon's help page <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Connect, Browse, and Transfer Files on E-Reader</a> says that on a Mac the models released before 2024, apart from the Kindle Scribe, need no extra app for that. The Scribe and the Kindles released in 2024 need a separate app, such as Amazon's Send to Kindle for Mac. The page doesn't say which group the later models are in. Two warnings from the user guide: the Kindle can't be used for reading while it's in USB drive mode, and a file that you copied over USB is \"permanently deleted\" when you remove it. A Kindle won't open an EPUB copied over the cable, so <a href=\"/blog/mobi-azw3-or-epub-for-kindle-usb/\">pick the format first</a>. If the Kindle doesn't appear on the Mac at all, see <a href=\"/blog/kindle-not-showing-up-on-mac/\">why a Kindle doesn't show up on a Mac</a>.</p>\n<p>A Kobo is simpler over a cable. Kobo's page <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">Add non-protected PDF and ePub files to your Kobo eReader</a> has you tap Connect on the Kobo, drag the files onto the drive named KOBOeReader, and eject it. The page doesn't mention Wi-Fi at all. Books with DRM are the exception: the page covers non-protected files only. If the drive doesn't show up, see <a href=\"/blog/kobo-not-showing-up-on-mac/\">why a Kobo doesn't show up on a Mac</a>.</p>\n<h2 id=\"your-reading-without-the-cloud\">Your reading, without the cloud</h2>\n<p>What an offline e-reader loses is the copy of your progress and highlights on its maker's servers. For a Kobo, Readwyn keeps a copy on your Mac instead, over the same USB cable: the Sync screen says “Connected over USB”, and “Device → Readwyn” brings the Kobo's progress, highlights, notes and time read into the library on your Mac. <a href=\"/blog/sync-kobo-with-mac-without-calibre/\">How to sync a Kobo with a Mac without calibre</a> shows the steps, and <a href=\"/blog/send-books-to-kobo-from-mac/\">sending books to a Kobo</a> covers the other direction.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-kobo-plus/","url":"https://readwyn.com/blog/what-is-kobo-plus/","title":"What is Kobo Plus, and what does it include?","summary":"Kobo Plus is Rakuten Kobo's monthly subscription: Read for ebooks, Listen for audiobooks, or both. You borrow the books, and they go when you cancel.","content_html":"<p>Kobo Plus is Rakuten Kobo's subscription for ebooks and audiobooks. You pay <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018976253\">a fixed price each month</a> and read or listen to as many books from its catalogue as you like, on a Kobo e-reader or in the Kobo app. It comes in three plans: Kobo Plus Read (ebooks), Kobo Plus Listen (audiobooks), and Kobo Plus Read &amp; Listen (both). Nothing you read through it is yours to keep: Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018976173-Cancel-or-renew-your-Kobo-Plus-subscription\">cancellation page</a> says plainly that once you cancel, \"You'll no longer have access to your Kobo Plus books.\"</p>\n<h2 id=\"which-books-are-in-it\">Which books are in it</h2>\n<p>Not the whole Kobo store. Only a title that shows \"Read with Kobo Plus\" on its page is included, and Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018976153\">common questions page</a> puts the ebook side at \"hundreds of thousands of eBooks\", with more being added. Everything else in the store still has to be bought.</p>\n<p>You add Kobo Plus books on kobo.com, on a Kobo e-reader, or in the Kobo app for Android, according to the same page. It doesn't list the iOS app for that.</p>\n<h2 id=\"what-it-costs\">What it costs</h2>\n<p>Kobo's help pages give no prices. They say the price depends on where you live, that it excludes tax, and they send you to <a rel=\"external\" href=\"https://www.kobo.com/plus\">kobo.com/plus</a> for your local figure. When Kobo Plus reached the United States in April 2023, <a rel=\"external\" href=\"https://techcrunch.com/2023/04/04/kobo-plus-subscription-plans-are-now-available-in-the-u-s/\">TechCrunch reported</a> $7.99 a month for Read, $7.99 for Listen, and $9.99 for both. Check the store page for today's price before you sign up.</p>\n<p>A free trial needs a valid card, and prepaid Mastercard and Visa cards are refused. How long it lasts <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018811614-Subscribe-to-Kobo-Plus-on-kobo-com\">depends on the region</a>: the common questions page gives 14 days for customers in the US, Taiwan, and Hong Kong, and for FNAC customers, and kobo.com/plus shows the offer for your country. Residents of Quebec get no trial at all, and someone who has subscribed to Kobo Plus before <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018976173-Cancel-or-renew-your-Kobo-Plus-subscription\">can't have a second one</a>.</p>\n<h2 id=\"changing-plans\">Changing plans</h2>\n<p>From Read or Listen you can move up to Read &amp; Listen, and from Read &amp; Listen down to either single plan, under “My Account” and “My Subscriptions” on kobo.com with “Change Plan”. If you change in the middle of a billing period, you pay the difference. Going straight from Read to Listen, or back, isn't possible: Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/11547634913047-Change-your-Kobo-Plus-subscription-plan\">plan page</a> says you cancel the one and sign up for the other.</p>\n<p>The <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018976153\">common questions page</a> adds that you can't manage the subscription on a Kobo e-reader or in the iOS app. The Android app shows its details under Settings and “Kobo Plus Subscription”, but the plan page gives only kobo.com for a change of plan.</p>\n<h2 id=\"devices-and-offline-reading\">Devices and offline reading</h2>\n<p>The <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018811614-Subscribe-to-Kobo-Plus-on-kobo-com\">subscription page</a> names two Kobo e-readers that don't support it: the Kobo Mini and the Kobo Wi-Fi. The common questions page adds that Kobo Plus books can't be downloaded and imported to a Sony Reader.</p>\n<p>The Listen plan doesn't reach every Kobo. Only the Clara 2E, Sage, Libra 2, Elipsa, Elipsa 2E, Clara BW, Clara Colour, and Libra Colour <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/4406292712471-Listen-to-audiobooks-on-your-Kobo-eReader\">play audiobooks</a>, and only through Bluetooth headphones or a Bluetooth speaker. If you aren't sure which model you have, <a href=\"/blog/which-kobo-do-i-have/\">which Kobo do I have</a> tells them apart.</p>\n<p>Reading without a connection has a cap. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018811734-Reading-Kobo-Plus-books-while-offline\">offline reading page</a> lets you open at most 15 Kobo Plus books during the trial period. After 15 books, or when the trial ends, you reconnect to Wi-Fi and sync before you can open more. <a rel=\"external\" href=\"https://eftm.com/2021/11/rakuten-kobo-launches-its-all-you-can-read-subscription-service-in-australia-218388\">EFTM's report of the Australian launch</a> gives the same cap of 15 books, over a 30-day period.</p>\n<h2 id=\"where-it-s-sold\">Where it's sold</h2>\n<p>Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018811614-Subscribe-to-Kobo-Plus-on-kobo-com\">subscription page</a> lists 37 countries in November 2026, among them the US, Canada, the UK, Ireland, Australia, New Zealand, India, South Africa, much of Europe (though not Germany or Austria), Mexico and several South American countries, and parts of Asia. If you change your billing address to a country without Kobo Plus, the <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018976153\">common questions page</a> says you keep your books until the current billing period ends, but you can't add new ones, and the subscription is then cancelled.</p>\n<h2 id=\"what-stays-when-you-cancel\">What stays when you cancel</h2>\n<p>The books go. Your highlights and notes can come with you if you export them first: Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/29991333812631-Export-annotations-from-your-books\">export page</a> covers books bought from Kobo or read through Kobo Plus, as PDF, HTML, TXT, or Markdown, though it warns that the export depends on the author's approval and may not work for every book. <a href=\"/blog/moving-from-kobo-to-kindle/\">Moving from Kobo to Kindle</a> walks through that export and what else does or doesn't move.</p>\n<p>For the author's side (how Kobo Plus pays for each minute read, and how that compares with Amazon's subscription), see <a href=\"/blog/how-kindle-unlimited-pays-authors/\">how Kindle Unlimited pays authors</a>.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-koreader/","url":"https://readwyn.com/blog/what-is-koreader/","title":"What is KOReader?","summary":"KOReader is a reading app for e-ink devices, made by volunteers, that runs on Kindle, Kobo, PocketBook, Android and Linux and opens formats such as FB2 and DjVu that a Kobo's own reader does not.","content_html":"<p>KOReader is a reading app for e-readers, separate from the software that the device came with. Its own site calls it <a rel=\"external\" href=\"https://koreader.rocks/\">“a document viewer for E Ink devices”</a> and names Kindle, Kobo, PocketBook, Android and desktop Linux as the places it runs. Volunteers write and support it. Compared with the reader that a Kobo ships with, it opens more file formats, FB2 and DjVu among them, and it lets you change much of how a page looks and what a tap does.</p>\n<p>The name started as “Kindle/Kobo Open Reader”. The <a rel=\"external\" href=\"https://koreader.rocks/user_guide/\">user guide</a> admits that it has outgrown that: it spells out a longer joke version that adds PocketBook, reMarkable, Cervantes, Sony, Android, Linux and macOS.</p>\n<h2 id=\"the-formats-it-opens\">The formats it opens</h2>\n<p>The guide's list is PDF, PDB, XPS, EPUB, MOBI, FB2, DjVu, CBR, CBT, CBZ, DOCX, RTF, HTML, HTMLZ, TXT, Markdown and CHM, plus ZIP archives and plain image files. Neither <a href=\"/blog/what-is-fb2/\">FB2</a> nor <a href=\"/blog/what-is-djvu/\">DjVu</a> is on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">Kobo's list of the formats its eReaders support</a>, and KOReader on the same Kobo opens both.</p>\n<p>PDFs get their own chapter in the guide. KOReader can crop the margins automatically or by hand, reflow the text of a page, run OCR on a scanned one, and zoom column by column. Comics and manga have a panel zoom.</p>\n<h2 id=\"getting-books-onto-it\">Getting books onto it</h2>\n<p>You can copy books the same way as for the device's own reader, for example over USB. The guide then lists other ways, some only on certain devices:</p>\n<ul>\n<li>a USB mass storage mode inside KOReader, on a Kobo and some Cervantes devices</li>\n<li>cloud storage: Dropbox, FTP and WebDAV</li>\n<li>SSH and SFTP, on a Kobo, a Kindle or a Cervantes</li>\n<li>calibre, which can send books from its library to KOReader over Wi-Fi</li>\n<li>OPDS catalogues, public ones or your own server</li>\n<li>a news downloader for RSS and Atom feeds, and Wallabag for saved web articles</li>\n</ul>\n<p>Books then turn up in a file browser, as a list or as a mosaic of covers, with Favorites, History and Collections on top of it.</p>\n<h2 id=\"what-reading-in-it-is-like\">What reading in it is like</h2>\n<p>Long-press a word and KOReader looks it up in an installed dictionary, on Wikipedia, or in a translation. Highlights come in four styles (lighten, underline, strikeout and invert), with bookmarks and notes beside them, and a plugin exports them as text, Markdown, HTML, JSON or the Kindle clippings format, or sends them to services such as Joplin and Readwise. The guide also notes that calibre's KOReader Sync plugin can pull highlights into a calibre database.</p>\n<p>The typography settings take in your own fonts, font weight, kerning, word spacing, line spacing, margins, and CSS style tweaks. To get around a long book there's the Book Map, an overview of the whole book, and the Page Browser, which shows the pages “like in a photo album” as a grid of thumbnails.</p>\n<p>Taps, swipes and keys can be bound to more than 200 actions, and profiles bundle actions that can run automatically. Much of this is done by plugins, which the guide compares to browser add-ons. A reading statistics plugin records your sessions and your reading speed and uses them to estimate how long a chapter or a book has left.</p>\n<h2 id=\"syncing-between-devices\">Syncing between devices</h2>\n<p>A progress sync plugin keeps your place across KOReader devices. You make an account from inside the plugin while the device is online. The guide says the server keeps only an MD5 hash of your password, that the server software is open source, and that you can host your own.</p>\n<h2 id=\"where-it-comes-from\">Where it comes from</h2>\n<p>The guide names three official sources: koreader.rocks, the project's GitHub page, and its release on F-Droid for Android. It is blunt that other sites with the name aren't the project, koreader.com among them, and that “There is no Play Store version of KOReader!” The guide doesn't cover installing it on a Kindle or a Kobo; it starts once KOReader is running and points to the releases on GitHub. After that, KOReader can update itself over Wi-Fi (Update › Check for update in the top menu). Besides the stable releases, which usually come once a month, there are nightly builds, which the guide says are for testers and developers.</p>\n<p>One thing it won't do. On Kindles that show ads on the lock screen, the guide says that removing the ads is technically possible, but the project doesn't change the default sleep screen there: Amazon might react by locking down the Kindle, or the whole project could end.</p>\n<h2 id=\"koreader-and-a-library-on-a-mac\">KOReader and a library on a Mac</h2>\n<p>With OPDS, KOReader browses a library kept on a computer and downloads from it itself. Readwyn tested its own OPDS catalogue with KOReader in October 2026: KOReader browsed the library by author, series and shelf, searched it, and downloaded and opened an EPUB 3 (<a href=\"/blog/opds-catalogue-for-koreader/\">the setup and the test</a>). <a href=\"/blog/what-is-opds/\">What is OPDS?</a> explains the format itself.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-the-xteink-x4/","url":"https://readwyn.com/blog/what-is-the-xteink-x4/","title":"What is the Xteink X4?","summary":"The Xteink X4 is a 4.3 inch E Ink reader, about the size of a phone's back, that sticks on with magnets and turns pages with buttons. It reads EPUB from a microSD card.","content_html":"<p>The Xteink X4 is a very small E Ink e-reader made by Xteink, a Chinese hardware company. Its screen is 4.3 inches, it has buttons instead of a touchscreen, and its back is magnetic, so it clips onto an iPhone or a recent Pixel and rides around on the back of your phone. The first version cost $69 when it went on sale at the end of 2025. It reads EPUB and plain text files, copied onto its microSD card or sent over Wi-Fi. That original is no longer sold: since mid 2026 the name has meant the X4 Classic, which replaced it, and the X4 Pro, which adds a touchscreen and a front light.</p>\n<h2 id=\"the-original-x4\">The original X4</h2>\n<p>heise reported the device on <a rel=\"external\" href=\"https://heise.de/-11109221\">9 December 2025</a>, with deliveries from 24 December. These are heise's figures, with the image formats that <a rel=\"external\" href=\"https://www.notebookcheck.net/Xteink-X4-Compact-e-reader-attaches-to-iPhones-thanks-to-MagSafe-and-can-display-more-than-just-e-books.1177199.0.html\">Notebookcheck</a> adds:</p>\n<table><thead><tr><th></th><th>Xteink X4 (2025)</th></tr></thead><tbody>\n<tr><td>Screen</td><td>4.3 inch, black and white, 800 by 480 pixels, 220 ppi</td></tr>\n<tr><td>Size</td><td>114 by 69 by 5.9 mm</td></tr>\n<tr><td>Weight</td><td>74 g</td></tr>\n<tr><td>Controls</td><td>Buttons only, no touch, no light</td></tr>\n<tr><td>Storage</td><td>32 GB microSD card included</td></tr>\n<tr><td>Battery</td><td>650 mAh, up to 14 days at 1 to 3 hours of reading a day</td></tr>\n<tr><td>Files</td><td>EPUB and TXT, plus JPG and BMP images</td></tr>\n<tr><td>Charging</td><td>USB-C</td></tr>\n</tbody></table>\n<p>Inside is an ESP32 microcontroller rather than the processor of a phone or tablet. It runs Xteink's own software, which installs no apps. The magnets hold it to phones that have them, and for phones that don't, Xteink offers stick-on magnetic strips. Notebookcheck also noted that it can show boarding passes, membership cards and QR codes, which suits a screen that stays on your phone all day.</p>\n<p>At 220 ppi the text is less sharp than on the 300 ppi screens of a Kindle or a Kobo Libra 2. Digital Trends, which gave it an Editors' Choice award in its <a rel=\"external\" href=\"https://www.digitaltrends.com/phones/xteink-x4-review/\">June 2026 review</a>, spotted a few jagged edges on fine text and faint pixelation in stylised fonts, but found the resolution more than enough for a screen this small. Why that density matters for small type is in <a href=\"/blog/e-reader-pixels-per-inch/\">how many pixels per inch an e-reader needs</a>.</p>\n<h2 id=\"what-it-is-like-to-use\">What it is like to use</h2>\n<p>The same review lists the rough edges. The stock software offers a file browser, a list of recent books and a settings page. Text comes in two sizes, small and medium, with three choices of line spacing and no control of the margins. Justified text can't be turned off, illustrations in books often don't appear, and the buttons carry no labels, so the reviewer called learning them \"steep (read: maddening)\". There is no front light either, so in a dark room you need a lamp or a clip-on light.</p>\n<p>Getting books onto it takes one of two routes, both of which Digital Trends describes: copy EPUB files onto the microSD card from a computer, or type the reader's local IP address into a browser on a computer or phone and drop the files over the network. The card slot is recessed so deep that you need a fingernail or a SIM pin, and that reviewer preferred the wireless route. TechCrunch's reviewer of the later X4 Pro <a rel=\"external\" href=\"https://techcrunch.com/2026/07/21/the-xteink-x4-pro-could-be-the-tiny-e-reader-of-your-dreams/\">found the card easier</a>. That article is clear about what won't work: Kindle books can't be read on Xteink devices, and library ebooks, which come as protected EPUB files, can't be added. A plain EPUB without DRM is what it wants; <a href=\"/blog/where-to-buy-drm-free-ebooks/\">where to buy DRM-free ebooks</a> lists the shops that sell them.</p>\n<p>TechCrunch describes the X3 and X4 as having gone viral on social media, with a cult following. Owners who want more than the stock software can swap it for CrossPoint, an open-source firmware. By May 2026 Xteink had <a rel=\"external\" href=\"https://liliputing.com/?p=199704\">blocked such firmware</a> on X3 and X4 units sold through Chinese retailers such as Taobao, though it says the units sold on its own website are not affected. Digital Trends put the install at about 20 minutes, flashed from a web browser, and said it adds margin controls, more fonts, labels for the buttons on screen, and better wireless transfer, and that it plays well with calibre.</p>\n<h2 id=\"the-x4-classic-and-the-x4-pro\">The X4 Classic and the X4 Pro</h2>\n<p>Xteink has moved on from the first X4. In September 2026 Notebookcheck reported that the original was <a rel=\"external\" href=\"https://www.notebookcheck.net/1392365.html\">discontinued shortly after the X4 Pro appeared</a> in July, and that a replacement, the X4 Classic, was on sale at $79 with a 920 mAh battery and a weight of 68 g. Notebookcheck lists EPUB, PDF, MOBI and TXT for the Classic, and an accelerometer that turns the page when you flick the device.</p>\n<p>The X4 Pro is the one that answers the main complaints. TechCrunch's review gives it a 4.26 inch touchscreen and a front light whose colour can be warmed (how such a light works is in <a href=\"/blog/front-light-and-warm-light/\">front light and warm light</a>), in a body 5.95 mm thick that weighs 72 g. It charges from a magnetic pogo connector rather than USB-C, and pages turn with a tap at the side of the screen or with the side buttons.</p>\n<p>On 11 November 2026, <a rel=\"external\" href=\"https://www.xteink.com/\">Xteink's own shop</a> listed three readers, all at sale prices below their list prices:</p>\n<ul>\n<li>X4 Pro: $99 (list price $129), touchscreen and buttons, front light, 1,100 mAh, about 72 g.</li>\n<li>X4 Classic (also called the X4 V2): from $79 (list price $109), buttons, no light, 920 mAh, about 68 g.</li>\n<li>X3: $69 (list price $79), a smaller 3.7 inch screen, buttons, 650 mAh, about 58 g.</li>\n</ul>\n<p>All three have a magnetic back and 16 GB of storage that a microSD card can take to 256 GB. So if someone says \"Xteink X4\" today, they may mean the discontinued original, the Classic that replaced it, or the Pro. None of them has the page of a 6 or 7 inch e-reader (<a href=\"/blog/e-reader-screen-size/\">screen sizes compared</a>): the whole original X4 measured 114 by 69 mm, small enough, as heise put it, to sit on the back of a current smartphone.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/wide-margins-on-kobo-or-kindle/","url":"https://readwyn.com/blog/wide-margins-on-kobo-or-kindle/","title":"Why the margins of a book are so wide on a Kobo or a Kindle","summary":"The e-reader's margin is only the outer layer. A book's stylesheet can add side margins of its own inside it, and in ems they grow with the font.","content_html":"<p>If one book has wide margins and the rest of your library doesn't, the book is adding them. The margin setting of a Kobo or a Kindle draws the outer edge of the page, and the book's own stylesheet can put a second margin, or padding, on the text inside that edge. Turning the device's slider down to its lowest step shrinks only the first layer. If the book sets its margin in ems, the gap also grows every time you make the type bigger. If every book looks this way, start with the device setting.</p>\n<h2 id=\"the-device-setting\">The device setting</h2>\n<p>On a Kobo, tap the middle of the page to open the Reading Menu, tap the Text icon, and drag the slider beside Margins, as <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017639913-Adjust-font-size-and-change-the-font-style-on-your-Kobo-eReader\">Kobo's help page</a> describes. On a Kindle, tap the top of the screen, then the Aa icon, and open the Layout tab: Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">Kindle Colorsoft user guide</a> lists “Increase or decrease margin size” there, beside the choice of screen orientation. The guide adds that these display settings are device specific, so a second Kindle on the same account keeps margins of its own.</p>\n<h2 id=\"margins-the-book-brings-with-it\">Margins the book brings with it</h2>\n<p>An EPUB, and the Kindle formats made from one, are styled with CSS. In CSS a padding is space around a box's content and a margin is space outside its border. A paragraph's box sits inside the body's box. The <a rel=\"external\" href=\"https://www.w3.org/TR/CSS2/box.html#collapsing-margins\">CSS 2.1 specification</a> lets vertical margins merge into one, but says “Horizontal margins never collapse.” So a left margin on the body and another on each paragraph both stay, and the e-reader draws its own margin outside them both. The same stylesheet can also <a href=\"/blog/why-ebook-fonts-look-different/\">keep your font choice from taking effect</a>.</p>\n<p>Publishers are told not to do this. Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GH4DRT75GWWAGBTU\">text guidelines for reflowable books</a>, part of the Kindle Publishing Guidelines, say body text “must not have an imposed left/right margin or padding throughout the book.” The page tells publishers to set the side margins of normal body text to 0 so that readers keep “the full range of margin selection”. A book that ignores the rule takes some of that range away: even at the lowest step of the slider, its own margin is still there.</p>\n<h2 id=\"margins-that-grow-with-the-font\">Margins that grow with the font</h2>\n<p>The unit makes it worse. The <a rel=\"external\" href=\"https://www.w3.org/TR/css-values-4/#font-relative-lengths\">CSS Values specification</a> defines one em as the font size of the element that uses it. A margin of 3em is 60 pixels at a font size of 20 pixels, and 120 pixels on each side when you double the size. Large type and wide margins then squeeze the line from both ends at once, and a few words to the line is the result.</p>\n<p>A percentage behaves differently: <a rel=\"external\" href=\"https://www.w3.org/TR/CSS2/box.html#margin-properties\">CSS calculates it</a> from the width of the containing box, so it stays put when the font changes. That's why Amazon's guidelines ask for percentages where a block (a recipe list or a block quote, in its examples) does need to stand in from the text, so “that the margins do not grow too wide with large font sizes and impair reading”. If only the quotations and poems of a book run into a narrow column at a large size, an em margin on those blocks is the likely cause.</p>\n<h2 id=\"what-you-can-change\">What you can change</h2>\n<ol>\n<li>Set the device margin to its lowest step, then compare the book with one that looks right. If the gap is still wider, it's in the book.</li>\n<li>Try a smaller font size. If the margins shrink with it, they're set in ems.</li>\n<li>On a Kindle, turn the screen sideways from the same Layout tab. A wider screen gives a longer line.</li>\n<li>With a DRM-free file of your own, an EPUB editor lets you clear the side <code>margin</code> and <code>padding</code> of <code>body</code> and of the paragraph classes, as Amazon's page asks of publishers. Turn a quotation's em margin into a percentage instead of deleting it.</li>\n<li>If you converted the book with calibre, its <code>ebook-convert</code> adds a page margin of 5 pt on each side by default (<a rel=\"external\" href=\"https://manual.calibre-ebook.com/generated/en/ebook-convert.html\">manual</a>). A value below zero for <code>--margin-left</code> and <code>--margin-right</code> adds none and keeps the book's own.</li>\n</ol>\n<p>None of this applies to a PDF or a fixed-layout book. There the white border is part of the page itself, and the e-reader can only scale the whole page or let you zoom. <a href=\"/blog/fixed-layout-vs-reflowable/\">Fixed layout and reflowable ebooks</a> explains the difference, and <a href=\"/blog/send-pdf-to-kindle/\">sending a PDF to a Kindle</a> covers turning the screen and zooming into a page with big margins. Wide margins also set how long each line is, and <a href=\"/blog/ideal-line-length-for-reading/\">the best line length for reading</a> gives the numbers to aim for.</p>\n","date_published":"2026-11-13T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/change-title-of-sideloaded-kindle-book/","url":"https://readwyn.com/blog/change-title-of-sideloaded-kindle-book/","title":"How to change the title of a sideloaded book on a Kindle","summary":"A Kindle lists a sideloaded MOBI book by the title inside the file. Change the title in Readwyn’s “Edit details”, then send the book to the Kindle again.","content_html":"<p>The title that a Kindle lists for a sideloaded MOBI book is written inside the book file, so renaming the file in Finder doesn't change it. The title has to change in the file, and the file has to be copied again. In Readwyn, correct the title in “Edit details”, then choose “Send to device again” with the Kindle plugged in. The copy that lands on the Kindle carries the new title. This works for a Kindle that the Mac mounts as a disk, which is the only way Readwyn reads one. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page on USB transfers</a> says a Kindle Scribe and the Kindles released in 2024 need a separate application on a Mac, and that models released earlier than 2024 (the Scribe aside) don't. Readwyn works with the earlier ones.</p>\n<h2 id=\"change-it-and-send-the-book-again\">Change it and send the book again</h2>\n<ol>\n<li>Open the book in Readwyn and click “Edit details”, or press ⌘E.</li>\n<li>Type the new title. The field shows “Was” with the old title struck through, and the button counts the change: “Save 1 change”. If the Kindle is connected and the book is on it, the footer adds “Your Kindle keeps the old details until you send the book again.”</li>\n<li>Save. Then click “…” (“More actions”) on the book screen and choose “Send to device again”.</li>\n<li>When Readwyn says “Copied to the device. Eject the device so it adds the book.”, click “Eject device” on the Sync screen and unplug the Kindle (<a href=\"/blog/eject-kindle-or-kobo-from-mac/\">more on ejecting</a>).</li>\n</ol>\n<p>Readwyn replaces the file in place. It can't make the Kindle read the file again. If the Kindle still lists the old title after you unplug it, take the book off the Kindle and send it fresh, by the steps under “The file name keeps the old title”, which also say what that costs.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book.webp?h=6c9b53f864b64d2872b9\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book.webp?h=1970bca0dae546ce847c\" alt=\"The book screen of Middlemarch, with a Kobo Clara 2E connected and the line On Kobo Clara 2E under the cover: the title and author, and the buttons Continue reading, Status: Reading, and Edit details\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>“Edit details” is on the book screen, and “Send to device again” is in the “…” menu beside it</figcaption>\n</figure>\n<p>The other fields of the sheet are covered in <a href=\"/blog/edit-ebook-metadata-on-mac/\">how to edit ebook metadata on a Mac</a>. The author works the same way as the title: the copy on the Kindle gets the authors that the library has when you send.</p>\n<h2 id=\"why-renaming-the-file-isn-t-enough\">Why renaming the file isn't enough</h2>\n<p>A MOBI file has a name on the disk and a title inside, and the two are independent. The MobileRead wiki's <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Mobi2Mobi_Advanced\">notes on Mobi2Mobi</a>, a tool that edits this metadata from the command line, make the distinction: you can rename a <code>.mobi</code> file in Windows Explorer or in Finder, but “you can only change the books Metadata with Mobi2Mobi”. The same page uses the inner title for a trick. It puts a short genre code at the front of each title, so that the Kindle's sort by title groups the books by genre.</p>\n<p>Readwyn's copy holds the title in three places. Two of them, the full name and the record that the <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/MOBI\">wiki's page on the MOBI format</a> lists as 503, “updatedtitle”, take the whole title exactly as you typed it, in any script. The third is the name of the Palm database, which the format limits to 31 bytes. Readwyn fills it with the ASCII letters and digits of the title and puts <code>_</code> for everything else. A title with no such letter or digit, a Tamil one for example, gets the name <code>book</code> there.</p>\n<p>None of this touches the EPUB in your library. Readwyn keeps a book's details in its own library and never writes them into your book files, so “Send to device” takes the title from the library at the moment you send. The text of the copy comes from the documents of the EPUB, character for character, and the title you typed isn't added to it. A title page inside the book still shows the title as the EPUB prints it.</p>\n<h2 id=\"the-file-name-keeps-the-old-title\">The file name keeps the old title</h2>\n<p>The first time you sent the book, Readwyn wrote it to <code>documents/&lt;title&gt; - &lt;first author&gt;.mobi</code> and recorded that path. A later send goes to the recorded path, so the new copy replaces the old one and the Kindle doesn't end up with the book twice. The file is still named after the old title, though, which shows when you browse the Kindle in Finder.</p>\n<p>If you want the file renamed too, take the book off the Kindle and send it fresh:</p>\n<ol>\n<li>With the Kindle connected, open the book's Details tab. In the row “On the device”, click “Remove from device”. Readwyn deletes the file and forgets its path.</li>\n<li>Choose “Send to device”. The book goes to <code>documents/&lt;new title&gt; - &lt;first author&gt;.mobi</code>. A character that the Kindle's disk refuses in a file name, a colon for example, becomes <code>_</code>.</li>\n</ol>\n<p>This route costs something. “Remove from device” also deletes the files that the Kindle kept beside the book: its <code>.mbp</code> file on an older Kindle, or its <code>.sdr</code> folder on a newer one. They hold the Kindle's place in the book and the highlights and notes you made there. So run “Device → Readwyn” on the Sync screen first, which brings <a href=\"/blog/kindle-highlights-in-ebook-library/\">progress, highlights and notes into the library</a>. They stay in the library, but they don't all go back. Readwyn sends no highlights or notes to any Kindle. Progress returns only to a Kindle on firmware 1 to 4, and only after that Kindle has opened the new copy once (<a href=\"/blog/sync-kindle-with-mac-without-calibre/\">what a Kindle sync carries</a>).</p>\n<h2 id=\"a-book-that-you-copied-by-hand\">A book that you copied by hand</h2>\n<p>Readwyn doesn't patch a book file on the Kindle. “Send to device” always writes a whole copy, made from the EPUB in the library. A book that you dragged to the Kindle yourself therefore has to be in the library first: <a href=\"/blog/add-books-by-dragging/\">add the file</a>, set its title, and send it. A MOBI or an AZW3 gets an EPUB 3 at import. A file with DRM doesn't convert, and without that EPUB the send stops with “Not copied: the book has no EPUB file”.</p>\n<p>If you highlighted in the book on the Kindle, run “Device → Readwyn” on the Sync screen after you add the file and before you change the title. Readwyn finds a book that it didn't send by the title inside the file, so the sync reaches the book only while the two titles still agree. The same sync records where the file is, and the send that follows replaces it in place.</p>\n<p>Afterwards, look in the Kindle's <code>documents</code> folder in Finder. If it holds two files for the book, the old one and a new one named after the title you set, the Kindle has the book twice, once under each title. Delete the one that you copied by hand. The Kindle kept its place in the book and the highlights you made there beside that file, in its <code>.mbp</code> file or its <code>.sdr</code> folder, so the new copy starts without them.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/does-kindle-sync-sideloaded-books/","url":"https://readwyn.com/blog/does-kindle-sync-sideloaded-books/","title":"Does a Kindle sync the progress of sideloaded books?","summary":"A book copied to a Kindle over USB keeps your place on that Kindle only. A file sent through Send to Kindle syncs its last page read, notes and highlights.","content_html":"<p>Not if you copied the book over a USB cable. Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">user guide for the Kindle Colorsoft</a> says files transferred by cable \"are not stored in the Cloud\", and it offers syncing of notes, highlights, bookmarks and the last page read only for personal documents \"archived in Kindle format\": the ones that came through Send to Kindle with archiving switched on. The Kindle still records your place in a book that came over the cable, in a file beside the book, but that record stays on the one Kindle. Readwyn, the ebook library app for Mac and Linux that this blog belongs to, can read it, but only from a Kindle that the computer mounts as a drive. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">help page on USB transfers</a> says the Kindle Scribe and the Kindles released in 2024 need a separate application for a cable transfer on a Mac, and that models released before 2024 (the Scribe excepted) don't. Those earlier models are the ones Readwyn works with.</p>\n<h2 id=\"two-routes-onto-a-kindle\">Two routes onto a Kindle</h2>\n<p>A book that didn't come from the Kindle Store reaches a Kindle by one of two routes, and Amazon treats them differently.</p>\n<p>With a cable, you drag a file into the Kindle's documents folder, and Amazon's cloud keeps no copy of it. The Colorsoft guide gives the consequences for the file itself. Remove it from the Kindle and it's permanently deleted. Buy a new Kindle and anything \"transferred directly to your previous Kindle\" has to be carried across by hand.</p>\n<p>With Send to Kindle, the file goes to Amazon first: by email, from the web page, from the Kindle app on a phone, or from the desktop apps. Amazon keeps it in your library as a personal document, and you read it in Kindle format. Of these documents the guide says that notes, highlights and bookmarks \"are synchronized across devices along with the last page you read\". An EPUB qualifies, because <a href=\"/blog/send-epub-to-kindle/\">Send to Kindle has accepted EPUB files since 2022</a>.</p>\n<p>The split is an old one. In October 2011, software version 3.3 let the Kindle Keyboard download archived personal documents and, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2011/10/14/kindle-keyboard-software-update-version-3-3-adds-new-features/\">The eBook Reader reported</a>, brought Whispersync of their last page read, bookmarks and annotations, PDFs excepted. A year later Christopher Breen described the Send to Kindle app <a rel=\"external\" href=\"https://www.macworld.com/article/219438/syncing-all-books-to-kindle-apps-and-devices.html\">in Macworld</a> and wrote of the books it delivers: \"Unlike with sideloaded books, these will stay in sync.\"</p>\n<h2 id=\"what-a-sent-document-needs-before-it-syncs\">What a sent document needs before it syncs</h2>\n<ul>\n<li><strong>Archiving.</strong> In your Amazon account, open Manage Your Content and Devices, the Preferences tab, Personal Document Settings, then Edit Archive Settings. The box \"Enable personal document archiving to my Kindle library\" has to be ticked. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G6N7P2TFA2YTH5YW\">page on turning archiving off</a> gives that path. If you send from the Send to Kindle web page, <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G5WYD9SAF7PGXRNA\">Amazon's overview of the service</a> says to turn on \"Add to your library\" there, so that the document can be downloaded to any device linked to your account.</li>\n<li><strong>Sync.</strong> It's on by default and needs an internet connection. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GGFEXXS8Z7DPJSTN\">Amazon's page on the setting</a> puts it under More, Settings, Device Options on a Kindle e-reader, and says it saves your reading position, notes and highlights to Amazon's servers. The Colorsoft guide calls the setting Whispersync for Books and puts it under All Settings (in the Quick Actions menu), Device Options, Whispersync.</li>\n<li><strong>Kindle format.</strong> The guide limits syncing to documents archived in Kindle format. It adds that PDF and TXT files can also be read natively, and its promise of syncing doesn't cover a file kept that way.</li>\n</ul>\n<p>One gap remains even with all three in place. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TNtSm13k3txFI4EvDV\">table of the notes that each device and app supports in a sent document</a> has a row for Kindle for Web that reads \"personal documents aren't supported yet\", so a sent document syncs between Kindle e-readers and the Kindle apps but can't be opened at read.amazon.com.</p>\n<h2 id=\"the-place-that-stays-on-the-kindle\">The place that stays on the Kindle</h2>\n<p>A Kindle with firmware 5 or later (the Kindle Touch, the Paperwhite, the Voyage and the Oasis among them) keeps the position of a sideloaded book in a folder beside the book, named after it and ending in .sdr. That's the record Readwyn reads. With the Kindle connected, the direction “Device → Readwyn” on <a href=\"/blog/sync-kindle-with-mac-without-calibre/\">the Sync screen</a> brings the progress of each book that's also in your library onto the computer, for books in MOBI and AZW3 (KF8) format. The figure is a whole percent: the last-read position divided by the length of the book's text. Nothing travels the other way: Readwyn writes no progress to a Kindle of that generation.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/eject-kindle-or-kobo-from-mac/","url":"https://readwyn.com/blog/eject-kindle-or-kobo-from-mac/","title":"How to eject a Kindle or a Kobo from a Mac","summary":"Eject it in Finder with File > Eject, the Eject button in the sidebar, or a drag to the Trash. In Readwyn, “Eject device” on the Sync screen waits for a write to end.","content_html":"<p>A Kindle or a Kobo that your Mac shows as a drive is ejected like any other storage device. <a rel=\"external\" href=\"https://support.apple.com/guide/mac-help/connect-storage-devices-mac-mchl027f1d66/mac\">Apple’s guide</a> lists three ways: select it and choose File &gt; Eject, click the Eject button beside its name in the Finder sidebar, or drag its icon from the desktop to the Trash. Then unplug the cable. In Readwyn, “Eject device” at the top of the Sync screen ejects it from inside the app, and it waits if Readwyn is still writing to the e-reader. Readwyn works with a Kindle only when the Mac mounts it as a drive, and the Mac doesn’t mount every Kindle: <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon’s help page on USB transfers</a> says the Kindle Scribe and the Kindles released in 2024 need a separate application to transfer files over USB on a Mac, such as Amazon’s own Send to Kindle app. For one of those Kindles there’s nothing to eject in Finder: the same page says to select Disconnect on the Kindle’s screen or to unplug the cable.</p>\n<h2 id=\"in-finder\">In Finder</h2>\n<p>A Kobo joins the Mac only after you tap Connect on its screen, and its drive is called KOBOeReader. <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017765933-Eject-your-Kobo-eReader-when-connected-to-a-computer\">Kobo’s steps for a Mac</a> are to open a Finder window, right-click KOBOeReader in the sidebar, click “Eject KOBOeReader”, and unplug. The same article gives the reason in four words: you eject “to avoid losing data”.</p>\n<p>For a Kindle, Amazon has said two things over the years. The <a rel=\"external\" href=\"https://kindle.s3.amazonaws.com/Kindle_Paperwhite_Users_Guide.pdf\">user’s guide of the first Kindle Paperwhite</a>, from 2012, tells Mac owners to click the Eject button next to the Kindle in any Finder window, or to drag it from the desktop to the Trash. The Kindle then leaves what Amazon calls USB drive mode and carries on charging. Until it does, you can’t read on it: the Kindle, says the guide, “is not usable as a reading device while in USB drive mode”. Amazon’s <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/Ereader/Early+Generation+Device+Help/Kindle+1st+Gen+Help/enUS/Kindle_1st_Generation_Use_with_Your_Computer.html\">page for the first Kindle</a> puts the eject before the unplug: unmount or eject the Kindle with the computer, then disconnect the cable. Amazon’s current help page, the one linked above, ends with a note that there’s no need to select an eject button on the computer at all. The only Mac steps on that page are those of the Send to Kindle app, which shows the Kindle’s files in a window of its own.</p>\n<p>The same page says that a Kindle released before 2024, other than the Scribe, needs no separate application. If yours sits in Finder as a drive with a <code>documents</code> folder, it’s in the USB drive mode that the older guides describe, so eject it first and unplug it second. When it should be there and isn’t, <a href=\"/blog/kindle-not-showing-up-on-mac/\">check the cable, a macOS alert, and a Finder setting</a>. If you don’t know which model you have, <a href=\"/blog/which-kindle-do-i-have/\">its serial number or its Device Info screen will tell you</a>.</p>\n<h2 id=\"in-readwyn\">In Readwyn</h2>\n<p>The Sync screen has the button for a Kobo and for a Kindle alike.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-sync.webp?h=3748d9769c62e82b8755\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-sync.webp?h=25b1a7e75f31a1231622\" alt=\"The Sync screen of a Kobo Clara 2E, connected over USB, with the button Eject device at the top right\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>“Eject device”, at the top right of the Sync screen</figcaption>\n</figure>\n<p>When the eject is done, a message names the e-reader: “Ejected Kobo Clara 2E. You can unplug it.” Readwyn never reads a Kindle’s serial number, so it doesn’t know the model, and for a Kindle the message uses the name of the drive, for example “Ejected Kindle. You can unplug it.”</p>\n<p>The button knows when Readwyn itself is in the middle of a write. That can be a book going over with “Send to device”, the changes of a Readwyn → Device sync, a “Remove from device”, or the small file <code>.readwyn/device-id</code> that the first sync or send leaves on the e-reader. Click “Eject device” during one of these and an alert says “Readwyn is still writing to the Kindle” (or “the Kobo”), with the line “Wait for the sync to finish before you unplug it, or the device may lose its highlights.” “Wait” closes the alert and ejects nothing. “Stop and eject” ejects, but the eject still lets the write end first. If the write ends while the alert is open, the alert closes by itself.</p>\n<p>That wait lasts 30 seconds at most. If the write is still going then, nothing is ejected and Readwyn says “Readwyn is still writing to the device. Try again when the sync ends.” In a test on a copy of a real Kindle’s disk, with a write held open for 2 seconds, the eject waited 2.0 seconds and then took the volume away.</p>\n<h2 id=\"when-the-mac-refuses\">When the Mac refuses</h2>\n<p>Finder sometimes answers that the disk can’t be ejected. Apple’s guide says that another app or user may be using a file on it. Kobo’s article on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017605374-Problems-ejecting-your-Kobo-eReader\">problems ejecting</a> is more specific: a program is reading the e-reader, an anti-virus program is scanning it, a file manager window is showing its contents, a program has a file on it open, or the operating system is indexing it. Its advice is to wait and try again, then to close every window and app that might be using the e-reader. The last resort is the same at Kobo and at <a rel=\"external\" href=\"https://support.apple.com/guide/mac-help/eject-a-disk-mac-solutions-mchlp1285/mac\">Apple</a>: shut the computer down, then unplug. Apple puts one step before it, which is to log out, log in again, and try the eject once more.</p>\n<h2 id=\"what-the-e-reader-does-next\">What the e-reader does next</h2>\n<p>A Kobo takes in new books only once it has left USB mode. Kobo’s guide to <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">adding files from a computer</a> says you’ll see an “Importing content” message after you unplug the cable. That’s why Readwyn’s message after <a href=\"/blog/send-books-to-kobo-from-mac/\">a send</a> reads “Copied to the device. Eject the device so it adds the book.” Apply a Readwyn → Device sync straight after a send, before the Kobo has had that chance, and Readwyn stops with “The Kobo is still adding books. Eject it, let it finish, and connect it again.” That refusal comes from one of <a href=\"/blog/how-readwyn-writes-to-kobo-safely/\">the rules Readwyn keeps when it writes to a Kobo</a>.</p>\n<p>A classic Kindle, one with firmware 1 to 4, needs one more step after a sync that wrote collections to it, which is still an experimental feature. Readwyn’s message then ends with “Restart the Kindle to see the collections.” So eject it and unplug it. Then, on the Kindle, press Home, then Menu, choose Settings, press Menu again, and choose Restart. The power button doesn’t load the collections.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/export-kindle-scribe-notes/","url":"https://readwyn.com/blog/export-kindle-scribe-notes/","title":"How to export your notes from a Kindle Scribe","summary":"A Kindle Scribe exports notebooks from its Share icon: by email on every Scribe, and to Google Drive, OneDrive or OneNote on a Scribe from 2025 or later.","content_html":"<p>Open the notebook on the Scribe, tap the top of the screen to bring up the toolbar, and tap the Share icon in the top right corner. Amazon's help page on <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TJE2UYmdw0ppUuR3Rs\">sharing notes and notebooks</a> gives two choices at that point: “Quick send to” the email address of your Amazon account, or “Share via email” to anyone else. The email that arrives holds a download link, not the notebook itself, and the link stops working after 7 days. A Scribe released in 2025 or later can also send the notebook to Google Drive, Microsoft OneDrive or Microsoft OneNote from the same menu.</p>\n<h2 id=\"notes-in-a-book-or-a-document\">Notes in a book or a document</h2>\n<p>The same Share icon works inside a book or a personal document, with one extra tap: in a book, you tap the notebook icon first. What comes out is not the page as you see it. Amazon says your handwritten or text sticky notes appear at the end of the shared document, while on-page mark-ups and highlights stay in the original. Highlights you make in ordinary Kindle books follow other rules; <a href=\"/blog/export-kindle-highlights/\">where Kindle highlights are kept</a> covers those. If you wrote in the margins of a PDF and want those strokes on the pages, a Scribe from 2025 or later can send the annotated PDF to a cloud drive, as long as the PDF didn't come onto the Scribe over USB.</p>\n<p>The help page also lists two limits. You can share only one notebook at a time, and you can't share notebooks or personal documents with members of your Amazon Family, even if they own a Scribe too.</p>\n<h2 id=\"handwriting-as-text\">Handwriting as text</h2>\n<p>The Share menu has two more options that run handwriting recognition first. Amazon's page on <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TxYYWu3ZLZPBSTgp9H\">converting handwritten notes to text</a> describes them:</p>\n<ul>\n<li>“Convert to text and quick send” mails the text to your own address.</li>\n<li>“Convert to text and email” opens a preview first, where you can tap each page to correct what the Scribe read.</li>\n</ul>\n<p>Either way the text arrives as a .txt file. The corrections you make in the preview go into that email only and aren't saved on the Scribe. The sharing page adds that both options can also give you a PDF of the notebook in which your handwriting is searchable. Quick send includes it as a second link; with the email option you tick “Attach notebook as a searchable PDF”. That PDF leaves out the corrections you made in the preview.</p>\n<p>You set the recognition language under Settings, Device Options, Language &amp; Dictionaries, Handwriting Recognition, and the setting applies to every conversion. The list runs to 20 languages and variants, from Afrikaans and Arabic to Spanish and Swedish, with both British and American English.</p>\n<h2 id=\"google-drive-onedrive-and-onenote\">Google Drive, OneDrive and OneNote</h2>\n<p>Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TP8wuaQilIlwlITVTL\">page on the cloud services</a> says you need a Scribe released in 2025 or later, registered and online, and a Google or Microsoft account. If you aren't sure which Scribe you have, <a href=\"/blog/which-kindle-do-i-have/\">the model check</a> comes first. In a notebook, the Share icon then offers “Share current page” or “Share entire notebook” (to send only some pages, select them first with the multipage icon), and then a format:</p>\n<table><thead><tr><th>Service</th><th>What it receives</th></tr></thead><tbody>\n<tr><td>Google Drive or OneDrive</td><td>The handwriting as a PDF, the text, or a searchable PDF</td></tr>\n<tr><td>OneNote</td><td>The handwriting as a PNG image, the text, or both side by side</td></tr>\n</tbody></table>\n<p>Nothing syncs on its own. The <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TIh9JOMGAKr7bY4zqu\">cloud drive FAQ</a> says you export each time by hand, and that a document goes to the drive as a PDF with your annotations in it. The files land in a folder called Kindle Scribe (My Drive on Google Drive, My Files on OneDrive), and a second Scribe gets a folder of its own. A file you put on the Scribe over USB can't be exported to a drive at all. The cable isn't a documented route for notebooks either: Amazon's page on <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">USB transfers</a> doesn't mention notebooks, and says a Mac needs Amazon's Send to Kindle app to move files to and from a Scribe.</p>\n<p>OneNote works a little differently. Its <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=T4v7rH2oZuFFKot0gs\">FAQ</a> says exports go to a OneNote notebook called “Kindle Scribe”, with each Scribe notebook as a section of it, and that exporting again adds a new page with a timestamp instead of replacing the old one. OneNote takes notebooks only: <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TP8wuaQilIlwlITVTL\">the cloud services page</a> says personal documents can't go to it.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-many-books-read-on-kindle/","url":"https://readwyn.com/blog/how-many-books-read-on-kindle/","title":"How to see how many books you have read on a Kindle","summary":"Open Reading Insights in the Kindle app (More, then Reading Insights) or on Amazon's website for your books read by year. The Kindle itself only has a Read filter.","content_html":"<p>The count lives on Reading Insights, Amazon's page of your Kindle reading statistics, and not on the Kindle. In the Kindle app for iPhone or Android, tap More and then Reading Insights, <a rel=\"external\" href=\"https://goodereader.com/blog/kindle/kindle-reading-insights-is-an-excellent-system-that-monitors-reading-habits\">which sits under your name</a>. In a browser, sign in to Amazon and open <a rel=\"external\" href=\"https://www.amazon.com/kindle/reading/insights\">amazon.com/kindle/reading/insights</a>. Nathan Groezinger of The eBook Reader <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2025/01/06/free-prime-ebooks-reading-insights-and-the-neverending-kindle-challenge/\">describes the page</a> as showing how many books you've read by year, the days you read in each month, and your daily and weekly streaks. On the e-reader itself, the nearest thing is a filter that lists the books you've finished, with no total and no year.</p>\n<h2 id=\"what-the-kindle-shows\">What the Kindle shows</h2>\n<p><a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">Amazon's user guide for the Kindle Colorsoft</a> says that books you finish are marked as read automatically. To see them, open your library, tap the filter icon at the top left, and tap Read. Unread does the opposite. If the Downloaded filter is on, the list holds only what is on the device; the guide says to deselect it to see your Cloud content as well.</p>\n<p>That list is all the e-reader gives. A <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2025/05/19/ebook-readers-lack-useful-reading-history-and-reading-stats/\">May 2025 piece on The eBook Reader</a> puts it plainly: Kindle e-readers mark a book as read when you finish it, and show no reading stats at all, while the Kindle apps show the same insights as Amazon's website. You'd have to count the list by hand, and the guide's sort options (Most recent, Title, Author, Publication date) include no date of finishing, so you can't pick out one year.</p>\n<h2 id=\"kindle-challenges\">Kindle Challenges</h2>\n<p>Amazon also runs Kindle Challenges, a set of achievements with <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2025/01/06/free-prime-ebooks-reading-insights-and-the-neverending-kindle-challenge/\">a new challenge about every three months</a>. Reading on a Kindle e-reader counts towards them, but in <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2026/01/05/amazon-still-treating-kindles-like-second-class-devices/\">January 2026</a> you could still view the challenges and the achievements you'd earned only in the Kindle app for iOS and Android. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TdP7Fpp3mrAY4gibL9\">Amazon's help page</a> says that Kindle books bought or borrowed from Amazon count towards books read and streaks when you read them on a Kindle, in the app, or in Kindle for Web. Amazon doesn't say whether Reading Insights uses the same rule.</p>\n<h2 id=\"a-second-count-through-goodreads\">A second count through Goodreads</h2>\n<p>If your Kindle is linked to a Goodreads account, Goodreads keeps its own list. The Colorsoft guide says that rating a book on the Kindle's Goodreads Shelves screen adds it to your Read shelf. Goodreads' <a rel=\"external\" href=\"https://www.goodreads.com/blog/show/3064\">Year in Books</a> then counts the books shelved with a read date in that year, and its December 2025 post says that books you've chosen to keep private on Kindle are left out. The stats keep updating until 31 December, and the summary email goes only to members with at least three books read in the year.</p>\n<p>The two counts need not agree: Goodreads needs a read date and leaves out private books, and Amazon's books read and streaks count only Kindle books that you bought or borrowed from Amazon. What a Kindle sends back to Amazon in the first place is covered in <a href=\"/blog/does-your-e-reader-track-what-you-read/\">does your e-reader track what you read?</a>, and the <a href=\"/blog/how-many-books-does-the-average-person-read/\">average reader's yearly total</a> gives you something to measure against.</p>\n<h2 id=\"books-you-read-outside-the-kindle\">Books you read outside the Kindle</h2>\n<p>Readwyn, the ebook library app for Mac and Linux that this blog belongs to, has a “Books finished” figure on its Stats screen for a week, a month, a year, or all time. A book finished twice counts twice, and All time gives the year that your reading data starts (“Since 2009”). The finished reads come from the books you set to Read in Readwyn, from a Kobo sync, and from a calibre import, and once a year they fill the first card of <a href=\"/blog/year-in-books/\">your year in books</a>.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kindle-highlights-in-ebook-library/","url":"https://readwyn.com/blog/kindle-highlights-in-ebook-library/","title":"Get your Kindle highlights and notes into your ebook library","summary":"Plug in a Kindle that mounts as a drive, and a Device → Readwyn sync copies its highlights and notes into your library. Each book then exports as a Markdown file.","content_html":"<p>Connect the Kindle to your Mac with its cable, open “Sync” in Readwyn's sidebar, and review and apply the “Device → Readwyn” direction. It copies the highlights and notes of the books that are both on the Kindle and in your library, and from then on they sit in Readwyn's Notebook, by chapter, each with its page. This needs a Kindle that the Mac mounts as a drive, because Readwyn reads the files on the Kindle and has no other way in. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page on USB transfers</a> says the Kindle Scribe and the Kindles released in 2024 need a separate app on a Mac, such as Amazon's own Send to Kindle, to move files over a cable, and that models released earlier than 2024, the Scribe excepted, don't. A Kindle that needs such an app <a href=\"/blog/kindle-not-showing-up-on-mac/\">doesn't mount as a drive</a>, so it's out of Readwyn's reach.</p>\n<h2 id=\"where-the-highlights-come-from\">Where the highlights come from</h2>\n<p>A Kindle has no database of highlights. It keeps them in a file beside each book, and Readwyn reads that file: the <code>.mbp</code> file of a Kindle with firmware 1 to 4, or the files in the book's <code>.sdr</code> folder on firmware 5 and later (<a href=\"/blog/sync-kindle-with-mac-without-calibre/\">syncing a Kindle with a Mac</a> lists the models of each). For a MOBI or an AZW3 book these files record the exact position of each highlight, so Readwyn can find the same passage in the EPUB of your library and store its chapter, its page, and the paragraphs around it. A book that Readwyn itself put on the Kindle with “Send to device” lines up exactly, since the MOBI it sends has the same text as the library's EPUB. One case fails. Readwyn can't read the text of a MOBI or AZW3 file with DRM, and on firmware 5 and later the Kindle's file gives only where a highlight is, not its words. On those Kindles, the highlights and notes of a book with DRM don't come in.</p>\n<p>Books in Amazon's KFX format are different, because Readwyn can't read a KFX file. For those it falls back on <code>documents/My Clippings.txt</code>, the running list that a Kindle adds every highlight and note to (the <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">Kindle Colorsoft user guide</a> calls it My Clippings). The file gives the text of a highlight but only a Kindle location for its place, so Readwyn searches for the words in your EPUB. That route has a few quirks:</p>\n<ul>\n<li>A highlight you deleted on the Kindle is still in the file, so it comes into the library anyway.</li>\n<li>A Kindle can cut a clipping short. It writes “&lt;You have reached the clipping limit for this item&gt;” where the rest of the text should be, and Readwyn keeps the text before that mark and skips a highlight that has none. Amazon's own export has a limit of this kind: <a rel=\"external\" href=\"https://docs.readwise.io/readwise/docs/importing-highlights/kindle\">Readwise</a>, a service that imports Kindle highlights, says publishers require Amazon to limit how much highlighted text can be exported from a book.</li>\n<li>A note with no highlight under it is left out, as there's no text to look for.</li>\n<li>A highlight in a PDF has a page but no location, and Readwyn doesn't read it.</li>\n</ul>\n<p>Two things hold for every Kindle. Bookmarks aren't read. And a Kindle highlight carries no colour that Readwyn can read, so each new one arrives yellow. Give it another colour in Readwyn and later syncs leave your choice alone.</p>\n<h2 id=\"reviewing-the-sync\">Reviewing the sync</h2>\n<p>On the Sync screen, the “Device → Readwyn” card says “Bring progress, highlights and notes from the Kindle into the library.” A Kindle with firmware 5 or later adds My Words to that list. Below the sentence are a count of the changes and a button to review them.</p>\n<p>Nothing is copied until you've looked at the list. The review sheet groups the changes by book: a new highlight is a row marked “Add · Highlight” with its text, and one that has a note is “Add · Note”. Untick a row and it stays behind. The button at the foot counts the ticked rows, for example “Apply 7 changes”, and a click on it does the copy. At the top is “Merge highlights, notes and words instead of replacing them”, which is on by default. Leave it on, and the sync deletes no highlight. Turn it off, and the library's highlights of each synced book become exactly the Kindle's, so anything you <a href=\"/blog/highlight-and-take-notes-in-epub-on-mac/\">highlighted in the reader on the Mac</a> in those books would go. The review lists each deletion before you apply the sync.</p>\n<p>A book that is on the Kindle but not in your library appears under “Left out” with the reason “Not in this library.” Add it, and the next review has its highlights. For a KFX book that means an EPUB of the same title, since the EPUB is where Readwyn looks for the words. A book that you copied to the Kindle without Readwyn is paired with the library book of the same title.</p>\n<p>Highlights and notes travel one way only. Under “Not available on this Kindle”, the Sync screen lists “Highlights and notes to the device” with the reason: no public description tells how to write a highlight into a Kindle's files, and on firmware 1 to 4 a wrong record can cost the Kindle its own notes. So a note you add or edit on the Mac doesn't reach the Kindle. The Kindle still has its own version, and each later “Device → Readwyn” review lists the difference: “Update · Note” if the Kindle has another note for that highlight, “Delete · Note” if it has none. The merge option doesn't cover these rows. Untick the row, if it's ticked, to keep the text you wrote on the Mac. The first sync you apply does write one small file to the Kindle, <code>.readwyn/device-id</code>, a random ID by which Readwyn knows the device next time.</p>\n<h2 id=\"in-the-notebook\">In the Notebook</h2>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-notebook.webp?h=c5d907fb6a66e4603293\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-notebook.webp?h=758885220be4b17df8b7\" alt=\"The Notebook: the books with highlights at the left, and the highlights of Middlemarch at the right, grouped by chapter, with their colours, pages, and a note\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The Notebook, with Copy all and Export over the highlights of a book</figcaption>\n</figure>\n<p>“Notebook” in the sidebar lists every book with highlights, and a click on one shows its highlights in reading order under their chapter titles. The page at the right is the printed page if the book has a page list, and an estimate such as “~p. 196” if it doesn't. A clipping whose words aren't in your EPUB, from another edition for instance, has no page and no context. It goes to the end of the list under “Not found in this copy”. <a href=\"/blog/all-highlights-in-one-notebook/\">All your highlights in one notebook</a> goes through the rest of the screen.</p>\n<h2 id=\"a-markdown-file-for-obsidian-or-notion\">A Markdown file for Obsidian or Notion</h2>\n<p>Above a book's highlights, “Export” offers “Markdown…” and “Plain text…”. Either opens the macOS save panel in Documents with a name filled in, such as “Middlemarch - Highlights.md”. For the book in the picture, the Markdown file begins:</p>\n<pre><code data-lang=\"markdown\"># Middlemarch\n\nGeorge Eliot\n\n## Prelude\n\n&gt; Here and there is born a Saint Theresa, foundress of nothing, whose loving heart-beats and sobs after an unattained goodness tremble off and are dispersed among hindrances, instead of centring in some long-recognizable deed.\n&gt;\n&gt; — p. 3\n\nThe whole book in one sentence.\n\n## Book I · Chapter I\n\n&gt; Miss Brooke had that kind of beauty which seems to be thrown into relief by poor dress.\n&gt;\n&gt; — p. 7\n</code></pre>\n<p>The title is a first-level heading with the authors under it, each chapter is a second-level heading, each highlight is a block quote that ends with its page, and a note is an ordinary paragraph below its quote.</p>\n<p>Obsidian needs no import step. Its help says it keeps notes as <a rel=\"external\" href=\"https://obsidian.md/help/data-storage\">Markdown-formatted plain text files</a> in a vault, that a vault is a folder on your disk, and that it refreshes the vault by itself when something changes there from outside. Choose a folder inside your vault in the save panel, and the book's highlights turn up as a note. Notion takes the same file through an import: its <a rel=\"external\" href=\"https://www.notion.com/help/import-data-into-notion\">help page</a> gives the path as Settings, then Import, then “Text &amp; Markdown”, where you upload the <code>.md</code> file.</p>\n<p>The export covers one book at a time, and the file holds what the library had when you saved it. After a later sync brings in more highlights, export the book again.</p>\n<p>For a Kindle that doesn't mount on a Mac, <a href=\"/blog/export-kindle-highlights/\">where your Kindle highlights are</a> gives the routes that Amazon offers instead.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kindle-not-showing-up-on-mac/","url":"https://readwyn.com/blog/kindle-not-showing-up-on-mac/","title":"Why your Kindle does not show up on a Mac","summary":"A Kindle Scribe or a 2024 Kindle never mounts on a Mac: it needs an app such as Amazon's Send to Kindle. An older Kindle should mount. If not, check the cable, a macOS alert, and a Finder setting.","content_html":"<p>If it's a Kindle Scribe or a Kindle released in 2024, nothing is broken. These models don't appear in Finder at all: <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page on USB transfers</a> says that on a Mac they “require a separate application”, such as its own Send to Kindle app. A Kindle released before 2024 (the Scribe aside) needs no app and should appear as a disk. When one of those stays out of sight, the things to check are the cable, a macOS alert, and a Finder setting. Readwyn reads a Kindle only as a mounted disk, so it works with the Kindles that Finder shows and can't see a Scribe or a 2024 model.</p>\n<h2 id=\"why-a-scribe-or-a-2024-kindle-never-mounts\">Why a Scribe or a 2024 Kindle never mounts</h2>\n<p>A Kindle used to present itself to a computer as USB mass storage, which a Mac mounts as a disk. The Scribe stopped doing that with software update 5.16.3 in 2023, which <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2023/09/01/kindle-scribe-now-supports-mtp-instead-of-usb-mass-storage-after-update/\">moved it to MTP</a>, the Media Transfer Protocol. In The eBook Reader's words, it now “basically works like an Android device”. The basic Kindle, the Paperwhite, and the Colorsoft of the line-up that Amazon <a rel=\"external\" href=\"https://press.aboutamazon.com/2024/10/amazon-launches-entirely-new-kindle-lineup-including-reimagined-kindle-scribe-and-first-ever-color-kindle\">announced on 16 October 2024</a> also use MTP, as <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2024/11/01/how-to-connect-kindles-to-macs-to-transfer-connect-via-usb/\">The eBook Reader reported</a> on 1 November 2024. Windows supports MTP natively and a Mac doesn't, the same article notes. That's why Finder shows nothing, and why replugging won't help. The fix is an app.</p>\n<p>Amazon's page, as it stands in November 2026, names no model later than 2024. If a newer Kindle charges and never appears, try the app before you blame the cable. If you aren't sure which model you own, <a href=\"/blog/which-kindle-do-i-have/\">Which Kindle do I have?</a> tells them apart.</p>\n<h2 id=\"amazon-s-usb-file-manager\">Amazon's USB File Manager</h2>\n<p>The app that Amazon means is Send to Kindle for Mac, and the part of it that works over the cable is the USB File Manager. Amazon's steps are these:</p>\n<ol>\n<li>Open <a rel=\"external\" href=\"https://www.amazon.com/sendtokindle/mac\">amazon.com/sendtokindle/mac</a>. The page lists “Kindle app for Mac” first, but the USB File Manager is part of the other app: expand “Send to Kindle app for Mac” and click “Get Send to Kindle for Mac” to download the installer.</li>\n<li>Connect the Kindle with the USB cable that came with it.</li>\n<li>Open the app and choose Tools &gt; USB File Manager. If you've signed in with your Amazon account, “Open USB File Manager” at the top right of the app's window does the same.</li>\n<li>A separate window shows the Kindle's files and folders. Drag files between the Mac and the Kindle's “Documents” folder.</li>\n</ol>\n<p><a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page</a> sets the limits. Amazon lists Intel Macs and Macs with an M1 to M3 chip, with macOS 12 and above; the page names no later chip. Files move in and out by drag and drop only. To rename or delete a file on the Kindle, right-click it; Command + Delete also deletes the selected file. More than 2,000 files in one transfer aren't supported. If the window doesn't find the Kindle, Amazon's advice is to close other USB file transfer applications, click “Connect device” in the window, and unplug and replug the cable.</p>\n<p>When you've finished, there's nothing to eject on the Mac: select “Disconnect” on the Kindle's screen, or pull the cable.</p>\n<p>Two other routes skip Amazon's app. The help page says that third-party software can do the same job, and adds that Amazon can't vouch for any of it; The eBook Reader names OpenMTP as one such program. And a file of up to 200 MB can go without a cable from <a rel=\"external\" href=\"https://www.amazon.com/sendtokindle\">Amazon's Send to Kindle web page</a>. Amazon <a href=\"/blog/send-epub-to-kindle/\">converts an EPUB on the way</a>.</p>\n<h2 id=\"an-older-kindle-that-stays-hidden\">An older Kindle that stays hidden</h2>\n<p>A Kindle from before 2024 that isn't a Scribe should show up with no app at all. Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_all_new_paperwhite_v2/v1/en-US/html/kug.html\">user guide for the 10th-generation Paperwhite</a> says it “will appear as an external storage drive or volume on the computer's desktop”, with a folder called “documents” in it. When it doesn't, check these in turn.</p>\n<p><strong>The cable.</strong> The guide says the light on the bottom edge turns amber while the battery charges and green when it's full. No light at all points to a cable that isn't fully in at one end, or to a port that gives too little power. An amber light proves only that power arrives. Some cables are charge-only: they “provide power but do not support USB data transfer”, as <a rel=\"external\" href=\"https://kb.plugable.com/can-a-charge-only-cable-be-used-for-apple-carplay-or-android-auto\">Plugable's support page</a> puts it. Amazon tells you to use the cable that came with the Kindle.</p>\n<p><strong>The Kindle's screen.</strong> A Kindle that has joined the computer is in what the guide calls USB drive mode, and it “is not usable as a reading device while in USB drive mode” until you <a href=\"/blog/eject-kindle-or-kobo-from-mac/\">eject it</a>. So if you've only just plugged it in and you can still turn pages, no data connection exists, whatever the light says.</p>\n<p><strong>The alert.</strong> A Mac laptop with Apple silicon asks before a new USB accessory may connect, <a rel=\"external\" href=\"https://support.apple.com/en-us/102282\">Apple's support page says</a>. Click “Don't Allow” and the Mac won't recognise the Kindle, though the Kindle still charges. The Mac asks again the next time you connect it, so unplug the Kindle, plug it back in, and click “Allow”. If the Mac is locked, the same page says, “you need to unlock it before you can allow an accessory to connect”. You'll find the setting in System Settings &gt; Privacy &amp; Security &gt; Accessories, next to “Allow accessories to connect”, and its default is “Ask for new accessories”.</p>\n<p><strong>Finder.</strong> A mounted disk can still be out of sight. Choose Finder &gt; Settings: the General pane picks the items that show on the desktop and the Sidebar pane those in the sidebar, and <a rel=\"external\" href=\"https://support.apple.com/guide/mac-help/mchlp1039/mac\">external disks are one of the items</a>.</p>\n<p><strong>The port.</strong> <a rel=\"external\" href=\"https://support.apple.com/guide/mac-help/if-a-usb-device-doesnt-work-mchlp1641/14.0/mac/14.0\">Apple's guide to a USB device that doesn't work</a>, in its edition for macOS 14, says to try another port and to connect the device directly to the Mac, with any hub or extension cable taken out. It also gives a test: hold the Option key, choose Apple menu &gt; System Information, and see whether the device is listed below Hardware. If the Kindle isn't in that list, go back to the cable, the port, and the alert. If it is, the Mac can see it over the cable: check the Finder setting again, and whether your model is one that needs the app.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kindle-sdr-folders-and-mbp-files/","url":"https://readwyn.com/blog/kindle-sdr-folders-and-mbp-files/","title":"What are the .sdr folders and .mbp files on a Kindle, and can you delete them?","summary":"They hold a Kindle's place, highlights and notes for each book. Deleting one leaves the book intact, but a book copied over USB loses what that file held.","content_html":"<p>They're the Kindle's own notes on each book: where you stopped, what you highlighted, and the notes and bookmarks you left. A Kindle on firmware 1 to 4 (the Kindle 1 and 2, the DX, the Kindle Keyboard and the Kindle 4) writes an <code>.mbp</code> file beside the book. Firmware 5 and later (the Touch, the Paperwhite, the Voyage, the Oasis, the basic Kindle from 2014, the Scribe and the Colorsoft) keeps a folder ending in <code>.sdr</code> instead. You can delete either one without harming the book, and the Kindle starts a fresh one when you next open it. What goes with it is your place and, for a book you copied over a cable, the highlights and notes that only that file held. Readwyn, the Mac and Linux ebook library app behind this blog, can copy those into a library first, from a Kindle that the computer mounts as a drive. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">help page on USB transfers</a> says the Kindle Scribe and the Kindles released in 2024 need a separate application on a Mac, and that models released earlier (the Scribe aside) don't. Readwyn works with those earlier ones.</p>\n<h2 id=\"the-mbp-file-of-an-older-kindle\">The <code>.mbp</code> file of an older Kindle</h2>\n<p>On a classic Kindle the file sits in <code>documents</code> next to the book and shares its name: <code>Emma.mobi</code> gets <code>Emma.mbp</code>. The MobileRead wiki's page on <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Mbp\">the .mbp extension</a> describes it as a side file of Mobipocket books, made when a book is first opened, that holds “bookmarks, annotations, last read position”.</p>\n<p>Inside, it's a Palm database, the same container that a MOBI book uses. Your place is a count of bytes into the book's uncompressed text, and a date in the file's header records the last time you read. Each highlight carries a copy of the words you marked. A book that the Kindle has never opened has no <code>.mbp</code> file at all.</p>\n<h2 id=\"the-sdr-folder-of-a-newer-kindle\">The <code>.sdr</code> folder of a newer Kindle</h2>\n<p>From firmware 5 on, the Kindle makes a folder with the name of the book file, minus its extension, plus <code>.sdr</code>. The folder for <code>Emma.azw3</code> is <code>Emma.sdr</code>. Inside are two files, and their extensions follow the format of the book. jhowell, who wrote a parser for them, <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=3878110\">listed the pairs on MobileRead</a> in 2019: one file for the things that change at every page turn, such as the last page read and reading timing, and one for “less frequently changed data such as personal annotations”.</p>\n<table><thead><tr><th>Book format</th><th>Your place and reading time</th><th>Highlights, notes, bookmarks</th></tr></thead><tbody>\n<tr><td>MOBI (<code>.mobi</code>, <code>.azw</code>)</td><td><code>.mbs</code></td><td><code>.mbp1</code></td></tr>\n<tr><td>AZW3 (KF8)</td><td><code>.azw3f</code></td><td><code>.azw3r</code></td></tr>\n<tr><td>KFX</td><td><code>.yjf</code></td><td><code>.yjr</code></td></tr>\n</tbody></table>\n<p>The <code>.mbp1</code> file is not a newer <code>.mbp</code>: the digit marks a different format. A PDF gets a pair too, <code>.pdt</code> and <code>.pds</code>. In these files a highlight is a pair of positions in the book, with no copy of the words. The folder can hold more. jhowell <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=4342040\">wrote in 2023</a> that it can carry the pop-ups at the start and end of a book, page number data, X-Ray content, and even part of the book itself. The reader who asked had found folders larger than their books.</p>\n<h2 id=\"what-deleting-one-costs\">What deleting one costs</h2>\n<p>The book itself doesn't depend on these files. Your place in it does, and so do the highlights, notes and bookmarks you see inside it. How much of that comes back depends on where the book came from.</p>\n<p>A book from the Kindle Store, or a document sent through Send to Kindle and archived, has Amazon behind it, as long as Whispersync for Books is on (Settings, Device Options, Whispersync). Whispersync, Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">user guide for the Kindle Colorsoft</a> says, “synchronizes your last page read, notes, highlights, and bookmarks”. jhowell described the files on the Kindle as a cache, and <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=3963475\">told a reader who replaced them</a> that “the master copy is still on Amazon's server.”</p>\n<p>A book you copied over a cable has no such backup. The same guide says files transferred by USB “are not stored in the Cloud”, and it limits the syncing of notes and highlights to personal documents archived in Kindle format. Delete its <code>.mbp</code> file or <code>.sdr</code> folder and the Kindle forgets your place and shows the book clean. The words of your highlights survive in one other place: the Kindle adds each note and highlight to a file called My Clippings, which on the drive is <code>documents/My Clippings.txt</code>. Deleting a book's sidecar doesn't touch it. One MobileRead user deleted the two files in a <code>My Clippings.sdr</code> folder and saw no loss, and was told by shamanNS <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=4227716\">in the same thread</a> that such files “will just be recreated” at the next bookmark, highlight or note.</p>\n<p>Two cases are safe to clear. A <code>.sdr</code> folder whose book is gone is an orphan, and in a <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=3712674\">2018 thread on stray folders</a> the advice was to remove only those, never one that still has a book beside it. Empty <code>.sdr</code> folders also turn up. A forum member testing a Paperwhite 2 and 3 <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=3741092\">found they appeared</a> when the Kindle was connected by USB, or went to sleep, with a book open.</p>\n<p>What not to do is rename. The folder belongs to the book by its name, so a book renamed in Finder no longer has a folder of the same name. jhowell <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=4342040\">wrote</a> that the firmware is “somewhat tolerant” of changes in the documents folder, but that “renaming or moving those files has the potential to cause problems.” Moving the book and its folder together is another matter: in the thread where jhowell called the files a cache, a user who copied sideloaded AZW3 books with their <code>.sdr</code> folders to another Kindle <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=3962743\">reported</a> that it “works fine”, with the highlights showing on the new Kindle.</p>\n<h2 id=\"getting-the-contents-out-first\">Getting the contents out first</h2>\n<p><a href=\"/blog/export-kindle-highlights/\">Where your Kindle highlights are</a> covers My Clippings.txt. Readwyn reads the sidecars themselves. With the Kindle connected, “Device → Readwyn” on the Sync screen brings the reading progress, highlights and notes of each MOBI or AZW3 book that is also in the library, from the <code>.mbp</code> file of a classic Kindle or the <code>.sdr</code> files of a newer one (<a href=\"/blog/kindle-highlights-in-ebook-library/\">more on that sync</a>). Connecting the Kindle only reads it. When you then take a book off with “Remove from device”, in the row “On the device” of the book's Details tab, Readwyn deletes the book file and everything with its base name and one of the extensions <code>.mbp</code>, <code>.tan</code>, <code>.pdr</code>, <code>.ea</code>, <code>.apnx</code>, <code>.phl</code>, <code>.mbp1</code>, <code>.mbs</code> or <code>.han</code>, and the <code>.sdr</code> folder. The <code>.apnx</code> file is the one that gives a book <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/APNX\">page numbers tied to a print edition</a>. The highlights stay in the library.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kindle-vocabulary-builder-on-mac/","url":"https://readwyn.com/blog/kindle-vocabulary-builder-on-mac/","title":"Bring your Kindle Vocabulary Builder words to your Mac","summary":"A Device → Readwyn sync copies the Vocabulary Builder words of a Kindle into My Words on your Mac, for the MOBI and AZW3 books that are also in your library, each with its book and first lookup date.","content_html":"<p>The words that a Kindle collects in its Vocabulary Builder come across to your Mac in one Device → Readwyn sync over a USB cable, for each book that is both on the Kindle and in your library. They land in Readwyn’s My Words, each with the book you met it in and the date you first looked it up. The book’s file on the Kindle has to be a MOBI or AZW3 file: words from a book in Amazon’s KFX format stay on the Kindle. Readwyn reads a Kindle only as a drive that the Mac has mounted. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon’s help page on USB transfers</a> says the Kindle Scribe and the Kindles released in 2024 need a separate app, such as Amazon’s Send to Kindle, for a cable transfer on a Mac, so Readwyn can’t reach those. Models released before 2024, other than the Scribe, need no such app. If you don’t know which model you have, <a href=\"/blog/which-kindle-do-i-have/\">this is how to find out</a>.</p>\n<h2 id=\"a-list-the-kindle-writes-by-itself\">A list the Kindle writes by itself</h2>\n<p>Amazon’s <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/EM10/Kindle_User_Guide_en-US.pdf\">Kindle Paperwhite User’s Guide</a> (2nd edition) puts it in one line: “Words you look up in the dictionary are automatically added to Vocabulary Builder.” The same guide has the switch for it under All Settings, Reading Options, Language Learning, and says that no word is added while it’s off. A <a href=\"/blog/kobo-my-words-on-mac/\">Kobo works the other way round</a> and keeps only the words you choose to save. On a Kindle the list is a record of your lookups, not a selection.</p>\n<p>On the Kindle’s drive the whole list is one file, <code>system/vocabulary/vocab.db</code>, an SQLite database. Readwyn opens it with a connection that can’t write. One of Readwyn’s tests runs on a real <code>vocab.db</code> from a Kindle with firmware 5: all 410 words come out with the time of their first lookup, and the database and its folder are unchanged afterwards.</p>\n<p>Look a word up again in another book a year later and it still arrives once, under the first book and the first date. Nothing else is read: not the usage example that the Kindle shows with a word, and not whether you’re still learning the word or have mastered it.</p>\n<h2 id=\"which-words-come-across\">Which words come across</h2>\n<p>A sync covers a word only when its book is both on the Kindle and in your library, the same rule as for <a href=\"/blog/kindle-highlights-in-ebook-library/\">Kindle highlights and notes</a>. That takes two matches.</p>\n<p>First, Readwyn finds the word’s book among the files on the Kindle: by the unique ID in the header of the file, else by its ASIN, else by its title. A MOBI or an AZW3 file has that header. A book in Amazon’s KFX format, a file whose name ends in <code>.kfx</code>, doesn’t, so a word you looked up in such a book gets no book and stays on the Kindle. Only the book of the first lookup counts. If it’s no longer on the Kindle, the word is left out, even when a later book is still there.</p>\n<p>Second, that file has to match a book in your library. Readwyn tries the path it recorded when it sent the book, then the list that calibre leaves on a device, then the title.</p>\n<p>Nothing is dropped silently. Under “Left out”, the review names each word that stays behind and gives the reason: “Its book is not on the device.”, or a sentence that names the book and ends “is not in this library.” A word from a KFX book gets the first reason too, although its file is on the Kindle, because Readwyn can’t read that file. In the real database of that test, four words belonged to <em>What If?</em>, the one book that was on the test Kindle. The words of the other 31 books got no book, and in a sync they would all be left out.</p>\n<h2 id=\"run-the-sync\">Run the sync</h2>\n<p>Plug the Kindle in and click “Sync” in the sidebar. Connecting only reads the device. For a Kindle with a Vocabulary Builder, the “Device → Readwyn” card says “Bring progress, highlights, notes and My Words from the Kindle into the library.”, and its button counts the changes (“Review 8 changes”). In the review, each new word is a row under its book, marked “Add · Word”. Every change has its own checkbox, so you can untick the progress and the highlights and apply the words alone. A word that My Words already has, in any capitalisation, isn’t added a second time.</p>\n<p>Leave “Merge highlights, notes and words instead of replacing them” ticked at the top of the review, as it is by default. With it off, My Words is made to match the Kindle for each synced book: a word that you saved on the Mac from one of those books, and that the Kindle doesn’t have, shows as “Delete · Word”.</p>\n<p>The other card is short on such a Kindle: “Nothing goes from the library to this Kindle. The list below tells why.” That list, “Not available on this Kindle”, has the entry “My Words to the device” with the reason “The Vocabulary Builder of the Kindle has no public description of how to add a word.”</p>\n<p>On a Kindle with firmware 1 to 4, the list has “My Words” and “Kindle firmware 1 to 4 has no Vocabulary Builder.” Amazon <a rel=\"external\" href=\"https://press.aboutamazon.com/uk/2013/9/introducing-the-all-new-kindle-paperwhite-the-6th-generation-of-kindle\">announced the feature as new in September 2013</a>, with that year’s Kindle Paperwhite. By <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2023/02/06/list-of-latest-software-version-for-each-kindle/\">The eBook Reader’s list of the last software version of each model</a>, the Kindles that stopped at software 1 to 4 are the first Kindle, the Kindle 2, the Kindle DX, the Kindle Keyboard, and the basic Kindle of the 4th and 5th generation. Software 5 by itself doesn’t promise the feature. The Kindle Touch of the 4th generation stopped at 5.3.7.3 in that list. The first Kindle Paperwhite, the 5th generation in that list, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2014/03/06/1st-gen-kindle-paperwhite-gets-lots-of-new-features-with-software-update/\">got the Vocabulary Builder with software 5.4.4</a>, in March 2014, a higher version than the Touch’s last. Whatever the model, a Kindle without <code>system/vocabulary/vocab.db</code> gives a sync no words.</p>\n<h2 id=\"the-words-on-your-mac\">The words on your Mac</h2>\n<p>Click “Notebook” in the sidebar, then “My Words”. The words are grouped by headword, newest first, each with its book and its date. Readwyn stores no definitions. It looks every word up in the dictionaries on your Mac when it shows the list, in <a href=\"/blog/dictionary-order/\">the order you’ve set for them</a>, so the entry you read comes from the Mac’s dictionaries, not the Kindle’s. A word that none of them knows says “No entry in the dictionaries on this Mac.”</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-notebook-words.webp?h=cd3bb9313f6f7a11a159\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-notebook-words.webp?h=b4ba5a2db81a6383cd4a\" alt=\"My Words in the Notebook: All books with 103 words and four books at the left, and at the right the headword forlorn with its full entry open and the saved words forlornness in Middlemarch and forlorn in Persuasion\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>My Words in the Notebook: one group for each headword, with the book and the date of each word</figcaption>\n</figure>\n<p>Readwyn has no export of words to a file. What it has is “Copy all”, which puts the list on the clipboard as plain text, one block for each headword with its first sense, and under it each of your words with its book:</p>\n<pre><code>forlorn — adjective. sad and lonely, as if left behind by everyone.\n  forlornness, in Middlemarch\n  forlorn, in Persuasion\n</code></pre>\n<p>Searching the list, narrowing it to one book and removing words are covered in <a href=\"/blog/save-looked-up-words/\">Save the words you look up and review them later</a>.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/mobi-azw3-or-epub-for-kindle-usb/","url":"https://readwyn.com/blog/mobi-azw3-or-epub-for-kindle-usb/","title":"MOBI, AZW3, or EPUB: which format to put on a Kindle over USB","summary":"Copy AZW3 to any Kindle from the Kindle Keyboard on, and MOBI to the first Kindle, the Kindle 2 and the DX. No Kindle opens an EPUB that comes over the cable.","content_html":"<p>Over a cable, AZW3 is the file to copy to any Kindle from the Kindle Keyboard of 2010 onwards (the Keyboard and the fourth-generation Kindle need a software update first). The three models before it (the first Kindle, the Kindle 2 and the Kindle DX) read MOBI but not AZW3, and no Kindle opens an EPUB that arrives over USB: an EPUB gets there only through Amazon's wireless Send to Kindle service, which converts it on the way. On a Mac the cable has a catch of its own. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page</a> says the Kindle Scribe and the Kindles released in 2024 need a separate application there, such as its Send to Kindle app, and that the other models released before 2024 don't. Readwyn works only with a Kindle that the computer mounts as a drive, which on a Mac leaves out the Scribe and the 2024 models.</p>\n<h2 id=\"mobi-is-still-on-the-list\">MOBI is still on the list</h2>\n<p>A Kindle still reads MOBI. Amazon's page for <a rel=\"external\" href=\"https://www.amazon.com/dp/B0CFPJYX7P\">the twelfth-generation Paperwhite</a> gives the formats in two groups, under \"Content Formats Supported\". The first is \"Kindle Format 8 (AZW3), Kindle (AZW), TXT, PDF, unprotected MOBI, PRC natively\". The second group, which has Word documents, HTML, pictures and EPUB in it, works \"through conversion\". Converting is what Send to Kindle does, so EPUB is on the list without being a file the Kindle opens itself. Standard Ebooks is blunt about it in <a rel=\"external\" href=\"https://standardebooks.org/help/how-to-use-our-ebooks\">its Kindle FAQ</a>: Kindles \"still do not natively support epub files\", and you can't transfer one with a USB cable.</p>\n<p>Two words in that first group need a gloss. \"Unprotected\" means free of DRM: Amazon <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200735130\">tells publishers</a> that a Kindle recognises a transferred file only if it has none. PRC is MOBI under another extension, one that dates from Palm OS, which took only .prc and .pdb files (<a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/MOBI\">MobileRead wiki</a>).</p>\n<p>What did end is MOBI by Send to Kindle. An email from Amazon, which The eBook Reader <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2023/09/25/send-to-kindle-losing-mobi-support-probably-for-real-this-time\">reported in September 2023</a>, set the dates: the wind-down would start on 1 November 2023, and all Send to Kindle support for MOBI would end by 20 December 2023. The report added that the older formats could still go across by cable. <a href=\"/blog/send-epub-to-kindle/\">Can you send an EPUB to a Kindle now?</a> covers the wireless route.</p>\n<h2 id=\"why-azw3-is-the-better-file\">Why AZW3 is the better file</h2>\n<p>MOBI began as the format of the Mobipocket reader. The <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/MOBI\">MobileRead wiki</a> describes its markup as an extended HTML 3.2 with no CSS: nested tables aren't supported, a left margin moves only in fixed steps, and a phrase can't be set in a monospace font. An AZW3 file holds Kindle Format 8 (KF8) instead, which Amazon <a rel=\"external\" href=\"https://techcrunch.com/2011/10/20/amazon-throws-a-minor-curveball-with-html5-kindle-8-format\">announced in October 2011</a>. KF8 takes HTML5 and CSS3, and with them <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/KF8\">embedded fonts, floats, nested tables, merged cells, and SVG</a>. Standard Ebooks offers Kindle owners AZW3 and no MOBI at all, because MOBI lacks \"many basic features\", hyphenation among them.</p>\n<p>Amazon has been leaving MOBI behind as well. Its 2022 email about Send to Kindle, which <a rel=\"external\" href=\"https://www.fivefilters.org/2022/kindle-epub-support/\">FiveFilters quotes</a>, calls MOBI an older format that \"won't support the newest Kindle features for documents\". Kindle Direct Publishing <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GULSQMHU5MNH4EZM\">stopped taking MOBI uploads</a> for reflowable books on 1 August 2021 and for fixed-layout books on 18 March 2025 (<a href=\"/blog/kindle-direct-publishing-file-formats/\">the KDP formats guide</a> has the rest).</p>\n<p>Reach is the one thing MOBI still has. MobileRead's <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/KF8\">KF8 page</a> gives the support as every Kindle except the first and second generations and the Kindle DX. The Kindle 3, the model <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Amazon_Kindle_devices#Kindle_Keyboard\">later renamed the Kindle Keyboard</a>, got it with <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2012/09/11/software-update-for-kindle-keyboard-v3-4-adds-new-features/\">software update 3.4</a> in September 2012, and a fourth-generation Kindle needs firmware 4.1.0 or later. calibre's manual makes the same split in <a rel=\"external\" href=\"https://manual.calibre-ebook.com/generated/en/ebook-convert.html\">the options of <code>ebook-convert</code></a>: MOBI 6 \"is compatible with all devices\", while KF8 \"only works with newer Kindles\". If you don't know which Kindle is in your hand, <a href=\"/blog/which-kindle-do-i-have/\">the model guide</a> helps.</p>\n<h2 id=\"a-mobi-file-isn-t-always-mobi\">A .mobi file isn't always MOBI</h2>\n<p>The extension doesn't settle the format. Project Gutenberg's page for <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1342\">Pride and Prejudice</a> has two Kindle downloads, labelled \"Kindles (kf8)\" and \"Older Kindles\". The first arrives as <code>pg1342-images-kf8.mobi</code> and the second as <code>pg1342-images.mobi</code>: one extension, two formats. calibre does the same with its <code>--mobi-file-type</code> option, whose values <code>old</code>, <code>new</code> and <code>both</code> put MOBI 6, KF8, or the pair of them inside a .mobi file. The pair is a <a href=\"/blog/mobi-kf8-combo-files/\">combo file</a>. It is larger, and it exists so that an older Kindle can still open the book.</p>\n<h2 id=\"where-the-right-file-comes-from\">Where the right file comes from</h2>\n<p>Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/help/reading_options.html\">reading help</a> says to take one of those two downloads for USB, not the EPUB, and to copy it into the <code>documents</code> folder of the Kindle. Standard Ebooks <a rel=\"external\" href=\"https://standardebooks.org/help/how-to-use-our-ebooks\">adds a step</a>. Beside each AZW3 it offers a cover thumbnail, which goes into the <code>thumbnails</code> folder inside <code>system</code> after the book is copied, because a sideloaded book otherwise shows a default cover. For a Kindle DX it tells you to convert its EPUB to MOBI.</p>\n<p>A DRM-free EPUB from anywhere else has to be converted first. Readwyn's <a href=\"/blog/put-epub-on-kindle-from-mac-usb/\">“Send to device”</a> converts the book's EPUB to MOBI 6, the format that every Kindle reads, the oldest included, and writes it to <code>documents/&lt;title&gt; - &lt;first author&gt;.mobi</code>. That copy keeps the paragraphs, headings, lists, tables, bold, italic and pictures (SVG images apart), and drops the book's fonts. Of the CSS, only alignment, weight, italics, underline and strike-through carry over, rewritten as MOBI's own markup. The file has one part for each document of the EPUB, so a highlight made on the Kindle carries over to the exact place in the library's EPUB at the next sync.</p>\n<h2 id=\"the-cable-on-a-mac\">The cable on a Mac</h2>\n<p>The Kindle Scribe and the Kindles of 2024 don't show up as a drive. The Scribe moved from USB mass storage to the Media Transfer Protocol <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2023/09/01/kindle-scribe-now-supports-mtp-instead-of-usb-mass-storage-after-update/\">with a software update in 2023</a>, and the 2024 Kindle, Paperwhite and Colorsoft use it too. Windows works with MTP natively and a Mac doesn't, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2024/11/01/how-to-connect-kindles-to-macs-to-transfer-connect-via-usb/\">The eBook Reader explained</a> in November 2024. Amazon's answer is the USB File Manager in Send to Kindle for Mac, under Tools, where you drag the file into the Kindle's “Documents” folder. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Its help page</a> names EPUB only among the file types of the wireless service. The file to drag is still an AZW3. If a Kindle from before 2024 that isn't a Scribe doesn't appear on the Mac either, the things to check are <a href=\"/blog/kindle-not-showing-up-on-mac/\">the cable, a macOS alert and a Finder setting</a>.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/move-books-from-old-kindle-to-new-kindle/","url":"https://readwyn.com/blog/move-books-from-old-kindle-to-new-kindle/","title":"How to move your books from an old Kindle to a new Kindle","summary":"Register the new Kindle to the same Amazon account and download your books from the cloud. Files you copied to the old Kindle by cable have to be copied again by hand.","content_html":"<p>Nothing passes from one Kindle to the other. Register the new Kindle to the Amazon account that the old one is on, and the books you bought from the Kindle Store are in its library, ready to download from Amazon's cloud. Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">user guide for the Kindle Colorsoft</a> has a section called \"Transferring content from another Kindle\", and it sorts your content by where it came from. Store purchases are in the cloud. Documents you emailed to the Kindle are there too, if Personal Document Archiving was on. Whatever was \"transferred directly to your previous Kindle\" over a cable \"will need to be transferred manually\", through a computer. Readwyn, the ebook library app for Mac and Linux that this blog belongs to, can copy the reading progress, highlights and notes of those cable-loaded books into its own library, if they are MOBI or AZW3 files and the library has the same books. It reads only a Kindle that the computer mounts as a drive. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">help page on USB transfers</a> says the Kindle Scribe and the Kindles released in 2024 need a separate application on a Mac, and that models released earlier than 2024 (the Scribe excepted) don't. Those earlier models are the ones Readwyn works with.</p>\n<h2 id=\"sync-the-old-kindle-first\">Sync the old Kindle first</h2>\n<p>Your place in each book, your notes and highlights, and your collections reach Amazon when the Kindle syncs, so sync the old one while it's still in your hands. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GDCAMDFMC2LZP6BR\">Amazon's steps</a> are short: connect the Kindle to the internet, swipe down from the home screen to open the Quick Actions (or select Menu), and choose the Sync icon or \"Sync and Check for Items\".</p>\n<p>Sync is on by default, <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GGFEXXS8Z7DPJSTN\">Amazon's page on the setting</a> says. The Colorsoft guide calls the setting Whispersync for Books and describes it as the \"automatic backup of your annotations, most recent page read, furthest page read, and collections\". If you turned it off at some point, read that page before you turn it back on. It warns that notes, highlights and bookmarks made while Sync was off \"may be lost when turning this setting back on\".</p>\n<p>One group of Kindles can't do this any more. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TRXsYxKJr4WTdsVs2P\">Amazon's page on the Kindles released in 2012 or earlier</a> says that after 20 May 2026 they \"will no longer be able to connect to Amazon Services\", which leaves one of them nothing to sync with. The books you bought stay on the account: the same page says that your Kindle library is accessible on any Kindle released after 2012. For what is only on the old Kindle, the cable is the way out (see the files you copied by cable, below).</p>\n<h2 id=\"register-the-new-kindle-and-download\">Register the new Kindle and download</h2>\n<p>Setup asks for a language and then offers \"Set up on this Kindle\" or \"Set up with your phone\", which registers the Kindle through the Kindle app on an iOS or Android device. A Kindle that you bought online with your own Amazon account, the guide says, \"is already registered to you\". Your Amazon user name in the upper left corner of the Home screen confirms it. If that corner says \"My Kindle\" or has a previous owner's name, as a second-hand Kindle might, open the Quick Actions menu and tap All Settings, then Your Account, then Register/Deregister Device.</p>\n<p>After that the library lists everything on the account. By default it shows what's in the cloud and what's on the device together, and a downloaded book has a check mark: in the lower left corner of its cover in the grid view, and on the right side of the screen in the list view. To download a book, tap Your Library and tap the book. If the list shows only downloaded books, open the filter menu at the top left and deselect the Downloaded filter first. A browser gives a second route. Amazon's page on <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GGLBJ4VR79NSB26W\">sending books to a Kindle</a> has you open the Content tab of Manage Your Content and Devices, select Books, choose \"Deliver or Remove from Device\", pick the new Kindle from the list of your devices, and confirm with \"Make Changes\".</p>\n<p>The rest follows the account. Whispersync, in the guide's words, \"synchronizes your last page read, notes, highlights, and bookmarks\". Collections are stored in the cloud and \"synced between other devices and reading apps\" that are registered to the same Amazon account. Two choices may not come with them: how collections show in the library, and which ones you marked as favourites. The guide covers both in one passage and closes it with the note that \"this setting is device-specific\", so check both on the new Kindle.</p>\n<p>Documents that reached the old Kindle through Send to Kindle come down from the library as bought books do, as long as archiving was on. <a href=\"/blog/does-kindle-sync-sideloaded-books/\">Does a Kindle sync the progress of sideloaded books?</a> has the setting, and what syncs for such a document.</p>\n<h2 id=\"the-files-you-copied-by-cable\">The files you copied by cable</h2>\n<p>Amazon's cloud never held these, so the account can't bring them back. If the originals are still on your computer, copy those to the new Kindle. If the old Kindle has the only copies, the job has two steps: connect the old Kindle and copy the files out of its documents folder to the computer, then connect the new one and copy them into its documents folder.</p>\n<p>How a Kindle connects depends on the computer and on the Kindle. On Windows, Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">help page</a> says, a Kindle appears as an external storage drive. On a Mac, only a Scribe or a 2024 model needs a separate application, whether it's the old Kindle or the new one. Amazon's own Send to Kindle app will do: connect the Kindle, open the app, and choose Tools &gt; USB File Manager. That tool moves files by drag and drop only, and the help page adds that \"USB transfers of more than 2,000 files at a time are not supported\", so a very full documents folder goes across in batches. <a href=\"/blog/kindle-not-showing-up-on-mac/\">Why your Kindle does not show up on a Mac</a> has the steps. Which of your files a current Kindle opens is a matter of format, and <a href=\"/blog/mobi-azw3-or-epub-for-kindle-usb/\">MOBI, AZW3, or EPUB</a> goes through that.</p>\n<p>Two things are easy to forget. Fonts that you installed yourself are files as well: the guide has you connect the Kindle to a computer and points to the Readme in its fonts folder, so the new Kindle <a href=\"/blog/add-fonts-to-kobo-and-kindle/\">needs them copied again</a>. And what you marked in these books has no copy at Amazon. The guide offers syncing of notes, highlights, bookmarks and the last page read only for personal documents \"archived in Kindle format\", and a file that came over the cable isn't archived at all. Its highlights stay on the old Kindle, in a file beside the book and in the running list that the guide calls My Clippings. Copy <code>My Clippings.txt</code> from the documents folder to your computer along with the books (<a href=\"/blog/export-kindle-highlights/\">where Kindle highlights are kept</a>). If the books are also in a Readwyn library and the old Kindle mounts as a drive, connect it once more before you put it away: “Device → Readwyn” on <a href=\"/blog/sync-kindle-with-mac-without-calibre/\">the Sync screen</a> brings the reading progress, highlights and notes of its MOBI and AZW3 books into the library, for each book that is on both. They stay in the library: Readwyn writes none of them to a Kindle on firmware 5 or later, and that is every Paperwhite and every basic Kindle from 2014 on.</p>\n<h2 id=\"the-old-kindle-afterwards\">The old Kindle afterwards</h2>\n<p>It can stay registered and keep working beside the new one: the guide says you can synchronise your Kindle books with the other supported devices you own. If it's going to someone else, Amazon <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GP383NNSU4NSHP7C\">recommends a factory reset</a> before gifting or selling, \"to remove all content and user settings from the device\". The guide's note on the reset is the reason to do the cable copy first: \"you must back up any personal documents that you'd like to keep\". Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=201357520\">page on deregistering a device</a> says to do both, deregistering first and resetting second. Deregistering costs you nothing that the account holds: it \"removes content from your device but not your Amazon account\". The files you copied by cable aren't on the account, so copy them off before you deregister, not only before the reset.</p>\n<p>A Kindle released in 2012 or earlier is the exception here too. The Amazon page that recommends the reset now tells you not to deregister or reset one of them, and <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TRXsYxKJr4WTdsVs2P\">its page on these Kindles</a> gives the reason: after either, you can't register the Kindle again \"or use the device in any way\". The books already downloaded to such a Kindle still open while it stays registered (<a href=\"/blog/old-kindle-end-of-support/\">why an old Kindle can't reach the store</a>), so leave it as it is and it stays a working reader.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/moving-from-kindle-to-kobo/","url":"https://readwyn.com/blog/moving-from-kindle-to-kobo/","title":"Moving from a Kindle to a Kobo: which of your books can come with you","summary":"Kindle Store books with DRM stay with Amazon. Your own DRM-free files, Kindle books that offer an EPUB download, free classics and your library card can move to a Kobo.","content_html":"<p>The books you bought in the Kindle Store with DRM stay with Amazon. They won't open on a Kobo: Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GDDXGH9VR22ACM8U\">help for publishers</a> says that applying DRM \"limits reader access to Kindle apps and devices only\". Those books aren't lost. They stay on your Amazon account, readable on the old Kindle and in the Kindle apps. These can come with you: files without DRM that you put on the Kindle yourself, the Kindle Store books whose authors have switched on an EPUB download, free public-domain editions, and your library card.</p>\n<h2 id=\"kindle-store-books\">Kindle Store books</h2>\n<p>Whether a Kindle book carries DRM is the publisher's decision, made one book at a time. The <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=201014950\">Kindle Store Terms of Use</a> license a book for reading \"solely through Kindle Software or as otherwise permitted as part of the Service\". Where a book carries DRM, they also forbid any attempt to \"bypass, modify, defeat, or otherwise circumvent\" it, and the law <a href=\"/blog/is-removing-drm-legal/\">adds rules of its own</a>.</p>\n<p>Amazon's website used to let you download a file of a Kindle book you'd bought, through an option called “Download &amp; Transfer via USB”. Then Amazon gave notice that from 26 February 2025 the option \"will no longer be available\", as <a rel=\"external\" href=\"https://goodereader.com/blog/kindle/amazon-removing-download-and-transfer-on-the-kindle-feb-26th\">Good e-Reader reported</a>. The post on <a href=\"/blog/kindle-download-and-transfer-removed/\">why that option went</a> has the rest of the story.</p>\n<p>One group of books does come with a file. Since 20 January 2026, according to the same help page for publishers, someone who bought a book that its author has confirmed as DRM-free can download it as an EPUB and a PDF from the Manage Your Content and Devices page. A book published before 9 December 2025 gets the download only after its author turns it on, and a borrowed book never gets it. The page is written for authors who publish through Kindle Direct Publishing, so it describes their books.</p>\n<p>Amazon gave these books no label in the shop when the downloads began. A week after the start, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2026/01/27/amazon-needs-to-show-which-kindle-ebooks-are-available-as-epubs/\">The eBook Reader</a> found nothing on the product pages and no filter on the Manage Your Content page: the only way to find them was to go through your purchases there one at a time and look for the note “Download available in additional formats”. An EPUB from there is an ordinary file, and a Kobo opens it.</p>\n<p>Everything else keeps working where it is. The <a rel=\"external\" href=\"https://apps.apple.com/us/app/amazon-kindle/id302584613\">Kindle app</a> runs on iPhone, iPad and Mac, and its App Store listing says the Kindle books you've bought \"will automatically appear in your app\". There's <a rel=\"external\" href=\"https://play.google.com/store/apps/details?id=com.amazon.kindle\">an Android version</a> as well.</p>\n<h2 id=\"files-you-put-on-the-kindle-yourself\">Files you put on the Kindle yourself</h2>\n<p>A book that you copied to the Kindle over a cable is your own file, and the Kindle may hold the only copy of it. The <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">Kindle Colorsoft user guide</a> says files transferred by USB \"are not stored in the Cloud\". They sit in the folder called <code>documents</code>, so copy that folder to a computer before the Kindle goes into a drawer. On Windows, <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page</a> says the Kindle shows up in File Explorer with no extra software. On a Mac, the same page says a Kindle Scribe or a Kindle released in 2024 needs a separate application, such as the USB File Manager of Amazon's Send to Kindle app, and that the models released before 2024 (the Scribe excepted) need none.</p>\n<p>A document that went through Send to Kindle is another case. Amazon keeps a copy in your Kindle library, but only if you've enabled Personal Document Archiving, according to the Colorsoft guide. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G5WYD9SAF7PGXRNA\">Amazon's page on the service</a> puts that library on your Kindle devices and in the Kindle app, so for the Kobo, go back to the file that you sent.</p>\n<p>What happens next depends on the format. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">list of supported formats</a> for its e-readers includes EPUB, PDF and MOBI for books, CBZ and CBR for comics, and TXT, HTML and RTF for text. An EPUB, a PDF or a MOBI without DRM can therefore go across as it is. AZW3, the file that <a rel=\"external\" href=\"https://standardebooks.org/help/how-to-use-our-ebooks\">Standard Ebooks</a> makes for Kindles, isn't on the list and has to become an EPUB first.</p>\n<p>Readwyn, the ebook library app for Mac and Linux that this blog belongs to, does that conversion when you add the file. A MOBI or an AZW3 without DRM becomes an EPUB 3, and the original stays in the library unchanged. “Send to device” then puts the book on a connected Kobo as a <a href=\"/blog/kepub-or-epub-for-kobo/\">KEPUB</a>, unless you've turned that option off. A file with DRM doesn't convert.</p>\n<h2 id=\"getting-files-onto-the-kobo\">Getting files onto the Kobo</h2>\n<p>Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">instructions for a cable</a> are short. Connect the Kobo, tap “Connect” when it shows “Computer detected”, open the drive named KOBOeReader on the computer, drag the files onto it, and eject. A book added this way appears only on that Kobo. It doesn't reach your other Kobo e-readers or the Kobo apps, even after a sync.</p>\n<p>Five models can also collect files over Wi-Fi. The Forma, Sage, Elipsa, Elipsa 2E and Libra Colour link to <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/15335985512983-Add-books-to-your-eReader-using-Google-Drive\">Google Drive</a> or <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360033830114-Add-books-to-your-eReader-using-Dropbox\">Dropbox</a>, and a file of a supported type that you put in the “Rakuten Kobo” folder of that account arrives at the next sync. Both pages say the route is for files without DRM.</p>\n<h2 id=\"free-books-and-library-loans\">Free books and library loans</h2>\n<p>A public-domain classic is quicker to fetch again than to move. <a href=\"/blog/standard-ebooks/\">Standard Ebooks</a> publishes each of its books as a kepub file, which its help page calls \"specially prepared for the best reading experience on Kobos\", beside the azw3 for Kindles.</p>\n<p>A library loan stays behind, but the card moves. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017677053-Borrow-eBooks-from-the-public-library-using-your-Kobo-eReader\">page on borrowing</a> names 15 models that borrow from public libraries through OverDrive, in 13 countries, and <a href=\"/blog/library-ebooks-on-kobo-and-kindle/\">the post on library ebooks</a> has the steps.</p>\n<h2 id=\"highlights-and-notes\">Highlights and notes</h2>\n<p>The Kindle adds every note and highlight to a file \"in your library called My Clippings\", in the words of the Colorsoft guide, and <a href=\"/blog/export-kindle-highlights/\">where your Kindle highlights are</a> explains how to copy them out. For a document that didn't come from the Kindle Store, the same guide offers syncing of notes and highlights only if it is one of the \"personal documents archived in Kindle format\", which a file copied by cable isn't. Highlights in those books don't sync to your Amazon account, so take My Clippings along with the books.</p>\n<p>What that file gives you is the text. Copying it across won't put the marks back into the same book on the Kobo, because <a href=\"/blog/move-highlights-between-reading-apps/\">each reading app records a highlight as a position in its own copy</a>.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/moving-from-kobo-to-kindle/","url":"https://readwyn.com/blog/moving-from-kobo-to-kindle/","title":"Moving from a Kobo to a Kindle: what happens to your books","summary":"Books bought from Kobo stay in your Kobo account, and most won't open on a Kindle. The ones without DRM go across with Send to Kindle, and your notes export as text.","content_html":"<p>Most of the books you bought from Kobo won't open on a Kindle. Kobo's help says that <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019527954-Download-books-from-your-Kobo-account-to-export-to-another-device-or-app\">most of its books are protected by DRM</a>, and Amazon <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200735130\">tells publishers</a> that a Kindle recognises a file copied to it only if the file is free of DRM. What does cross is everything without it: the Kobo purchases that were sold unprotected, and any unprotected EPUB that you put on the Kobo yourself. The protected books aren't gone: your Kobo account keeps them, and they still open on the old Kobo, in Kobo's apps, and in a browser where the book supports it.</p>\n<table><thead><tr><th>What you have</th><th>What happens to it</th></tr></thead><tbody>\n<tr><td>A Kobo purchase with DRM</td><td>Stays readable on the Kobo and in Kobo's apps. It doesn't open on a Kindle.</td></tr>\n<tr><td>A Kobo purchase without DRM</td><td>Downloads from kobo.com as an EPUB or a PDF, and reaches the Kindle through Send to Kindle.</td></tr>\n<tr><td>A file you copied to the Kobo</td><td>Isn't in your Kobo account. If it has no DRM, it goes the same way. If you've lost the original, copy it back from the Kobo first.</td></tr>\n<tr><td>A Kobo Plus book</td><td>Goes when the subscription ends.</td></tr>\n<tr><td>Highlights and notes</td><td>Export as a document, for books from Kobo's shop.</td></tr>\n<tr><td>A library loan</td><td>Reaches a Kindle through Libby, in the United States only.</td></tr>\n</tbody></table>\n<h2 id=\"find-out-which-kobo-books-have-drm\">Find out which Kobo books have DRM</h2>\n<p>The way to tell is to download the book and look at what arrives. Sign in at kobo.com, open “My Account” and then “My Books”, click the Actions icon beside the book, and choose “Download” and “Download file”. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019527954-Download-books-from-your-Kobo-account-to-export-to-another-device-or-app\">download page</a> says you can download most of what you've bought this way.</p>\n<p>If the file ends in <code>.acsm</code>, the book is protected. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">page on file formats</a> says so in as many words, and adds that a book with Adobe's DRM goes onto an e-reader through Adobe Digital Editions. The ACSM file \"does not contain the eBook itself\", as <a rel=\"external\" href=\"https://www.fileinfo.com/extension/acsm\">FileInfo</a> puts it. It's <a href=\"/blog/what-is-an-acsm-file/\">a download ticket</a>: open it in Adobe's program and the protected book downloads. The result opens on a Kobo, which supports Adobe's DRM, and not on a Kindle. Whether you may break the lock yourself is a question of law, and <a href=\"/blog/is-removing-drm-legal/\">it has its own post</a>.</p>\n<p>If an <code>.epub</code> or a <code>.pdf</code> arrives instead, the book is probably unprotected, but the ending isn't proof: Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017814074-Add-eBooks-with-Adobe-Digital-Editions\">page on Adobe Digital Editions</a> says only that a protected book \"usually\" downloads as a file that ends in <code>.acsm</code>. Sending the file settles it, because a protected file doesn't get through Send to Kindle (see below). Kobo's download page says you can copy an unprotected EPUB or PDF from your computer to another maker's device, and warns that not every maker supports EPUB. A Kindle doesn't take an EPUB over the cable, only through Send to Kindle.</p>\n<h2 id=\"where-the-protected-books-still-open\">Where the protected books still open</h2>\n<p>Kobo's help <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017958774-How-do-I-read-my-eBook-or-listen-to-an-audiobook\">lists</a> the Kobo e-reader, the Kobo Books app for a phone or tablet, and Kobo Desktop for a PC or a Mac. The app <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017707914-Get-the-Kobo-Books-app\">needs</a> Android 10 or iOS 18, or later. There is also <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/4408228476567-About-Kobo-Web-Reader\">Kobo Web Reader</a>, for Chrome, Edge, Firefox or Safari. Once you've signed in to kobo.com, a book that supports it shows a “Read Now” button under “My Books”.</p>\n<h2 id=\"send-the-unprotected-books-to-the-kindle\">Send the unprotected books to the Kindle</h2>\n<p>An EPUB without DRM goes to a Kindle over the air. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G5WYD9SAF7PGXRNA\">Send to Kindle page</a> has EPUB and PDF on its list of supported file types. Four of the routes it names take a file: a web page that takes documents of 200 MB or smaller, an email address, the Kindle app for iOS and Android, and desktop apps for Mac and Windows. The document goes into your Kindle library, for your Kindle devices and the Kindle app.</p>\n<p>Amazon converts the EPUB on the way. When the service began to take EPUB in 2022, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2022/04/29/kindle-ereaders-now-support-epubs-sent-by-email/\">The eBook Reader</a> sent a DRM-free EPUB to a Kindle Paperwhite, found that it worked, and noted that Amazon turns such a file into its own AZW3 format before it reaches the Kindle. The same article warned of protected files: \"It's not going to work with Adobe ePubs or any protected files.\" The post on <a href=\"/blog/send-epub-to-kindle/\">sending an EPUB to a Kindle</a> has the limits of each route.</p>\n<p>A cable is the harder way round. Standard Ebooks' <a rel=\"external\" href=\"https://standardebooks.org/help/how-to-use-our-ebooks\">Kindle FAQ</a> says it's \"still not possible to use a USB cable to transfer an epub file\" to a Kindle, so the book has to be converted before you copy it. Which format to convert it to <a href=\"/blog/mobi-azw3-or-epub-for-kindle-usb/\">depends on the model of Kindle</a>.</p>\n<p>Books that you loaded onto the Kobo from a computer aren't in your Kobo account. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024775093-Add-non-protected-PDF-and-ePub-files-to-your-Kobo-eReader-using-your-computer\">page on adding files by cable</a> says such a book appears only on the device you copied it to, so there's no copy to download. Send the original files if you still have them. If you don't, plug the Kobo in, tap “Connect” when it shows “Computer detected”, and copy the files back from the drive named KOBOeReader.</p>\n<h2 id=\"take-your-highlights-and-notes\">Take your highlights and notes</h2>\n<p>Kobo's website <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/29991333812631-Export-annotations-from-your-books\">exports annotations</a> one book at a time. Sync the Kobo first: the page says that annotations reach Kobo's cloud when the e-reader or the app syncs. Under “My Books”, the Actions icon beside a book has “Export Annotations”. You pick PDF, HTML, TXT or Markdown under “File Format” and click “Generate File”, and the result waits under the Annotations tab with a “Download” button. Kobo keeps it for 30 days. The page sets three limits. The export covers only books bought from Kobo or read through Kobo Plus, it may not work for every book because it depends on the author's approval, and it leaves out markups.</p>\n<p>That leaves the books you loaded yourself. Readwyn, this blog's ebook library app for Mac and Linux, reads their highlights from the Kobo over the cable: if the book is in your Readwyn library, a “Device → Readwyn” sync copies its highlights and notes into the Notebook, and “Export” saves them as a Markdown or a plain-text file (<a href=\"/blog/export-kobo-highlights-on-mac/\">the steps</a>).</p>\n<p>Neither export puts the marks back into the book on the Kindle. A highlight is <a href=\"/blog/move-highlights-between-reading-apps/\">an address in one particular file</a>, and the copy that Send to Kindle delivers is a different file.</p>\n<h2 id=\"kobo-plus-and-library-books\">Kobo Plus and library books</h2>\n<p>Kobo Plus is a subscription, so there is nothing to move. Of cancelling, Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360018976173-Cancel-or-renew-your-Kobo-Plus-subscription\">help</a> says: \"You'll no longer have access to your Kobo Plus books.\" For a subscription bought from Kobo itself, the cancel button is on kobo.com under “My Account” and “My Subscriptions”, where “Cancel this subscription” is followed by “Yes, cancel subscription”. The page doesn't say whether the annotation export still works afterwards, so export first.</p>\n<p>Library books depend on where you live. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017677053-Borrow-eBooks-from-the-public-library-using-your-Kobo-eReader\">borrowing page</a> lists 13 countries and territories, from Canada to Thailand, where the Kobo models it names borrow from the public library on the device itself. A Kindle gets library loans through the Libby app, which hands them to Amazon to deliver, and Libby's help calls that route <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6017.htm\">\"available in the U.S. only\"</a>. The steps for both are in <a href=\"/blog/library-ebooks-on-kobo-and-kindle/\">how to read library ebooks on a Kobo or a Kindle</a>.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/organise-kindle-with-collections/","url":"https://readwyn.com/blog/organise-kindle-with-collections/","title":"How to organise books on a Kindle with collections","summary":"On a Kindle, open the library, tap the Menu icon and choose Create New Collection. A book can be in several collections, and they sync through your Amazon account.","content_html":"<p>Collections are how you group books on a Kindle: you make and name each one yourself, and one book can belong to several. On a current Kindle you open the library, tap the Menu icon, choose Create New Collection, type a name and tick the books that go in it. <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">Amazon's user guide for the Kindle Colorsoft</a> gives those steps, and says collections sync to the other devices and reading apps that are registered to the same Amazon account and support them. Readwyn, the ebook library app for Mac and Linux that this blog belongs to, reads a Kindle only when the computer mounts it as a drive: <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page on USB transfers</a> says the Kindle Scribe and the Kindles released in 2024 need a separate application on a Mac, and that models released before 2024 (the Scribe excepted) don't. Of the Kindles that do mount, only an old one, on firmware 2.5 to 4, keeps its collections in a file on that drive, where Readwyn can read and write them.</p>\n<h2 id=\"making-a-collection-on-the-kindle\">Making a collection on the Kindle</h2>\n<p>The Colorsoft guide lists five steps:</p>\n<ol>\n<li>On the Home screen, tap Your Library at the bottom right.</li>\n<li>On the library page, tap the Menu icon.</li>\n<li>Select Create New Collection.</li>\n<li>Type a name for the collection and tap OK.</li>\n<li>Tick the checkbox beside each item that belongs in it, then tap Done.</li>\n</ol>\n<p>You can change a collection afterwards. The Menu icon on its cover has Add/Remove Items, Rename and Delete. From the other side, the Menu icon on a single book's cover (or a press and hold on the cover) opens a list of options that, the guide says, may include Add to Collection.</p>\n<p>A collection takes \"as many items as you like\", books and personal documents both, and the same item can sit in more than one. Audiobooks are the exception: \"Audible content cannot be added to collections\", although the Kindle book of the same title can.</p>\n<p>Deleting a collection is safe for the books in it. The guide says it \"does not remove the content stored on your device or in the Cloud\". The grouping does go: a deleted collection also disappears from the Cloud and from your other devices and apps.</p>\n<h2 id=\"where-collections-show-up\">Where collections show up</h2>\n<p>The icon at the top right of the library changes how it looks, with two choices to make: All or Collections, and Grid or List. All shows every item on its own, and Collections shows your content in the groups you made.</p>\n<p>Whether collections also appear among the books in the ordinary views is a setting. The guide puts it at Settings, Home and Library, the Library tab, Collections, with three options: show collections only in Collections view, show your favourites in the library, or show all of them there. To make a collection a favourite, tap the Menu icon on its cover and choose \"mark collection as favorite\". The guide adds a note there: \"this setting is device-specific\", and it isn't saved when you deregister the Kindle.</p>\n<p>The library's filter menu has a Collections entry too, next to Books, Comics, Samples, Periodicals and Docs.</p>\n<h2 id=\"collections-made-somewhere-else\">Collections made somewhere else</h2>\n<p>You don't have to type on the Kindle. Amazon's help page <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=T4KX4ALDBNr9flMGqC\">What Are Collections?</a> says you can create them in Kindle for Windows \"and access them on other supported devices\". For Amazon's website, <a rel=\"external\" href=\"https://www.pcworld.com/article/399037/how-to-organize-your-kindle-library.html\">PCWorld's 2020 guide</a> gives the route: Your Content and Devices in your account, the Show drop-down set to Collections, then \"Create new collection\".</p>\n<p>A collection made elsewhere shows on the Kindle: the Colorsoft guide says collections on your cloud account \"will always display in Collections view\". Its books are a separate download. Amazon's page <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GQNG5LCNAE6JYM45\">Download Collections to Your Kindle E-Reader</a> says collections created on another registered device or on the website \"don't automatically download to your other devices\". The page gives three steps: open the library; select what it calls the Sort By menu, in the upper right-hand corner, and choose Collections; then tap the Menu icon on the collection you want and choose Download all items. In the Colorsoft guide, the Collections choice is under the icon at the top right of the library, not among the sort options.</p>\n<h2 id=\"shelves-from-a-library-on-your-computer\">Shelves from a library on your computer</h2>\n<p>If you keep your books in a library app, its shelves look like ready-made collections. Whether they can become collections depends on the age of the Kindle.</p>\n<p>A Kindle with firmware 5 or later (the Kindle Touch, the Paperwhite, the Voyage and the Oasis among them) keeps its collections in a database of its own, outside the storage that a USB cable shows. Connect such a Kindle, and Readwyn's <a href=\"/blog/sync-kindle-with-mac-without-calibre/\">Sync screen</a> lists “Shelves as collections” under “Not available on this Kindle”, with the reason: “Kindle firmware 5 and later keeps collections with Amazon, not in a file on its storage.”</p>\n<p>Collections came to the Kindle 2 and the Kindle DX with <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2010/06/19/kindle-2-and-dx-firmware-update-folders-finally/\">software 2.5 in June 2010</a>, and a Kindle on firmware 2.5 to 4 keeps them in one file on its drive, <code>system/collections.json</code>. Readwyn reads that file and brings the collections into the library as <a href=\"/blog/reading-status-and-shelves/\">shelves</a>, for the books that are in both places. Writing shelves back as collections is tagged “Experimental”, and Readwyn tells you to restart the Kindle from its Settings menu to see them, <a href=\"/blog/eject-kindle-or-kobo-from-mac/\">once you've ejected it</a>.</p>\n<h2 id=\"sorting-filters-and-series\">Sorting, filters and series</h2>\n<p>Collections are the part you build by hand. Sorting and filtering need no preparation.</p>\n<p>You can sort by Most recent, Title, Author or Publication date, and Most recent is the default: new items and the book you're reading sit at the top. The filter menu narrows the list to Read or Unread items (the guide says \"Books you finish are automatically marked as read\"), to one type of content, or to what's downloaded.</p>\n<p>Series are grouped as well. The setting came with firmware 5.13.4 in December 2020, under the name Group Series in Library, and <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2020/12/16/kindles-now-have-group-series-in-library-option-in-settings\">The eBook Reader's report of the update</a> described a series as the cover of its first book with a number in the corner. In the Colorsoft guide it's Group series, on the Library tab of the Home and Library settings. That guide's section on comic books says a comics series is shown by \"the cover of the book you most recently interacted with\", with an icon in the lower left corner for how many of its books you own or have downloaded. The report also noted a limit that matters if you load your own files: at the time, the grouping \"does not appear to work with sideloaded books, only books purchased from Amazon\".</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/put-epub-on-kindle-from-mac-usb/","url":"https://readwyn.com/blog/put-epub-on-kindle-from-mac-usb/","title":"How to put an EPUB on a Kindle from a Mac with a USB cable","summary":"A Kindle won’t open an EPUB copied over USB. Readwyn’s “Send to device” converts the book to MOBI and writes it to the Kindle’s documents folder.","content_html":"<p>A Kindle won’t open an EPUB that you copy to it over a cable: <a rel=\"external\" href=\"https://blog.vellum.pub/2022/05/send-to-kindle-now-accepts-epub-files/\">Vellum’s note on sideloading</a> says to use a MOBI file for USB. So the book has to be converted first and then put in the Kindle’s <code>documents</code> folder. Readwyn does both steps at once: plug the Kindle in, open the book, and choose “Send to device”. That works for a Kindle that shows up on the Mac as a disk, which is the only way Readwyn reads one. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon’s help page on USB transfers</a> says the Kindle Scribe and the Kindles released in 2024 need a separate application on a Mac, and that models released earlier than 2024 (the Scribe aside) don’t. Readwyn works with the earlier ones.</p>\n<h2 id=\"send-the-book\">Send the book</h2>\n<ol>\n<li>Connect the Kindle to the Mac with its USB cable. A card named “Kindle” appears in Readwyn’s sidebar. Connecting only reads the Kindle. In the Mac App Store version, the Sync screen asks once, the first time the Kindle connects: “Give Readwyn access to Kindle”. Click “Select Kindle…”, then “Allow” in the Finder dialog, which opens at the Kindle.</li>\n<li>Open the book from your library. If the EPUB isn’t there yet, <a href=\"/blog/add-books-by-dragging/\">drag the file from Finder onto the Readwyn window</a> first.</li>\n<li>On the book screen, click “…” (“More actions”) and then “Send to device”. The line under it names where the book goes: “Kindle”.</li>\n</ol>\n<p>A book that came into the library as a MOBI, an AZW3, or a PDF goes the same way, because Readwyn sends the EPUB 3 that it made from the file at import. Two kinds of book stay behind, each with its reason on screen. A comic gets “Not copied: a Kindle does not read comics”. A file of those formats that didn’t convert at import, such as one with DRM or a scanned PDF without text, gets “Not copied: the book has no EPUB file”.</p>\n<h2 id=\"what-lands-on-the-kindle\">What lands on the Kindle</h2>\n<p>One file: <code>documents/&lt;title&gt; - &lt;first author&gt;.mobi</code>, or <code>documents/&lt;title&gt;.mobi</code> for a book without an author. The name keeps the letters of every script, so a Tamil title stays in Tamil, and a character that the FAT file system refuses becomes <code>_</code>.</p>\n<p>The file is a MOBI 6 book. Readwyn writes that older format because every Kindle reads it. The newer KF8, the format inside an AZW3 file, doesn’t open on the first Kindle, the Kindle 2, or the Kindle DX, according to the <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/KF8\">MobileRead wiki</a>. On the Kindle 3, which Wikipedia’s <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Amazon_Kindle\">table of Kindle models</a> lists as the Kindle Keyboard, it needs a software update, and on the Kindle 4 it needs firmware 4.1.0 or later. If you’d sooner copy a file by hand, there’s a separate post on <a href=\"/blog/mobi-azw3-or-epub-for-kindle-usb/\">which format to put on a Kindle over USB</a>.</p>\n<p>The text arrives character for character. Paragraphs, headings, lists, tables, block quotes, bold, italics, links between chapters, and the cover all carry over. MOBI 6 has no place for much else, so embedded fonts go, and so does most of the CSS apart from alignment, weight, italics, underline, and strike-through. A JPEG, GIF, or PNG of up to 127 KB is copied byte for byte. A larger image, or one in another format such as WebP, becomes a JPEG of at most 127 KB and 1,600 pixels on its longest side. An SVG image is dropped. The EPUB in your library isn’t touched.</p>\n<p>On ten test EPUBs, the conversion took 14 to 300 ms a book, and the 300 ms was a 2.9 MB Complete Works of Shakespeare.</p>\n<p>Each chapter of the MOBI has the same text as its chapter in the EPUB. A highlight that you later make on the Kindle therefore comes back to the exact place in the library’s EPUB <a href=\"/blog/kindle-highlights-in-ebook-library/\">when you sync</a>.</p>\n<p>The first sync or send with a Kindle also leaves one small file on it, <code>.readwyn/device-id</code>, so that Readwyn knows this Kindle the next time. It never shows up as a book.</p>\n<h2 id=\"after-the-copy\">After the copy</h2>\n<p>When the copy is done, Readwyn says “Copied to the device. Eject the device so it adds the book.” The button for that, “Eject device”, is on the Sync screen. If Readwyn is still writing when you click it, an alert says “Readwyn is still writing to the Kindle”. Its “Stop and eject” button waits for the write to end before it ejects (<a href=\"/blog/eject-kindle-or-kobo-from-mac/\">more on ejecting a Kindle</a>).</p>\n<p>For a book that’s already there, the menu item reads “Send to device again”. If the file on the Kindle is the same, byte for byte, nothing is written and the message is “The book is already on the device.”</p>\n<p>To take the book off again, open its Details tab while the Kindle is connected. The row “On the device” has the button “Remove from device”, which deletes the book and the files the Kindle kept beside it, such as its <code>.sdr</code> folder. The book stays in your library.</p>\n<h2 id=\"if-your-kindle-doesn-t-appear-as-a-disk\">If your Kindle doesn’t appear as a disk</h2>\n<p>For an older Kindle that stays out of Finder, the things to check are the cable, a macOS alert, and a Finder setting: <a href=\"/blog/kindle-not-showing-up-on-mac/\">why a Kindle doesn’t show up on a Mac</a>. The Scribe and the 2024 models are a different case. They talk to a computer over MTP instead of acting as a USB disk, and a Mac doesn’t support MTP on its own, as <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2024/11/01/how-to-connect-kindles-to-macs-to-transfer-connect-via-usb/\">The eBook Reader explained</a> in November 2024. Readwyn can’t see such a Kindle. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon’s page</a>, as of November 2026, names no model later than 2024. If a newer Kindle never shows up in Finder, the same goes for it.</p>\n<p>Amazon’s own route is its Send to Kindle app for Mac. The same page describes the app’s USB File Manager, where you drag files into the Kindle’s “Documents” folder, and says it works on Intel and M1 to M3 Macs with macOS 12 or later. It’s still a cable copy, so an EPUB that goes in this way won’t open: Standard Ebooks’ <a rel=\"external\" href=\"https://standardebooks.org/help/how-to-use-our-ebooks\">Kindle FAQ</a> says Kindles “still do not natively support epub files”. The wireless side of Send to Kindle is the route for an EPUB on these models, since Amazon converts the file on the way: <a href=\"/blog/send-epub-to-kindle/\">what it accepts and what its limits are</a>.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/send-pdf-to-kindle/","url":"https://readwyn.com/blog/send-pdf-to-kindle/","title":"How to send a PDF to a Kindle, and how to make its text bigger","summary":"Send to Kindle takes a PDF by email, from a web page, or from Amazon's Mac app, and a USB cable works too. For bigger text, have the PDF converted or read it sideways.","content_html":"<p>Amazon's Send to Kindle service takes a PDF as an email attachment, through a web page, or from its Mac app, and a USB cable works too. Sent as it is, a PDF arrives as fixed pages that the Kindle shrinks to fit its screen, so there's no font size to turn up. For bigger text, you either turn the PDF into a book that reflows, which Amazon does when the subject of your email is “Convert”, or you keep the pages and read them sideways. Readwyn, the ebook library app this blog belongs to, is one more way to convert. It reaches a Kindle only as a mounted disk, and on a Mac, according to <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page on USB transfers</a>, that means the models released before 2024, the Kindle Scribe aside.</p>\n<h2 id=\"ways-to-send-it\">Ways to send it</h2>\n<p><strong>By email.</strong> Each Kindle on your account has its own Send to Kindle address, and <a rel=\"external\" href=\"https://www.makeuseof.com/kindle-hidden-email-address-delivers-any-document/\">MakeUseOf notes</a> that each Kindle app gets one too. <a rel=\"external\" href=\"https://www.amazon.com/sendtokindle/email\">Amazon's page for the email route</a> says where to find it: Manage Your Content &amp; Devices &gt; Preferences &gt; Personal Document Settings. The same settings hold the Approved Personal Document E-mail List, and only an address on that list can send. Attach the PDF and send it. The <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G7NECT4B4ZWHQ8WV\">limits</a> are 25 attachments in one email and 50 MB in total.</p>\n<p><strong>From a browser.</strong> Sign in at <a rel=\"external\" href=\"https://www.amazon.com/sendtokindle\">amazon.com/sendtokindle</a> and upload the file. The page takes up to 200 MB. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G5WYD9SAF7PGXRNA\">overview of the service</a> says to turn on “Add to your library” there, so that every device on the account can download the document. The same overview names one more route, for a PDF that is on a phone or a tablet: the free Kindle app for iOS and Android sends documents to your library too.</p>\n<p><strong>From Finder.</strong> Amazon's <a rel=\"external\" href=\"https://www.amazon.com/sendtokindle/mac\">Mac page</a> lists two apps. With the Send to Kindle app for Mac installed, right-click the PDF in Finder and choose “Send to Kindle”, or drop the file on the app. With the Kindle app for Mac, which the page recommends, the command is under Share in the same menu, once you have ticked “Send to Kindle” in System Settings &gt; Privacy &amp; Security &gt; Extensions &gt; Sharing.</p>\n<p><strong>Over a cable.</strong> Copy the PDF into the Kindle's <code>documents</code> folder. Nothing has to be converted first: Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_all_new_paperwhite_v2/v1/en-US/html/kug.html\">user guide for the 10th-generation Paperwhite</a> says a Kindle reads PDF “natively”. On a Mac, a Scribe or a 2024 Kindle needs a separate application for the copy. Amazon's own is the same Send to Kindle app, which has a <a href=\"/blog/kindle-not-showing-up-on-mac/\">USB File Manager</a> under its Tools menu.</p>\n<p>A PDF with a password won't go through the service at all. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=T48rsVm3gY7KeGkKUk\">list of Send to Kindle errors</a> gives the cure: save a copy without the security settings and send that.</p>\n<p>A Scribe owner who wants to write on the pages has one more thing to check before choosing a route: the software. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TNtSm13k3txFI4EvDV\">page of note types</a> still says, in November 2026, that “Handwriting isn't supported in PDF documents loaded via USB-C”. The release notes of software update 5.19.3 say otherwise. <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2026/03/27/kindle-software-update-5-19-3-released-with-improved-pdf-support/\">The eBook Reader quoted them in March 2026</a>: a PDF that came over the cable now gets what the notes call the “latest reading features”, and on a Scribe “you can write directly on your PDFs”. Its reviewer met two catches. Most of the PDFs that he had copied before the update wouldn't open, though the ones he deleted and copied again then did. And annotations on a copied PDF “are not transferable to other devices/software”. <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2026/04/07/bug-fix-update-released-for-kindle-ereaders-version-5-19-3-0-1/\">Version 5.19.3.0.1 replaced that update in April</a> with the same notes, so look for that number or a later one. On older software, send the PDF: the same page of Amazon's says that a Scribe can write directly on the page of a PDF, unless it came over USB-C or was sent before 11 November 2022.</p>\n<h2 id=\"let-amazon-reflow-it\">Let Amazon reflow it</h2>\n<p>Put the word “Convert” in the subject line of the email, and Amazon's servers reflow the text of the PDF instead of delivering its pages. Saikat Basu described the result in <a rel=\"external\" href=\"https://www.makeuseof.com/kindle-hidden-email-address-delivers-any-document/\">MakeUseOf in May 2026</a>: “resizable text, adjustable fonts”, where the plain PDF is a shrunken page that you pinch to zoom. He also reported “mixed results”. PDFs set in several columns sometimes came out scrambled, and heavy charts and tables lost their formatting.</p>\n<p>Scans are out. The eBook Reader's <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2017/03/23/heres-how-to-use-amazons-kindle-pdf-reflow-feature/\">guide to the feature</a>, written in 2017, says reflow “certainly isn't going to work for scanned PDFs”. It adds that without the word in the subject, the PDF is sent as it is. For a document where the text matters and the figures do too, <a rel=\"external\" href=\"https://www.techlicious.com/tip/turn-pdfs-into-epub-ebooks-for-your-kindle/\">Techlicious suggests</a> sending it twice, once with the word and once without, so that the original pages sit beside the readable copy.</p>\n<p>Amazon's <a rel=\"external\" href=\"https://www.amazon.com/sendtokindle/email\">step-by-step page for the email route</a>, as it stands in November 2026, doesn't mention the word. It says “You don't need to include a subject line.” The table on Amazon's page of note types does list “PDF (adjustable layout)” among the documents that come in through Send to Kindle, with highlights and text notes on a Kindle e-reader.</p>\n<h2 id=\"convert-it-before-you-send-it\">Convert it before you send it</h2>\n<p>The other way to get a reflowing book is to make it yourself and send that. Readwyn converts a PDF that has a text layer to EPUB 3 as you add it to your library. With a Kindle that mounts as a disk, “Send to device” then <a href=\"/blog/put-epub-on-kindle-from-mac-usb/\">writes the book to the <code>documents</code> folder as a MOBI</a>. For a Scribe or a 2024 model, open the book's Details tab: “Files” lists the EPUB as “For reading”, with a “Show in Finder” button beside it. Upload that file from there, because <a href=\"/blog/send-epub-to-kindle/\">Send to Kindle accepts an EPUB</a>. A scan doesn't convert here either, since there's no OCR. A separate post has <a href=\"/blog/convert-pdf-to-epub-without-calibre/\">what the conversion does, and the results on 50 test PDFs</a>.</p>\n<h2 id=\"when-the-pdf-has-to-stay-a-pdf\">When the PDF has to stay a PDF</h2>\n<p>A scan, or a page full of tables, has to be read as it was laid out. What you can change is how the Kindle shows it. In its <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2021/11/11/kindle-paperwhite-5-pdf-review-video/\">PDF review of the 2021 Paperwhite</a>, The eBook Reader says that turning the screen on its side is “a good way to minimize large margins and increase the text size”, with one catch: the screen rotates in one direction only. In portrait, a double tap zooms in, and you can pinch to zoom and then scroll around the page. The review says the double tap gets confused when the formatting changes from page to page, so it suits PDFs of plain text best.</p>\n<p>Faint print has its own fix: the text adjustment window has a contrast setting that makes the text darker (images darken with it). Bookmarks, the dictionary, search, notes and highlights all work in a PDF, and so does Dark Mode, which turns the page to white text on black. The review's verdict on that 6.8-inch screen is that Kindles are limited for PDFs, but “if your needs are basic they can get the job done”.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/set-up-kindle-for-a-child/","url":"https://readwyn.com/blog/set-up-kindle-for-a-child/","title":"How to set up a Kindle for a child","summary":"Create a child profile in Amazon Kids on the Kindle, set a parental controls PIN and choose the books. The steps, what a profile blocks, and the setup for a Kindle Scribe.","content_html":"<p>Create a child profile on the Kindle. From its Home screen, open the Menu, choose Parental Controls, then Amazon Kids, and add a profile with the child's name and birthday. On the way the Kindle has you set a parental controls PIN and pick which of your books the child may read. After that the child reads inside the profile, where the Kindle Store, the web browser, Goodreads and Wikipedia are blocked, and only the PIN leads back out. A profile comes \"at no additional cost\", Amazon's help page <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GHRVRLACAYKJXF7Y\">What Is Amazon Kids?</a> says, and you don't need one of the Kindle Kids models for it: <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/amazon-kindle-kids-setup-questions-answered\">Amazon's answers for parents</a>, published in March 2026 and updated that September, say that a Kindle, a Kindle Paperwhite or a Kindle Colorsoft gets the same parental controls. The Kindle Scribe is the exception: Amazon says it doesn't support child profiles for now, and what it has instead is a PIN and four locks.</p>\n<h2 id=\"the-steps-on-a-kindle-you-already-own\">The steps on a Kindle you already own</h2>\n<p>Amazon's help page <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G27XUUGLQ5TG84XC\">Create a Child Profile on Your Kindle E-Reader</a> gives six:</p>\n<ol>\n<li>From the Home screen, select the Menu icon.</li>\n<li>Select Parental Controls, then Amazon Kids.</li>\n<li>Select Add New Profile or Add Kid Profile.</li>\n<li>Enter your parental controls PIN if the Kindle asks for it, then the child's name and birthday.</li>\n<li>Select Next, and add the books from your own library that the child may read.</li>\n<li>Review the profile and its reading settings, then select Done.</li>\n</ol>\n<p>If your menus don't match those steps, Amazon describes two other routes to the same place. Its answers for parents start from the settings: swipe down from Home for Quick Actions, then Settings, Your Account, Amazon Household. The <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_11th/v1/en-US/html/kug.html\">user guide of the 11th-generation Kindle</a> puts Amazon Kids straight into the Home menu with a Get Started button, calls the household Amazon Family, and records the feature's earlier names, Amazon FreeTime and (in the UK) Amazon Fire for Kids. The guide adds that the Kindle prompts you to set a Parental Controls password (the PIN) if you don't have one yet. Both sources put the limit at four child profiles.</p>\n<p>To hand the Kindle over, tap Start reading next to the child's name. To take it back, open the Menu, select Exit Amazon Kids, and enter the PIN.</p>\n<h2 id=\"if-you-bought-a-kindle-kids\">If you bought a Kindle Kids</h2>\n<p>The Kindle Kids, the Kindle Paperwhite Kids and the Kindle Colorsoft Kids make the profile during their first start. The user guides of the Kindle and of the <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">Kindle Colorsoft</a> list two setup steps for the Kids models only, after you register the Kindle to your Amazon account: activate the Amazon Kids+ subscription, then create child profiles and set the parental control PIN. Amazon's answers for parents add that web access starts blocked and the Store starts switched off.</p>\n<p>A Kids model adds a case and six or 12 months of Amazon Kids+, the subscription library of children's books and audiobooks. After the included months, Amazon's answers for parents give the price in dollars only: $7.99 a month for Prime members and $8.99 for everyone else.</p>\n<h2 id=\"what-a-profile-blocks-and-what-it-adds\">What a profile blocks, and what it adds</h2>\n<p>The user guide's list of blocked places is short: the Kindle Store, the web browser, Goodreads and Wikipedia. Search leaves out results from the Store and from Wikipedia. While the profile is active a lock icon sits at the top of the screen, and the settings that deregister the Kindle, reset it or open the Family Library are disabled. Your own library, reading history, notes and highlights stay out of the child's view, according to Amazon's answers for parents.</p>\n<p>Next to each child's name, a settings icon opens the options of the profile, among them Add Books, Store Access, Set Daily Goals, Set Bedtime, View Progress, Edit Profile and Delete Profile. Store Access lets the child browse and ask for a book, and the request comes to the main email address of your Amazon account for approval. Delete Profile is permanent, and it takes the child's reading statistics and achievements with it.</p>\n<p>Goals, bedtime and books can also be set away from the Kindle, in the <a rel=\"external\" href=\"https://parents.amazon.com/\">Parent Dashboard</a>, a free app and website. The help page <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GHRVRLACAYKJXF7Y\">What Is Amazon Kids?</a> carries one warning: a child profile that goes unused for more than 18 months is removed automatically.</p>\n<h2 id=\"putting-books-into-the-profile\">Putting books into the profile</h2>\n<p>Without Amazon Kids+, you build the child's library yourself, \"by purchasing or borrowing books individually\" in the words of Amazon's answers for parents. For a book that you bought, the article names three places to share it: Household Sharing, under the three-dot menu at the top right of Home on the Kindle; Manage Your Content and Devices on the web, under Settings, then Household and Family Library; and the Content menu of the Parent Dashboard. Shared books turn up under \"Added for You\" on the child's Home screen the next time the profile is online. Amazon's page <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G84RX2PWJ3WAU76K\">Add Content to a Child Profile</a> says that <a href=\"/blog/library-ebooks-on-kobo-and-kindle/\">public library books</a> can be added to a profile with Amazon Kids+, and that a borrowed one leaves the profile when its lending period ends or when you return it.</p>\n<p>Your own files can't be shared. The page lists \"Personal documents\" first among the things a profile can't hold. That's Amazon's name for the files you bring yourself, such as a <a href=\"/blog/what-is-project-gutenberg/\">Project Gutenberg</a> EPUB that went through <a href=\"/blog/send-epub-to-kindle/\">Send to Kindle</a>. They stay in your own library.</p>\n<h2 id=\"the-kindle-scribe-and-a-kindle-without-a-profile\">The Kindle Scribe, and a Kindle without a profile</h2>\n<p>The Scribe \"doesn't support Child Profiles at this time\", Amazon's help page on creating one says, and the user guide of the 11th-generation Kindle warns that Amazon Kids isn't available in some countries. The <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_scribe/v1/en-US/html/kug.html\">Scribe's own guide</a> mentions a child's profile under Amazon Family, but it has no Amazon Kids section. What it describes, in the same words as the 11th-generation guide, is a set of four locks. Swipe down for Quick Actions and tap All Settings (or open the Menu, which the Scribe calls More, and choose Settings), then Your Account, then Parental control. Under Restrictions you can block the web browser, the Kindle and Audible stores, Cloud content and Goodreads, each on its own, and the first one you switch on has you create the PIN.</p>\n<p>That's a lighter setup than a profile. The child reads in your own library, the one that holds your personal documents. None of a profile's options comes with it: no daily goal, no bedtime, no progress view. The Parent Dashboard can't manage these locks either, Amazon's article adds.</p>\n<p>Write the PIN down either way. Both guides give the same remedy for a forgotten one: deregister the Kindle from your account on Amazon's website, or contact Kindle Customer Service. Amazon's page on <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GV3ESZM3CABSBR87\">resetting the parental controls password</a>, for a Kindle of the <a href=\"/blog/which-kindle-do-i-have/\">6th generation</a> or newer, gives a third way, a reset of the whole device. Deregistering and the reset each empty the Kindle: the guides say that deregistering removes all content from the device, and the help page says that the reset removes the downloaded content along with the PIN. Books bought from Amazon stay in the cloud and can be downloaded again.</p>\n<h2 id=\"two-settings-for-a-new-reader\">Two settings for a new reader</h2>\n<p>Word Wise prints short definitions above the harder words of a book. The user guide says it helps \"kids reading their first chapter books\", and it's switched on in the Aa menu, under the More tab, for the English-language books that support it. The Font tab of that menu holds the fonts, OpenDyslexic among them, though <a href=\"/blog/do-dyslexia-fonts-work/\">the studies of dyslexia fonts</a> found that wider letter spacing helped where the letter shapes did not.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/sideloaded-book-not-showing-on-kindle/","url":"https://readwyn.com/blog/sideloaded-book-not-showing-on-kindle/","title":"Why a sideloaded book does not show up on your Kindle","summary":"A Kindle lists a file copied over USB only if it's in a format it reads (not EPUB), has no DRM, and sits in the documents folder. Then eject, and clear the Library's filters.","content_html":"<p>A Kindle lists a book that came over a cable only if three things are true of the file: it's in a format the Kindle reads, which rules out EPUB; it has no DRM; and it sits in the Kindle's <code>documents</code> folder. Even then the book appears only after the Kindle has been ejected from the computer. If all of that holds and the book still isn't in the Library, a filter may be hiding it. A book that was there last month and has since gone is a different case: owners have reported Kindles that deleted their sideloaded books after a long spell in airplane mode.</p>\n<h2 id=\"eject-before-you-look\">Eject before you look</h2>\n<p>Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_all_new_paperwhite_v2/v1/en-US/html/kug.html\">user guide for the 10th-generation Paperwhite</a> says the Kindle “is not usable as a reading device while in USB drive mode”, and it leaves that mode when you eject it. Standard Ebooks ends <a rel=\"external\" href=\"https://standardebooks.org/help/how-to-use-our-ebooks#kindle-method-1\">its cable instructions</a> with the eject, and only then says that the ebook “should now be visible”. So <a href=\"/blog/eject-kindle-or-kobo-from-mac/\">eject the Kindle</a> first and check the Library second.</p>\n<p>Where you eject depends on the model. If the computer shows your Kindle as a drive, eject it on the computer: for a Mac, the Paperwhite guide gives the Eject button beside the Kindle in a Finder window. A Kindle Scribe or a 2024 model needs a separate application on a Mac, as Amazon's current <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">help page on USB transfers</a> says, and there's nothing to eject on the computer: a note on that page says to select the Disconnect button on the Kindle's screen, or to unplug the cable. Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">guide for the Kindle Colorsoft</a>, a 2024 model, has the same sentence about USB drive mode, so these models are no exception.</p>\n<h2 id=\"a-file-the-kindle-reads\">A file the Kindle reads</h2>\n<p>The Paperwhite guide says you can add “Kindle-compatible files” to the <code>documents</code> folder, and an EPUB isn't one. <a rel=\"external\" href=\"https://standardebooks.org/help/how-to-use-our-ebooks#kindle-epub\">The Kindle FAQ of Standard Ebooks</a> has the question “I heard that Kindles support epubs now. Is that true?” Its answer is no: it's “still not possible to use a USB cable to transfer an epub file” to a Kindle. The computer can't tell you this. To it, a connected Kindle is “an external storage drive or volume”, in the guide's words, and a drive doesn't judge the files put on it.</p>\n<p>An EPUB reaches a Kindle through Send to Kindle, where <a href=\"/blog/send-epub-to-kindle/\">Amazon converts it on the way</a>. Amazon's USB page names EPUB only among the file types of that wireless service. The other route is to convert the book yourself, and there's a separate post on <a href=\"/blog/put-epub-on-kindle-from-mac-usb/\">putting an EPUB on a Kindle from a Mac with a USB cable</a>.</p>\n<p>For the cable, the file to copy is an AZW3 or a MOBI. Amazon's guide gives a list where it describes downloads in the Kindle's own browser: Kindle content (.AZW, .AZW1, .AZW3, and .KFX), “unprotected Mobipocket books” (.MOBI, .PRC), and text files. There's no EPUB on it. Standard Ebooks offers AZW3 files for the purpose. <a rel=\"external\" href=\"https://www.gutenberg.org/help/reading_options.html\">Project Gutenberg's reading help</a> points to two downloads, “Kindles (kf8)” and “Older Kindles”, and says to use the Kindle format, “not EPUB”. Which of the two to take depends on the age of the Kindle, and <a href=\"/blog/mobi-azw3-or-epub-for-kindle-usb/\">which format to put on a Kindle over USB</a> goes through the cases. PDF and TXT files need no conversion: Amazon's guide says a Kindle reads both “natively”.</p>\n<p>Two more things can get in the way of a file in a good format. One is DRM. Amazon tells publishers who preview a book on their own device that a <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200735130\">Kindle “will only recognize transferred files”</a> if they're free of DRM, so a file that still has DRM on it won't appear, however you copy it. The other is an archive. If the download arrived as a ZIP, unpack it on the computer and copy the book file itself. Amazon's USB page says that any “renaming, compressing, or extracting of the file” needs to happen on the computer, not on the Kindle's drive.</p>\n<h2 id=\"the-documents-folder\">The documents folder</h2>\n<p>The guide names the place. The Kindle's drive contains a folder called <code>documents</code>, and files are added “to this directory”. Amazon's page for publishers makes it a condition: a transferred file has to be “added to the correct folder type”. A book left at the top of the drive, beside the folder, isn't in it.</p>\n<p>In the USB File Manager of Amazon's Send to Kindle app for the Mac, the folder to drag into is shown as “Documents”. If the Mac doesn't show your Kindle at all, <a href=\"/blog/kindle-not-showing-up-on-mac/\">that's a separate problem</a>. On Windows, Amazon's page adds a limit that a long file name inside nested folders can reach: “the entire file path shouldn't exceed 256 characters”.</p>\n<h2 id=\"a-filter-or-an-unfamiliar-cover\">A filter, or an unfamiliar cover</h2>\n<p>By default the Library shows everything, in the cloud and on the device. Its filter menu can narrow that by read state (Read or Unread) and by content type, and the guide lists the types: Books, Comics, Samples, Periodicals, Docs, and Collections. The guide doesn't say which of those types a file from a computer gets, and it may not be Books. calibre's manual, for one, <a rel=\"external\" href=\"https://manual.calibre-ebook.com/generated/en/ebook-convert.html\">describes an option</a> under which a MOBI file “will be marked as a personal document when sent to the Kindle”. So don't rely on a type filter to find the book. While a filter is active, a toolbar names it, and the X on the filter clears it. Clear every filter before you decide the book isn't there.</p>\n<p>Amazon has a page called <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TsdpGRrbNmNohmj2Q9\">Troubleshoot Kindle Content Not Appearing in Your Library</a>. It's written for store purchases, and its steps about Wi-Fi, syncing, and payment can't help with a file that never went through Amazon. Three of its steps carry over to a book that came by cable, though the page words them for the Kindle app: remove any active filters, search for the title, and sort by Recent “to bring new items to the top”. Its last step is a restart. On the Paperwhite that the guide covers, that means holding the power button for 9 seconds until the Power dialog appears, then tapping Restart.</p>\n<p>Search for the title, not the picture. Standard Ebooks supplies a cover thumbnail to copy after the book itself, and warns that without it “your Kindle will use a default cover”. It blames a bug in the Kindle's software.</p>\n<h2 id=\"a-book-that-was-there-and-has-gone\">A book that was there and has gone</h2>\n<p>Since October 2023, The eBook Reader has reported on <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2023/10/23/kindle-ereaders-deleting-sideloaded-ebooks/\">Kindles that delete sideloaded books</a>. The reports usually describe the same sequence: books go on over USB, the Kindle spends a long stretch in airplane mode, and when Wi-Fi comes back on the books have vanished. Some owners found the books' folders still on the drive with the files gone. “Nobody seems to know what the exact cause is”, Nathan Groezinger wrote in <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2024/02/01/beware-of-airplane-mode-on-kindles-if-you-sideload-books/\">a follow-up in February 2024</a>, adding that it “doesn't appear to be a widespread issue”. His <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2024/12/19/kindle-bugs-list-common-software-issues-and-how-to-fix-them/\">list of Kindle bugs</a> from December 2024 gives the workaround: keep the Kindle in airplane mode all the time, or keep it regularly on Wi-Fi, and avoid the long gap in between. The same list notes that the bug seems to affect only books that went on over USB, so a book that can go through Send to Kindle is out of its reach.</p>\n<p>Keep the file on your computer whichever you choose. Amazon's guide says files transferred by cable “are not stored in the Cloud”, so there's nothing to download again.</p>\n<h2 id=\"if-the-book-went-through-send-to-kindle\">If the book went through Send to Kindle</h2>\n<p>A document sent through Send to Kindle goes missing for other reasons, because it travels by way of your Kindle library at Amazon. Here the store steps do apply: the Kindle needs Wi-Fi, and Amazon's troubleshooting page says to open Settings and select “Sync My Kindle” or “Sync and Check for Items”. On the Send to Kindle web page, <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G5WYD9SAF7PGXRNA\">Amazon's overview of the service</a> says to turn on “Add to your library”, so that the document can be downloaded to any device linked to your account.</p>\n<p>By email, the sender matters. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=T48rsVm3gY7KeGkKUk\">list of Send to Kindle errors</a> has a code, E014, for a document from an address that isn't on your approved list, and that document isn't delivered. The first codes of the list are for a file in a format that the service doesn't support, and for a file that is invalid or corrupted. As of November 2026, the overview's list of supported file types has no MOBI and no AZW3 on it, and Standard Ebooks warns at the top of its Kindle instructions that its AZW3 files can't go this way.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/sync-kindle-with-mac-without-calibre/","url":"https://readwyn.com/blog/sync-kindle-with-mac-without-calibre/","title":"How to sync a Kindle with a Mac without calibre","summary":"Readwyn syncs a Kindle over USB: progress, highlights and notes come into the library. No status, rating or time read, and only a classic Kindle takes progress back.","content_html":"<p>Readwyn syncs a Kindle over its USB cable, and on a Kindle with firmware 5 or later the sync runs one way. Connect the Kindle, review “Device → Readwyn” on the Sync screen that opens, and apply it: reading progress, highlights and notes come into your library straight from the Kindle's own files, so calibre doesn't need to be installed. Only a Kindle on firmware 1 to 4 takes anything back in a sync: progress and shelves. Three things never sync, in either direction: reading status, rating and time read. A Kindle keeps none of them in a file that Readwyn can reach over USB. Readwyn also has to see the Kindle as a drive on the Mac, and <a href=\"/blog/kindle-not-showing-up-on-mac/\">not every Kindle is one</a>. <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page on USB transfers</a> (November 2026) says a Kindle Scribe and the Kindles released in 2024 need a separate application on a Mac, and that earlier models other than the Scribe don't. The newer ones <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2024/11/01/how-to-connect-kindles-to-macs-to-transfer-connect-via-usb/\">use MTP</a>, which a Mac doesn't support natively, in place of USB mass storage. So they don't mount on a Mac, and Readwyn can't find them.</p>\n<h2 id=\"what-the-sync-screen-shows\">What the Sync screen shows</h2>\n<p>Plugging in only reads the Kindle. In the Mac App Store build, the Sync screen asks once, the first time the Kindle connects: “Give Readwyn access to Kindle”. Click “Select Kindle…”, then “Allow” in the Finder dialog that opens at the Kindle. After that, the top of the screen has a drawing, the name of the Kindle's volume (“Kindle”), and a line such as “5 books on the device · 1.4 GB free”. There's no model name, because a Kindle names its model only in its serial number and Readwyn never reads that. The drawing is the nearest thing: a keyboard model for a Kindle on firmware 1 to 4, a plain slab for any other. On a Kindle with firmware 5 or later the line also gives the version, for example “firmware 5.16”, read from the Kindle's <code>system/version.txt</code>. A Kindle on firmware 1 to 4 writes no version to its storage, so its line has none.</p>\n<p>From there the steps are the ones <a href=\"/blog/sync-kobo-with-mac-without-calibre/\">a Kobo gets</a>. Under “Choose what to sync” are two cards, “Device → Readwyn” and “Readwyn → Device”. A card's button opens <a href=\"/blog/kobo-sync-preview/\">a review sheet</a> that lists every change book by book, with a checkbox on each row, and nothing is written until you apply it. What differs is the wording of the cards, which name only what this Kindle carries.</p>\n<p>Take a Kindle with firmware 5 or later. On <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2023/02/06/list-of-latest-software-version-for-each-kindle/\">The eBook Reader's list of the last software version of each model</a>, which dates from February 2023, that covers the Kindle Touch, the Paperwhites, the Voyage, the Oasis, and the plain Kindle from its 7th generation on (<a href=\"/blog/which-kindle-do-i-have/\">how to tell which Kindle you have</a>). The Scribe is on the list too, but it and the 2024 models don't mount on a Mac, so nothing here applies to them. On one of these that does mount, the first card reads “Bring progress, highlights, notes and My Words from the Kindle into the library.” The second has nothing to send. While Readwyn counts the changes it reads “Nothing goes from the library to this Kindle. The list below tells why.” Once the count is done, the card reads “Nothing to change: the device and the library agree.” and is marked “Up to date”. The list under the cards is headed “Not available on this Kindle”, and each line has its reason. “Reading status”, for instance, says “A Kindle keeps no reading status in the files that Readwyn can reach over USB.”</p>\n<p>So on these Kindles the sync runs one way. Progress comes in for MOBI and AZW3 books. <a href=\"/blog/kindle-highlights-in-ebook-library/\">Highlights and notes</a> arrive yellow, since the Kindle's files record no colour that Readwyn can read, and the words of <a href=\"/blog/kindle-vocabulary-builder-on-mac/\">the Vocabulary Builder</a> land in My Words. A KFX book is the exception. Readwyn can't read it, so it gives no progress, and the words you looked up in it stay on the Kindle. Its highlights are taken from <code>documents/My Clippings.txt</code> and placed in your library's EPUB by their text.</p>\n<h2 id=\"which-books-take-part\">Which books take part</h2>\n<p>A sync covers the books that are both in your library and on the Kindle. It never adds a book and never deletes one. A book that Readwyn sent is known by the path that Readwyn recorded for it. For any other, Readwyn reads <code>metadata.calibre</code> if the Kindle has that file. It's calibre's list of the books on the device, and <a href=\"/blog/import-calibre-library/\">the calibre import</a> kept each book's calibre UUID, so a book on that list matches by its UUID. Without the file, matching falls back to the title and the first author.</p>\n<p>Books on the Kindle that aren't in the library are counted in one line under the cards, with “Show” to list them, and a sync skips them. To put a library book on the Kindle, pick “Send to device” from the “…” menu of the book screen (<a href=\"/blog/put-epub-on-kindle-from-mac-usb/\">putting an EPUB on a Kindle</a>). It converts the EPUB to MOBI and writes it to the Kindle's <code>documents/</code> folder.</p>\n<h2 id=\"what-only-a-classic-kindle-syncs\">What only a classic Kindle syncs</h2>\n<p>A classic Kindle runs firmware 1 to 4. On the same list, that's the first Kindle (last version 1.2.1), the Kindle 2 and the Kindle DX (2.5.8), the Kindle Keyboard (3.4.3), and the Kindle of the 4th and 5th generation (4.1.4). Readwyn can't read that number from the Kindle, so it goes by the files: a classic Kindle keeps an <code>.mbp</code> file beside each book it has opened. Its second card then reads “Send progress and shelves from the library to the Kindle.” A Kindle that hasn't opened a book yet has no such file and takes nothing. Its list has “Progress to the device” and “Shelves to the device”, both with the reason “This Kindle has not opened a book yet, so Readwyn cannot tell which files it keeps.”</p>\n<p>Progress goes back as the last-read position in that <code>.mbp</code> file. Readwyn changes the position in place and never creates the file, so a book the Kindle hasn't opened takes nothing. The review leaves it out and says why: “Progress stays in Readwyn. The Kindle has not opened this book yet. Open it once on the Kindle.” With two directions the order matters, because the receiving side takes the sender's value. Run “Device → Readwyn” first.</p>\n<p>Shelves travel both ways as collections of the same name, through the Kindle's <code>system/collections.json</code>. Collections came to the Kindle 2 and the Kindle DX with <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2010/06/19/kindle-2-and-dx-firmware-update-folders-finally/\">software 2.5 in 2010</a>, and a classic Kindle on firmware 2.5 to 4 keeps them in that file. The first Kindle stopped at 1.2.1, so it has none. Sending shelves is tagged “Experimental”: it has been checked on a copy of a Kindle's disk, which can't run, and not on a running Kindle. The review carries the notice “Restart the Kindle to see the collections” with its steps: “After you apply, eject the Kindle and unplug it. On the Kindle, press Home, then Menu, choose Settings, press Menu again, and choose Restart. The power button does not load them. Readwyn backs up the old collections file first.”</p>\n<p>Two things stay out even here. Highlights and notes made in Readwyn don't go to the Kindle, because nothing public describes how to add a record to its annotation file and a damaged file would lose the Kindle's own notes. And there are no words to bring: “Kindle firmware 1 to 4 has no Vocabulary Builder.”</p>\n<p>Every write of progress or collections to a classic Kindle is guarded. The old file is copied to <code>kobo-backups</code> in Readwyn's app folder on the Mac, the new file is written under a temporary name and renamed into place, and Readwyn reads it back. If that check fails, the old file returns. After a “Readwyn → Device” sync the message says “The Kindle’s files are backed up.”, and “Show backups in Finder” in Settings › E-readers opens the folder.</p>\n<p>The progress write was tried on a copy of a real classic Kindle's disk. “Device → Readwyn” brought in the progress of four books: 1%, 0%, 30% and 2%. “Readwyn → Device” then moved one of them from 30% to 60%. Afterwards the Kindle's file gave 60% and had kept its length, the old file sat in the backup folder, and <code>fsck_msdos -n</code> found the volume clean.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-an-asin/","url":"https://readwyn.com/blog/what-is-an-asin/","title":"What is an ASIN, and how does it differ from an ISBN?","summary":"An ASIN is the 10-character code Amazon gives every item in its store. An ISBN is the book trade's number for one edition in one format, and a Kindle ebook can sell without one.","content_html":"<p>An ASIN (Amazon Standard Identification Number) is Amazon's own catalogue number: a <a rel=\"external\" href=\"https://sell.amazon.com/blog/what-is-an-asin\">unique 10-character code of letters and digits</a> that Amazon gives every product in its store, books included. An ISBN belongs to the book trade as a whole. It names the publisher and <a rel=\"external\" href=\"https://www.isbn-international.org/content/what-isbn/10\">the specific title, edition and format</a>, so a paperback and an EPUB of the same text get different ones. Amazon gives a book an ASIN whether or not it has an ISBN, and Kindle Direct Publishing says plainly that <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G7DMSKCM9DVS65TC\">eBooks don't require an ISBN</a>: publish one without, and Amazon assigns only the ASIN.</p>\n<h2 id=\"why-an-asin-has-10-characters\">Why an ASIN has 10 characters</h2>\n<p>The length is a leftover from Amazon's first years, when it sold only books. Rebecca Allen, an engineer there from 1996 to 1998, has written <a rel=\"external\" href=\"https://inventlikeanowner.com/blog/the-story-behind-asins-amazon-standard-identification-numbers/\">the story behind ASINs</a>. The catalogue was keyed on the ISBN, then a 10-digit number, and the code stored those keys as 10-character strings. When Amazon began to sell things that had no ISBN, she proposed a new key with the same length and the same set of characters, so that the code could stay as it was. A book that had an ISBN kept it as its ASIN. Everything else got a serial number in base 36, the 26 letters plus the 10 digits. She had met many objections to base 62, which adds the small letters, because it would have made the key case sensitive: <code>b000000000</code> and <code>B000000000</code> would have been two different numbers. The counter started at <code>B000000000</code>.</p>\n<p>You can still see that split. Amazon's seller guide says an ASIN <a rel=\"external\" href=\"https://sell.amazon.com/blog/what-is-an-asin\">typically starts with “B0”</a>, while a book's begins with a number. And when two formats of a self-published book fail to appear on one product page, KDP's help asks for <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GTQGN866DSS6XBD2\">the ASIN of the eBook and the 10-digit ISBN</a> of the paperback or hardcover: the print book is still known by its ISBN, the ebook by Amazon's number.</p>\n<h2 id=\"the-two-numbers-side-by-side\">The two numbers side by side</h2>\n<table><thead><tr><th></th><th>ASIN</th><th>ISBN</th></tr></thead><tbody>\n<tr><td>Who gives it</td><td>Amazon</td><td>The publisher, through the ISBN system</td></tr>\n<tr><td>What it names</td><td>One product in Amazon's store</td><td>One title, in one edition and one format</td></tr>\n<tr><td>Form</td><td>10 letters and digits</td><td>13 digits since 1 January 2007, starting 978 or 979</td></tr>\n<tr><td>Needed for a Kindle ebook</td><td>Yes, Amazon assigns it</td><td>No</td></tr>\n</tbody></table>\n<p><a href=\"/blog/isbn-for-ebooks/\">Does an ebook need its own ISBN?</a> covers who requires one for an ebook and what it costs.</p>\n<h2 id=\"when-an-isbn-has-no-10-digit-form\">When an ISBN has no 10-digit form</h2>\n<p>Since 1 January 2007, <a rel=\"external\" href=\"https://www.isbn-international.org/content/what-isbn/10\">every ISBN has 13 digits</a> and starts with 978 or 979. A 978 number still has an older 10-digit form. A 979 number has none. In a <a rel=\"external\" href=\"https://www.isbn-international.org/node/331\">notice of November 2019</a>, the International ISBN Agency said that the United States agency would start giving out 979 numbers in early 2020. The same notice warns that a 979 number does not have an equivalent 10-digit ISBN, and that converting one leads to duplicated ISBNs and orders for the wrong book. A book whose only ISBN starts with 979 therefore has no 10 characters that could double as its ASIN, which is one more reason not to treat the two numbers as interchangeable.</p>\n<h2 id=\"finding-a-book-s-asin\">Finding a book's ASIN</h2>\n<p>On a product page, the ASIN is the part of the address <a rel=\"external\" href=\"https://sell.amazon.com/blog/what-is-an-asin\">after “/dp/”</a>, and it is also listed in the product details further down the page. An author who publishes through KDP finds it <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200652190\">under the book's title on the KDP Bookshelf</a>. The same KDP page gives the short link for each of Amazon's stores: the store's address, <code>/dp/</code>, and the ASIN, for example <code>amazon.co.uk/dp/</code> followed by the ASIN.</p>\n<h2 id=\"inside-a-kindle-file\">Inside a Kindle file</h2>\n<p>A MOBI or AZW3 file can carry both numbers in its header. The <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/MOBI\">MobileRead wiki's description of the format</a> lists one record for the ISBN (104), one for the ASIN (113), and one for the content type (501), where <code>EBOK</code> means an ebook and <code>PDOC</code> a personal document.</p>\n<p>Readwyn, the ebook library app this blog belongs to, reads that ASIN to match books on a Kindle that the Mac mounts as a drive (<a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon says</a> that a Kindle Scribe and the Kindles released in 2024 need a separate application on a Mac, and that earlier models don't). A word from the Vocabulary Builder finds its book by the unique ID in the header, else by the ASIN, else by the title (<a href=\"/blog/kindle-vocabulary-builder-on-mac/\">more on that</a>). The ASIN field of a sideloaded file isn't always an Amazon number, though: in the four calibre-made books on the Kindle that Readwyn is tested against, that field holds calibre's own ID for the book, a UUID.</p>\n","date_published":"2026-11-11T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/add-fonts-to-kobo-and-kindle/","url":"https://readwyn.com/blog/add-fonts-to-kobo-and-kindle/","title":"How to add your own fonts to a Kobo or a Kindle","summary":"A Kobo or a Kindle reads TTF and OTF files from a folder named fonts on its USB drive. Kobo allowed this in 2011, and Amazon made it official in 2018.","content_html":"<p>Kobo's help article, <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/13009477876631-Load-fonts-onto-your-Kobo-eReader\">Load fonts onto your Kobo eReader</a>, gives the steps for a Kobo: connect the reader to your computer with its USB cable, tap Connect, and open the drive called KOBOeReader. Put your .ttf or .otf files in a folder named <code>fonts</code> at the top level of that drive (create the folder if it isn't there) and eject. The font then shows in the Text menu, where you tick it under Supplemental before you choose it. A Kindle takes the same steps, except that its <code>fonts</code> folder is already on the drive once the software is version 5.9.6 or later, and it accepts .ttc files as well. Kobo allowed this in 2011; Amazon made it official only in 2018.</p>\n<h2 id=\"kobo\">Kobo</h2>\n<p>On 22 June 2011, The eBook Reader <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2011/06/22/how-to-install-new-fonts-on-the-kobo-touch-ereader/\">reported</a> that a folder called <code>fonts</code> at the root of a Kobo Touch, next to the <code>.kobo</code> and <code>.adobe-digital-editions</code> folders, was all it took. After an eject, the extra faces appeared under Aa beside the seven that came with the device: Georgia, Avenir, Amasis, Delima, Felbridge, Gill Sans and Rockwell. A couple of weeks later <a rel=\"external\" href=\"https://www.adweek.com/performance-marketing/kobo-touch-update-lets-users-add-fonts/\">Adweek wrote up</a> firmware 1.9.5 for the Touch, which let owners add fonts and change the margins, and which had brought five of those seven built-in faces with it.</p>\n<p>Kobo's current help page still sends you to the same folder, though the menus have changed. You open a book, tap the middle of the page, tap the Text icon, tick the new font in the drop-down beside “Supplemental”, and then choose it beside “Font Face”. Kobo lists .otf and .ttf as the supported types, asks you to unzip a downloaded font before copying it, and admits that some fonts don't render properly; its advice in that case is to find another version of the same font. The same page names dyslexia-friendly fonts as one reason to load your own (whether such fonts help readers is <a href=\"/blog/do-dyslexia-fonts-work/\">a separate question</a>).</p>\n<p>A new font won't show up in a PDF: Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017639913\">text settings</a> says you can't change the size or style of the text in a PDF.</p>\n<h2 id=\"kindle\">Kindle</h2>\n<p>Before Amazon supported it, owners found a way round. In January 2013 The eBook Reader <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2013/01/06/how-to-easily-add-custom-fonts-to-kindle-paperwhite-no-hacks/\">described a trick</a> for the first Kindle Paperwhite on firmware 5.3.0 or later: an empty file named <code>USE_ALT_FONTS</code> at the root of the drive, a <code>fonts</code> folder beside it, files named Fontname-Regular, Fontname-Italic, Fontname-Bold and Fontname-BoldItalic, and a restart. It needed no jailbreak, but Amazon never documented it, and the site later added a note that it didn't work on the second Paperwhite, released later in 2013.</p>\n<p>The supported way arrived with a Kindle software update in June 2018. On 6 June, TeleRead <a rel=\"external\" href=\"https://teleread.org/2018/06/06/kindle-owners-can-now-enjoy-custom-fonts-not-just-those-on-the-menu-heres-how/\">reprinted the readme</a> that the update left inside the new <code>fonts</code> folder. It accepts OpenType and TrueType fonts with the extensions .ttf, .otf and .ttc, and nothing else. You unzip the font if it came compressed, copy the files into the folder, disconnect the Kindle, and choose the font in the Display Settings (Aa) menu. Three days later The eBook Reader <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2018/06/09/how-to-use-kindles-new-custom-font-feature-video/\">listed the models</a>: every Kindle back to the Paperwhite 2, which then meant the sixth to ninth generations, as long as it ran software version 5.9.6 or newer. If your Kindle is one of those and still has no <code>fonts</code> folder, update it first. If you aren't sure which generation you own, <a href=\"/blog/which-kindle-do-i-have/\">the model number tells you</a>. <a rel=\"external\" href=\"https://www.howtogeek.com/734645/how-to-install-custom-fonts-on-your-kindle/\">How-To Geek's 2021 guide</a> gives the newer path on the device: Aa, then Font, then Font family.</p>\n<h2 id=\"what-the-reader-does-with-the-files\">What the reader does with the files</h2>\n<p>Copy the whole family, not only the regular file. The 2011 Kobo guide gave the reason plainly: most books use italic and bold, so those files belong in the folder too. Amazon's readme makes the same recommendation, and The eBook Reader's 2018 Kindle guide used a set of four as its example: Georgia-Regular.ttf, Georgia-Italic.ttf, Georgia-Bold.ttf and Georgia-BoldItalic.ttf.</p>\n<p>A Kindle also checks what you give it. The readme says a custom font applies “for most books”, and that the Kindle draws the text in its default system font when the font can't be used, for example when it lacks characters the book needs or the file is damaged. The eBook Reader found that a font the Kindle rejects doesn't appear in the font list at all.</p>\n<p>The readme ends with a line about rights: you, not Amazon, are responsible for having the right to use the fonts you install. A font under the <a rel=\"external\" href=\"https://openfontlicense.org/open-font-license-official-text/\">SIL Open Font License</a> is the easy case, because that licence lets anyone use, copy, embed and redistribute the font, as long as the licence goes with it, and forbids selling the font on its own. A font you bought, or one that came with your computer, has its own licence, and its terms decide whether a copy on an e-reader is allowed.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/amazon-deleted-1984-from-kindles/","url":"https://readwyn.com/blog/amazon-deleted-1984-from-kindles/","title":"The time Amazon deleted 1984 from Kindles","summary":"What happened in July 2009 when Amazon wiped Orwell's 1984 and Animal Farm from Kindles, how Bezos apologised, and what the lawsuit that followed won.","content_html":"<p>In July 2009 Amazon reached into its customers' Kindles and deleted two George Orwell novels, Nineteen Eighty-Four and Animal Farm, then refunded what people had paid. The story broke on 17 July. Amazon's spokesman, Drew Herdener, <a rel=\"external\" href=\"https://www.theguardian.com/technology/2009/jul/17/amazon-kindle-1984\">told the Guardian</a> that a third party with no rights to the books had added them through Amazon's self-service platform, and that when the rights holder complained, Amazon removed the copies \"from our systems and from customers' devices\".</p>\n<p>The seller was a company called MobileReference. Before Herdener's statement, customer service had told buyers the publisher had pulled the titles, which <a rel=\"external\" href=\"https://www.engadget.com/2009-07-17-amazon-remotely-deletes-orwell-e-books-from-kindles-unpersons-r.html\">Engadget reported</a> along with a message saying the books now showed as \"not yet available\". Orwell died in 1950, so his work was already free of copyright in some countries, though not yet in the US or Europe, <a rel=\"external\" href=\"https://www.theguardian.com/technology/2009/jul/17/amazon-kindle-1984\">the Guardian noted</a>.</p>\n<p>Nobody missed the irony. In the novel, inconvenient records vanish down the memory hole, and Amazon had just made a book vanish from devices people thought they owned. On 23 July Jeff Bezos posted an apology on the Kindle forum, calling the company's fix \"stupid, thoughtless and painfully out of line with our principles\" (<a rel=\"external\" href=\"https://bits.blogs.nytimes.com/2009/07/23/amazon-chief-says-erasing-orwell-books-was-stupid/\">The New York Times</a>).</p>\n<p>A week later Justin Gawronski, a 17-year-old high school student, sued. He'd been writing notes on his Kindle copy of Nineteen Eighty-Four for a summer assignment, and the deletion left them useless; the suit <a rel=\"external\" href=\"https://www.engadget.com/2009-07-30-student-sues-amazon-after-kindle-eats-his-homework.html\">asked for class-action status</a> and a ban on Amazon getting into Kindles like that again. In September Amazon offered affected buyers a choice: the book delivered again with their notes intact, or <a rel=\"external\" href=\"https://www.engadget.com/2009-09-04-amazon-offers-to-give-back-your-kindles-copy-of-1984.html\">a $30 cheque</a>.</p>\n<p>Gawronski and a co-plaintiff, Antoine Bruguier, never got to trial. On 25 September 2009 the two sides filed a settlement in the US District Court in Seattle, subject to the court's approval (<a rel=\"external\" href=\"https://web.archive.org/web/20110424165344/http://www.techflash.com/seattle/2009/09/amazon_settles_lawsuit_over_deleted_1984.html\">TechFlash</a>). Amazon agreed to pay $150,000 in fees to the plaintiffs' lawyers, and the lead firm, KamberEdelson, said it would give its share to charity. The more lasting part was a written promise. The Kindle terms of the time granted \"the non-exclusive right to keep a permanent copy\" of each book, and for books bought under those terms on Kindles used in the US, the <a rel=\"external\" href=\"https://web.archive.org/web/20240808062154/https://cases.justia.com/federal/district-courts/washington/wawdce/2:2009cv01084/161529/15/0.pdf\">stipulation</a> says Amazon wouldn't remotely delete or change them unless the owner agreed, the owner asked for a refund or the payment failed, a court or regulator ordered it, or deletion was needed to protect the user, the device, or the network (harmful code hidden in an ebook, say).</p>\n<p><a rel=\"external\" href=\"https://www.engadget.com/2009-10-01-amazon-clarifies-kindle-book-deletion-policy-can-still-delete-b.html\">Engadget's Nilay Patel</a> saw the gap in it: a court order would still be enough to pull a book, while no judge can send a bookshop into your house to take back a paper copy. And the promise covered US devices only.</p>\n<p>The episode still comes up whenever people ask <a href=\"/blog/do-you-own-your-kindle-books/\">whether you own your Kindle books</a>. At the time, Cindy Cohn of the Electronic Frontier Foundation <a rel=\"external\" href=\"https://www.theguardian.com/technology/2009/jul/22/kindle-amazon-digital-rights\">told the Guardian</a> that DRM leads buyers to think they've purchased something when, to the seller and often the law, they've got much less. Orwell's books left copyright in the UK on <a rel=\"external\" href=\"https://www.theguardian.com/books/booksblog/2021/jan/01/george-orwell-is-out-of-copyright-what-happens-now\">1 January 2021</a>. In the US, a book published in 1949 with a copyright notice and then renewed is protected for 95 years (<a rel=\"external\" href=\"https://guides.library.cornell.edu/copyright/publicdomain\">Cornell's public-domain chart</a>), so Nineteen Eighty-Four stays protected there <a rel=\"external\" href=\"https://standardebooks.org/ebooks/george-orwell/nineteen-eighty-four\">until the end of 2044</a>.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/android-e-readers/","url":"https://readwyn.com/blog/android-e-readers/","title":"What is an Android e-reader, and who is it for?","summary":"An Android e-reader is an E Ink reader with the Play Store, so Kindle, Kobo and Libby apps sit side by side, at a cost in battery, refresh and buying.","content_html":"<p>If your ebooks come from Amazon, from Kobo and from a library card, an Android e-reader can hold all three. It has the same kind of E Ink screen as a Kobo, but it runs Google's Android with the Play Store, so the Kindle app, the Kobo app and Libby install side by side. In return you get a busier device, with more settings to tend.</p>\n<p>BOOX's Go 7 is a typical example. Its <a rel=\"external\" href=\"https://shop.boox.com/products/go7\">spec page</a> lists a 7 inch Carta 1300 screen at 300 ppi, physical page-turn buttons, a weight of about 195 g, Android 13 and a built-in Google Play Store. Kobo's <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-clara-bw\">Clara BW</a> uses a Carta 1300 panel as well, so the page looks much the same.</p>\n<p>On the Play Store you'll find <a rel=\"external\" href=\"https://play.google.com/store/apps/details?id=com.amazon.kindle\">Amazon Kindle</a>, <a rel=\"external\" href=\"https://play.google.com/store/apps/details?id=com.kobobooks.android\">Kobo Books</a> and <a rel=\"external\" href=\"https://play.google.com/store/apps/details?id=com.overdrive.mobile.android.libby\">Libby</a>, the library app, which <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6105.htm\">needs Android 7.1 or later</a>. Amazon <a rel=\"external\" href=\"https://www.engadget.com/amazon-removes-direct-purchasing-android-161708963.html\">switched off book purchases in the Android Kindle app</a> at the end of May 2022 rather than sell through Google's Play billing, and for a while a buyer had to <a rel=\"external\" href=\"https://www.slashgear.com/1694179/why-cannot-buy-kindle-books-on-amazon-app-mobile-android-iphone-ipad-mobile\">pay on Amazon's website</a> and then open the book in the app. Buying in the app came back in parts of Europe first, under a Google programme for the European Economic Area that lets an app use its own billing (<a rel=\"external\" href=\"https://www.androidauthority.com/buy-books-kindle-app-europe-3445793/\">Android Authority, May 2024</a>). In January 2025 Amazon told <a rel=\"external\" href=\"https://animecorner.me/amazon-kindle-android-in-app-purchases-united-states/\">Anime Corner</a> that US users could buy in the app again. Whether the app sells books in your country is worth checking before you rely on it. Kobo and Kindle e-readers borrow library books by other routes, covered in <a href=\"/blog/library-ebooks-on-kobo-and-kindle/\">library ebooks on Kobo and Kindle</a>.</p>\n<p>An Android reader can also run apps that read OPDS catalogues, such as KOReader or Moon+ Reader. Readwyn shares its libraries that way: with Readwyn server on, turn on “Share with reader apps (OPDS)” under Settings › Other devices › Reader apps, add the app, and it browses the books by author, series or shelf over the home Wi-Fi and downloads each as EPUB 3 or in its original format (<a href=\"/blog/opds-catalogue-for-koreader/\">setting it up in KOReader</a>).</p>\n<p>\"Android\" covers a wide spread of versions, even inside one brand. The comparison table on the Go 7 page puts the first Go Color 7 on Android 12, beside the Go 7 on Android 13. BOOX says its Note Air5 C <a rel=\"external\" href=\"https://shop.boox.com/blogs/news/note-air5-c-gets-smoother-new-smooth-refresh-mode-reduces-visual-lag\">runs Android 15</a>, and the <a rel=\"external\" href=\"https://shop.boox.com/blogs/news/unveiling-boox-palma-3\">Palma 3</a>, a phone-shaped reader announced on 25 October 2026, ships with Android 16. Bigme's European shop had its 7 inch colour <a rel=\"external\" href=\"https://euroshop.bigme.vip/products/bigme-b751c-s-upgraded-7inch-color-ereder-with-android-14-os\">B751C S</a> on Android 14 at €229 (the price on 1 November 2026), beside a B7 Pro with 4G calling, which is closer to a phone than a reader. Before you buy an older model, check the minimum Android version of the apps you rely on.</p>\n<p>The screen costs you something too. Phone apps assume a screen that redraws instantly, and E Ink can't keep up with a scrolling feed (the reasons are in <a href=\"/blog/why-e-ink-flashes/\">why E Ink flashes</a>). Android readers give you a choice of refresh modes instead. BOOX introduced its Super Refresh technology in 2022 with <a rel=\"external\" href=\"https://shop.boox.com/blogs/news/boox-super-refresh-bsr-technology\">four modes</a>: HD for long reading with the least ghosting, Balanced, Fast for scrolling websites, and Ultrafast, which drops detail so that video and app switching keep up. A firmware update that BOOX <a rel=\"external\" href=\"https://shop.boox.com/blogs/news/note-air5-c-gets-smoother-new-smooth-refresh-mode-reduces-visual-lag\">announced in December 2025</a> adds a mode called “Smooth” to EinkWise, the display settings of the Note Air5 C, for third-party apps with moving content. Speed costs sharpness: even BOOX's own description of Fast mode warns of some ghosting.</p>\n<p>Then there is the battery. The Clara BW gets by with a 1,500 mAh cell and a 1.0 GHz processor, and <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-clara-bw\">Kobo says</a> one charge lasts up to 53 days, counting 30 minutes of reading a day with the front light at 10% and Wi-Fi and Bluetooth off. The Go 7 carries 2,300 mAh and an octa-core processor, and its spec page gives no reading time in days at all.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/apple-ebook-price-fixing/","url":"https://readwyn.com/blog/apple-ebook-price-fixing/","title":"Why Apple lost the ebook price-fixing case","summary":"How a US court found in 2013 that Apple organised five publishers to raise ebook prices, why its appeals failed, and where the $450 million went.","content_html":"<p>Apple lost because the courts found that it had organised five of the largest US publishers into an agreement to raise ebook prices. Under the Sherman Act, price fixing among competitors is illegal per se: a court doesn't ask whether it did any good. Apple's contracts with the publishers weren't unlawful in themselves. What sank it was the evidence that Apple used them, knowingly, to get the publishers to move together against Amazon's $9.99 price.</p>\n<h2 id=\"the-2010-contracts\">The 2010 contracts</h2>\n<p>By November 2009 Amazon sold 90% of all ebooks in the US, according to the facts found at trial and repeated in the <a rel=\"external\" href=\"https://portal.ct.gov/-/media/ag/press_releases/2015/20150630appleebookssecondcircuitopinionaffirmingdistrictcourtpdf.pdf\">Second Circuit's 2015 opinion</a>. It bought ebooks at a wholesale price and set its own retail price, and since <a href=\"/blog/first-kindle-2007/\">the first Kindle in late 2007</a> it had sold new releases and New York Times bestsellers for $9.99. Publishers saw that price as a threat to hardcover sales.</p>\n<p>Apple wanted a bookstore on the iPad, which it unveiled on 27 January 2010. Its offer was the agency model: the publisher sets the consumer price and keeps 70% of each sale, and Apple takes 30% as a commission. Apple's first emails to the publishers, on 4 and 5 January, proposed caps of $14.99 for books whose hardcover list price was above $35, $12.99 below that, and $9.99 for other trade books. Several publishers balked at those limits, and Eddy Cue, the Apple senior vice president who led the talks, got Steve Jobs to accept looser ones. A new release could now cost $12.99 if its hardcover was priced $25 to $27.50, $14.99 if it cost more but no more than $30, and $16.99 to $19.99 above that.</p>\n<p>The clause that mattered most was a most-favoured-nation (MFN) clause. If any other retailer sold an ebook for less, the publisher had to cut the iBookstore price to match. The court of appeals worked through the arithmetic: on a $25 hardcover with a $12.50 wholesale price, a publisher selling through Apple at the $12.99 cap kept 70% of the price, about $9.09 a copy (the opinion prints $8.75), less than the $12.50 that Amazon paid. If Amazon stayed at $9.99, the MFN would drag the iBookstore down to $9.99 too, leaving the publisher with less money and no control of the price. In practice, signing with Apple meant forcing Amazon onto agency terms as well. Cue said as much: “any decent MFN forces the model”.</p>\n<p>Hachette, HarperCollins, Macmillan, Penguin and Simon &amp; Schuster signed before the launch. Random House stayed out until early 2011. Asked after the iPad event why anyone would pay Apple $14.99 when Amazon charged $9.99, Jobs replied: “the price will be the same”.</p>\n<h2 id=\"prices-after-the-launch\">Prices after the launch</h2>\n<p>The day after the launch, Macmillan's chief executive, John Sargent, flew to Seattle and told Amazon to accept agency terms or wait seven months for Macmillan's new releases. Amazon took away the option to buy Macmillan's print and ebook titles, then gave way: it had agency terms with Macmillan by 5 February 2010. Three more publishers followed by March and Penguin in June.</p>\n<p>According to the <a rel=\"external\" href=\"https://portal.ct.gov/-/media/ag/press_releases/2015/20150630appleebookssecondcircuitopinionaffirmingdistrictcourtpdf.pdf\">court of appeals</a>, in the five months after the iBookstore opened those publishers priced 85.7% of their new releases on the Kindle at or just below Apple's caps, and 96.8% of their bestsellers. Random House's prices barely moved. An expert for the Justice Department measured the change from February 2010 to February 2011: the weighted average price of the five publishers' new releases rose 24.2%, bestsellers 40.4%, and all their ebooks together 23.9%. Some publishers also raised hardcover list prices so that the ebook edition fell into a higher price band.</p>\n<h2 id=\"the-suit-and-the-settlements\">The suit and the settlements</h2>\n<p>The Justice Department sued Apple and the five publishers on 11 April 2012, and <a rel=\"external\" href=\"https://portal.ct.gov/-/media/ag/press_releases/2015/20150630appleebookssecondcircuitopinionaffirmingdistrictcourtpdf.pdf\">33 states and territories</a> brought their own case. <a rel=\"external\" href=\"https://www.npr.org/2012/04/11/150445827/justice-department-sues-apple-publishers-for-price-fixing\">Hachette, HarperCollins and Simon &amp; Schuster settled the same day</a>. <a rel=\"external\" href=\"https://phys.org/news/2012-12-penguin-settlement-e-books-lawsuit.html\">Penguin followed on 18 December 2012</a>, agreeing to end its contracts with Apple and other retailers and, for two years, to sign none that limited retailers' discounts. <a rel=\"external\" href=\"https://www.writerswrite.com/macmillan-settles-ebook-price-fixing-20920131\">Macmillan settled in February 2013</a>; Sargent said a loss could have cost more than the company was worth. The publishers paid <a rel=\"external\" href=\"https://fortune.com/2014/11/21/apple-450-million-e-book-settlement-gets-final-court-approval\">$166 million</a> to ebook buyers in all.</p>\n<h2 id=\"the-trial\">The trial</h2>\n<p>Apple went to trial alone. The bench trial before Judge Denise Cote in Manhattan began on 3 June 2013 and lasted three weeks; she ruled on 10 July 2013, <a rel=\"external\" href=\"https://ago.nebraska.gov/node/146\">as the Nebraska attorney general's office summarised the case</a>. She found that Apple had orchestrated a conspiracy among the publishers to raise prices, a horizontal price-fixing agreement that was <a rel=\"external\" href=\"https://portal.ct.gov/-/media/ag/press_releases/2015/20150630appleebookssecondcircuitopinionaffirmingdistrictcourtpdf.pdf\">illegal per se</a>, and that it failed the more forgiving rule-of-reason test as well.</p>\n<p>Apple's defence was that its terms made commercial sense for a newcomer: an agency model, caps and an MFN guaranteed it a small profit on every sale, and if the publishers then used them against Amazon, Apple had at most facilitated that without meaning to. Both courts rejected this because of how Apple behaved. It assured each publisher that the others were getting identical terms, Sargent told Cue about the Seattle trip in advance, and Cue's team watched Kindle prices change. When Penguin's Amazon deal lagged, Cue told Jobs that Apple was moving a batch of Penguin titles to $9.99. Judge Cote found Cue's testimony on several of these points not credible. Her injunction of 5 September 2013 barred Apple from enforcing MFNs with publishers, made it treat ebook apps like any other app in its store, and provided for an outside monitor to check that Apple complied.</p>\n<h2 id=\"the-appeal\">The appeal</h2>\n<p>A panel of the Second Circuit <a rel=\"external\" href=\"https://portal.ct.gov/-/media/ag/press_releases/2015/20150630appleebookssecondcircuitopinionaffirmingdistrictcourtpdf.pdf\">upheld the ruling 2 to 1 on 30 June 2015</a>. Judge Debra Ann Livingston wrote that a company at one level of a market which organises a price-fixing cartel among firms at another level (a hub-and-spoke conspiracy) is judged by the per se rule just as its co-conspirators are. Raymond Lohier, the other judge in the majority, agreed on that point; Livingston alone added that Apple would also lose under the rule of reason. The third judge, Dennis Jacobs, dissented. As the majority described his view, he accepted that Apple had organised the conspiracy but thought it was entitled to, as a way into a market Amazon dominated. Livingston called that idea “marketplace vigilantism”.</p>\n<p>Apple petitioned the Supreme Court in October 2015, and on 7 March 2016 the court <a rel=\"external\" href=\"https://www.supremecourt.gov/docketfiles/15-565.htm\">declined to hear the case</a>. That left the 2013 judgment in place.</p>\n<h2 id=\"the-450-million\">The $450 million</h2>\n<p>In June 2014, two months before a damages trial at which the states and a class of buyers were expected to seek up to $840 million, Apple agreed to an unusual settlement that depended on its appeal. <a rel=\"external\" href=\"https://fortune.com/2014/11/21/apple-450-million-e-book-settlement-gets-final-court-approval\">As Reuters reported</a>, Apple would pay $400 million to consumers and $50 million to lawyers if it lost, $50 million and $20 million if the case went back to Judge Cote, and nothing if it won outright. Cote approved it on 21 November 2014.</p>\n<p>The Supreme Court's refusal triggered the $400 million payment. From June 2016, people who had bought the affected ebooks between 1 April 2010 and 21 May 2012 got <a rel=\"external\" href=\"https://news.delaware.gov/?p=28619\">$6.93 for each New York Times bestseller and $1.57 for every other title</a>. Amazon, Apple, Barnes &amp; Noble and Kobo customers received account credits automatically; Sony and Google customers were sent cheques.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/atkinson-hyperlegible/","url":"https://readwyn.com/blog/atkinson-hyperlegible/","title":"What is Atkinson Hyperlegible, and does it help low-vision readers?","summary":"The Braille Institute's free font makes I, l and 1 look different, but its makers have published no study, and in a survey of 2,022 people it ranked behind Verdana.","content_html":"<p>Its makers haven't published evidence that it does. Atkinson Hyperlegible is a free sans-serif typeface from the <a rel=\"external\" href=\"https://www.brailleinstitute.org/freefont/\">Braille Institute</a>, dating from 2019 and drawn so that characters which look alike in most fonts look different: the capital I, the lowercase l and the digit 1, or the capital O and the zero. The institute's font page lists awards and download counts but no study, and in a large survey that included it, people with poor near vision picked it a little less often than five other fonts, Verdana among them.</p>\n<h2 id=\"where-it-came-from\">Where it came from</h2>\n<p>The font grew out of a rebrand. J. Robert Atkinson founded the Braille Institute in 1919. Ahead of its centenary, <a rel=\"external\" href=\"https://www.allaboutvision.com/low-vision/atkinson-hyperlegible-typeface/\">All About Vision reported</a> in November 2019, the institute hired a New York design firm to update its identity, with Elliott Scott as lead designer. That firm, Applied Design Works, calls Atkinson a blinded cowboy on its <a rel=\"external\" href=\"https://helloapplied.com/projects/braille-institute\">project page</a>, which also records the font's 2019 Innovation by Design Award from Fast Company. In 2024, <a rel=\"external\" href=\"https://www.brailleinstitute.org/freefont/\">the institute says</a>, the Cooper Hewitt, Smithsonian Design Museum added the font to its permanent collection.</p>\n<p>By late July 2025, Munich's transport company MVG had moved the destination displays in its U-Bahn trains and on its U-Bahn platforms to a slightly adapted version of Atkinson Hyperlegible. A spokesperson told the <a rel=\"external\" href=\"https://www.abendzeitung-muenchen.de/muenchen/aufgefallen-was-bei-den-u-bahn-anzeigen-in-muenchen-ploetzlich-anders-ist-art-1071181\">Abendzeitung</a> that looks had little to do with it. The paper explains that the font was made for accessibility and legibility, and that the switch should make the U-Bahn easier to use for passengers with impairments.</p>\n<h2 id=\"how-the-letters-differ\">How the letters differ</h2>\n<p>Most typefaces aim for harmony, with letters that share strokes, and that likeness is the trouble. All About Vision's example is \"Illness\" in Helvetica, where the first three letters can look the same.</p>\n<p>At its core Atkinson Hyperlegible is a traditional grotesque sans-serif, as the <a rel=\"external\" href=\"https://github.com/google/fonts/blob/main/ofl/atkinsonhyperlegible/DESCRIPTION.en_us.html\">description on Google Fonts</a> puts it, and it breaks that model wherever two characters could be confused. The <a rel=\"external\" href=\"https://github.com/googlefonts/atkinson-hyperlegible\">project's README</a> names the pairs: B and 8; 1, L, l and I; Q and G; E and F; p and q; i and r; O and 0.</p>\n<p><a rel=\"external\" href=\"https://www.maxkohler.com/notes/2021-02-16-atkinson-hyperreadable/\">Max Kohler's notes</a> on a talk by the designers point out fixes that are easy to miss at reading size: the lowercase i and l each carry a serif, the zero carries a slash, and the 8 is exaggerated so it can't pass for a B. <a rel=\"external\" href=\"https://www.dezeen.com/2020/09/11/atkinson-hyperlegible-typeface-applied-design-works/\">Dezeen noted</a> that the face puts distinct letters ahead of visual consistency: the capital I has serifs while the capital T has none, and round details on many letters recall braille dots.</p>\n<h2 id=\"what-the-evidence-shows\">What the evidence shows</h2>\n<p>Dezeen says the fonts were developed with a low-vision specialist from the Braille Institute and a panel of people with visual impairments, and Kohler's notes describe the team judging letterforms through glasses that simulated eye conditions. The designers said they'd used the institute's standard tests of reading speed and retention, but the notes give no results, and in 2019 the institute told All About Vision it still planned to collect data on how well the typeface worked. In March 2024 two presenters from the Braille Institute gave a session at the CSUN accessibility conference <a rel=\"external\" href=\"https://www.csun.edu/cod/conference/sessions/2024/index.php/public/presentations/view/3108.html\">billed as a review of the results of a quantitative study</a> with low-vision readers, but the programme entry gives no figures. Asked whether scientific studies shaped the design, the designers said it came from the tradition of type design.</p>\n<p>As of November 2026 the institute's font page cites use rather than results: more than 153,000 downloads from its site, more than 10,000 websites, and over 40 million impressions through Google Fonts.</p>\n<p>The nearest thing to a comparison is a survey by the Readability Group, three type and accessibility specialists, presented at axe-con in 2021. It showed 2,022 participants words in 20 fonts and scored each font by how often it was selected. In <a rel=\"external\" href=\"https://github.com/thibaudcolas/readability-group-survey\">the published summary</a>, Atkinson Hyperlegible scored 62 per cent across all participants, behind SF Pro (69), Segoe UI (65), Verdana and BBC Reith Sans (64 each); the <a rel=\"external\" href=\"https://github.com/thibaudcolas/readability-group-survey/blob/main/presentation.md\">slides of the talk</a> put it level with Red Hat Text and Roboto. Among the 372 participants with significant poor near vision it scored 61 on the slides, against 64 for Verdana and 68 for SF Pro. Fonts promoted as aids for dyslexia came last, with OpenDyslexic at 12. The survey recorded which fonts people chose, not how fast or accurately they read; the dyslexia fonts have also been <a href=\"/blog/do-dyslexia-fonts-work/\">tested on speed and accuracy</a>, with no gain.</p>\n<h2 id=\"licence-and-the-2025-family\">Licence and the 2025 family</h2>\n<p>The original <a rel=\"external\" href=\"https://github.com/googlefonts/atkinson-hyperlegible\">family</a> has four fonts, regular and bold with an italic for each, with 335 glyphs per font and accents for 27 languages. The institute's own download comes with an end-user licence agreement, but Google Fonts has carried the font since <a rel=\"external\" href=\"https://github.com/google/fonts/blob/main/ofl/atkinsonhyperlegible/METADATA.pb\">30 April 2021</a> under the <a rel=\"external\" href=\"https://openfontlicense.org/\">SIL Open Font License</a>. On an e-reader, the font is one of several <a href=\"/blog/reading-with-low-vision-on-an-e-reader/\">settings that help with low vision</a>.</p>\n<p>Atkinson Hyperlegible Next, made with the foundry Letters From Sweden, launched on 10 February 2025, <a rel=\"external\" href=\"https://www.printmag.com/type-tuesday/atkinson-hyperlegible-next-applied-design/\">PRINT reported</a>. It has seven weights, each with an italic, plus a variable version, and the <a rel=\"external\" href=\"https://www.brailleinstitute.org/freefont/\">institute's font page</a> says it covers more than 150 languages and recommends installing Next first. Google Fonts lists <a rel=\"external\" href=\"https://github.com/google/fonts/blob/main/ofl/atkinsonhyperlegiblenext/METADATA.pb\">Next</a> and its monospaced sibling, Atkinson Hyperlegible Mono, under the same open licence.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/audiobooks-vs-reading/","url":"https://readwyn.com/blog/audiobooks-vs-reading/","title":"Is listening to an audiobook the same as reading?","summary":"What studies from 2010 to 2021 found when they compared listening to an audiobook with reading the same text, and where reading still comes out ahead.","content_html":"<p>For understanding and remembering a book, mostly yes. A 2021 review of 46 studies found no reliable difference in comprehension between listening and reading, although reading did better when the reader set the pace and when the test asked for inferences rather than plain facts.</p>\n<h2 id=\"the-experiments-2010-to-2016\">The experiments, 2010 to 2016</h2>\n<p>The early classroom and lab comparisons did not favour the ear. David Daniel and William Douglas Woody gave students either a podcast or the text of the same primary source and then quizzed them. The <a rel=\"external\" href=\"https://doi.org/10.1080/00986283.2010.488542\">podcast group scored lower</a>, and the students' first preference for podcasts changed straight after the quiz. Their paper, in Teaching of Psychology in 2010, judged podcasts better for extra material than for the main content of a course.</p>\n<p>At the University of Waterloo in 2013, Trish Varao Sousa, Jonathan Carriere and Daniel Smilek expected a listener, with no page to follow, to drift off without noticing, so their main measure was attention. The 211 students in <a rel=\"external\" href=\"https://www.frontiersin.org/articles/10.3389/fpsyg.2013.00892/full\">that study</a> took in three excerpts of about 1,800 words from Bill Bryson's <em>A Short History of Nearly Everything</em>: one read silently, one read aloud, and one heard as Bryson's own recording. Up to ten times in each excerpt, at random intervals of 30 to 90 seconds, a box popped up to ask whether the student's mind had been wandering. Listeners said yes at about half of the probes, silent readers at about a third, and students reading aloud least often. Listening also gave the worst scores on the 10-question true or false test after each excerpt, and less interest in the material than reading aloud.</p>\n<p>Then a 2016 study by <a rel=\"external\" href=\"https://doi.org/10.1177/2158244016669550\">Beth Rogowsky, Barbara Calhoun and Paula Tallal</a> found no gap at all. Its 91 participants got the preface and one chapter of a non-fiction book as a digital audiobook, as e-text on a tablet, or as both at once, sat a written test straight away, and sat the same test two weeks later. Neither test showed a statistically significant difference between the three groups, and men and women recalled the same amount.</p>\n<h2 id=\"the-brain-and-the-meta-analysis\">The brain and the meta-analysis</h2>\n<p>Brain scans from 2019 suggest why the results can be so close. In Jack Gallant's lab at Berkeley, people lay in an fMRI scanner while they listened to, and while they read, several hours of the same stories. The patterns of meaning across the cortex were highly correlated between listening and reading, and a model trained on one modality predicted brain responses in the other. The paper, by Fatma Deniz, Anwar Nunez-Elizalde, Alexander Huth and Gallant, calls the two representations <a rel=\"external\" href=\"https://doi.org/10.1523/JNEUROSCI.0675-19.2019\">\"almost identical\"</a>.</p>\n<p>The review behind the first paragraph is Virginia Clinton-Lisell's <a rel=\"external\" href=\"https://doi.org/10.3102/00346543211060871\">meta-analysis</a>, which the Review of Educational Research published online in December 2021. It pooled 46 studies with 4,687 participants of all ages. Across all of them, the gap between reading and listening was an effect size of 0.07, too small to be reliable (p = .23). The differences showed up in the subgroups. When readers could go at their own speed, reading had a small advantage (g = 0.13); when the experimenter set the pace, it did not. Going at your own speed means slowing down on a hard sentence or looking back a paragraph, which a recording played straight through doesn't allow. <a href=\"/blog/text-to-speech-on-e-readers/\">Text-to-speech on an e-reader</a> lets you choose the speed and rewind, which gets part of the way there. Reading also helped more on questions of inference (g = 0.36) than on literal recall, where the two were level. There was some sign, too, that listening and reading came closer still in languages whose spelling matches their sounds closely (g = 0.001) than in languages with irregular spelling (g = 0.10). Clinton-Lisell's own conclusion is practical: accessible teaching needs both a written and an audio version of the material.</p>\n<h2 id=\"talking-books-in-the-1930s\">Talking books in the 1930s</h2>\n<p>Books made for listening are much older than any of these studies. In the United States, the <a rel=\"external\" href=\"https://www.loc.gov/nls/about/organization/history/\">Pratt-Smoot Act became law on 3 March 1931</a> and set up a national library service for blind adults at the Library of Congress. By 1933 the American Foundation for the Blind had a player ready, in two kinds: one driven by a spring and one combined with an electric radio. The Library's history quotes a description of the talking book as \"the recording on a disc of the voice of a good reader\", and the format was built for length. Records cut at 150 grooves to the inch and played at 33⅓ rpm held about thirty minutes of reading each, so a book of 60,000 words fitted on eight or nine double-sided twelve-inch records.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/bookerly-kindle-font/","url":"https://readwyn.com/blog/bookerly-kindle-font/","title":"What font does a Kindle use?","summary":"A Kindle sets books in Bookerly, a serif that Dalton Maag drew for Amazon to replace Caecilia in 2015; Kobo readers choose from about a dozen fonts or add their own.","content_html":"<p>A Kindle sets books in Bookerly unless you choose another font under the <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/how-to-use-kindle-paperwhite\">Aa button</a>. The type foundry <a rel=\"external\" href=\"http://luc.devroye.org/fonts-89157.html\">Dalton Maag</a> drew it for Amazon. The Kindle app for iOS <a rel=\"external\" href=\"https://thenextweb.com/news/buy-the-book\">got it on 27 May 2015</a>, and three weeks later Amazon built it into a new Kindle Paperwhite. It has stayed the default on the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Bookerly\">Kindles released since</a>. Before it the default was Caecilia, which The Next Web's report called a slab serif better suited to headlines than to body text.</p>\n<h2 id=\"caecilia-the-font-it-replaced\">Caecilia, the font it replaced</h2>\n<p>The first Kindle had no font menu at all. Its <a rel=\"external\" href=\"http://g-ecx.images-amazon.com/images/G/01/digital/fiona/general/Kindle_User_Guide.pdf\">user's guide</a> offers six text sizes, and Caecilia turns up only on the trademark page at the back, credited to Linotype along with Neue Helvetica. When Bookerly appeared, John Brownlee of Fast Company pointed out that slab serifs often look bold even when they aren't, and Kirk McElhearn, looking at the same page in both fonts, found Bookerly's a little lighter (<a rel=\"external\" href=\"https://mjtsai.com/blog/2015/05/28/kindle-typography-improvements/\">both collected by Michael Tsai</a>). Marco Arment was less impressed. In <a rel=\"external\" href=\"http://www.marco.org/2015/05/27/kindle-bookerly\">his view</a> the font had never been the real problem: the trouble was justified text without hyphenation, which left wide gaps between words.</p>\n<h2 id=\"how-bookerly-reached-each-kindle\">How Bookerly reached each Kindle</h2>\n<p>Bookerly wasn't brand new even in May 2015. The Next Web's report of that day said that Amazon had already put it on Fire tablets in late 2014 without an announcement, and that the iOS app was the first place outside Amazon's own hardware to get it.</p>\n<p>E-ink came next. Amazon's <a rel=\"external\" href=\"https://press.aboutamazon.com/2015/6/amazon-introduces-new-kindle-paperwhite-the-most-popular-kindle-now-even-better-still-only-119\">press release of 17 June 2015</a> launched the third Paperwhite, a $119 reader with a 300 ppi screen that shipped on 30 June, and said Bookerly would help people “read faster with less eyestrain”. The release cited no study for that. It did give one number about readers: more than half of Kindle owners read at a font size larger than the default. Owners of the Voyage and the older Paperwhites got Bookerly in software version 5.6.5 on 10 August 2015 (<a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2015/08/10/major-kindle-5-6-5-update-finally-arrives-with-bookerly-font-and-hyphens/\">The eBook Reader</a>), and <a rel=\"external\" href=\"https://www.engadget.com/2015-08-11-kindle-update-bookerly.html\">Engadget</a> listed the seventh-generation basic Kindle as well. The same update brought the layout engine that Amazon now calls <a href=\"/blog/kindle-enhanced-typesetting/\">“Enhanced Typesetting”</a>, with hyphenation, kerning, and ligatures. Amazon's example of kerning was the word “quietly”, where the tail of the y tucks in under the l.</p>\n<h2 id=\"amazon-ember-and-opendyslexic\">Amazon Ember and OpenDyslexic</h2>\n<p>Bookerly has a sans serif sibling, Amazon Ember, which Dalton Maag drew for the Kindle Oasis of 2016 and later widened into a family of condensed and monospaced cuts (<a rel=\"external\" href=\"http://luc.devroye.org/fonts-89157.html\">Luc Devroye</a>). When the Oasis came out, Ember set the menus but <a rel=\"external\" href=\"https://teleread.com/amazon-introduces-new-sans-serif-font-ember-for-kindle-oasis/index.html\">not the books</a>. By December 2016 readers could pick it for books, and update 5.8.7 gave it a bold weight, which Bookerly then lacked (<a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2016/12/01/new-amazon-ember-bold-font-added-to-kindles-with-5-8-7-update\">The eBook Reader</a>). Beyond the Kindle, Amazon's <a rel=\"external\" href=\"https://developer.amazon.com/en-US/alexa/branding/echo-guidelines/identity-guidelines/typography\">brand guidelines</a> make Ember the master typeface for all its products and services. Most Kindle books can also be read in <a rel=\"external\" href=\"https://opendyslexic.org/\">OpenDyslexic</a>, a free font under the SIL Open Font License (<a rel=\"external\" href=\"https://www.aboutamazon.com/news/books-and-authors/kindle-accessibility-features-for-all-readers\">Amazon</a>). Whether fonts like it help dyslexic readers is <a href=\"/blog/do-dyslexia-fonts-work/\">a separate question</a>.</p>\n<p>A book can also carry its own font. On a KF8 Kindle you can turn publisher fonts on or off, and the <a rel=\"external\" href=\"https://kindlegen.s3.amazonaws.com/AmazonKindlePublishingGuidelines.pdf\">Kindle Publishing Guidelines</a> (version 2026.3) warn that a book which forces one typeface on its body text can stop you choosing your own, one of the reasons <a href=\"/blog/why-ebook-fonts-look-different/\">an ebook can ignore your font setting</a>.</p>\n<h2 id=\"fonts-on-a-kobo\">Fonts on a Kobo</h2>\n<p>Kobo counts its fonts rather than naming them on its product pages. As of November 2026, the <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-clara-bw\">Kobo Clara BW</a> page promises 13 fonts and more than 50 font styles under the name iType, and the <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-elipsa-2e\">Kobo Elipsa 2E</a> page promises 12 under the name TypeGenius. A Kobo also takes <a href=\"/blog/add-fonts-to-kobo-and-kindle/\">fonts of your own</a>: copy .otf or .ttf files into a folder named fonts on the KOBOeReader drive, then, in the Text menu, tick each one under Supplemental before it shows up in the Font Face list (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/13009477876631-Load-fonts-onto-your-Kobo-eReader\">Kobo help</a>). OpenDyslexic is already on the device (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360020048733-Use-the-OpenDyslexic-font-on-your-Kobo-eReader\">Kobo help</a>), though Kobo warns that it may not work in some books or languages, and not at all in PDFs.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/can-you-inherit-ebooks/","url":"https://readwyn.com/blog/can-you-inherit-ebooks/","title":"Can you inherit Kindle books, or leave ebooks in your will?","summary":"What store terms, family sharing, Apple's Legacy Contact and US digital-asset law say about passing on Kindle and other ebooks after a death.","content_html":"<p>Not through the stores. Apple gives the plainest answer, on its <a rel=\"external\" href=\"https://support.apple.com/en-us/102631\">support page for Legacy Contact</a>, the person you can name to get into your Apple Account after you die. With the access key you generated for them and your death certificate, that person can download your photos, messages, notes, files and device backups. The books you bought are on the list of what they can't reach, next to films, music and subscriptions. Kindle and Google Play Books sell you the same thing, a right to read a book on your account, and their terms don't let you pass that right to anyone else, by will or any other route. What heirs can take over is narrower: files you hold without DRM (subject to their own licence), the device itself, and in many US states a legal route into the account.</p>\n<h2 id=\"a-purchase-that-isn-t-a-copy-you-own\">A purchase that isn't a copy you own</h2>\n<p>The US Copyright Act draws the line in <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/109\">17 U.S.C. § 109</a>, the first-sale rule that <a href=\"/blog/can-you-resell-ebooks/\">can you resell an ebook</a> explains. It lets the owner of a lawfully made copy give it away, which is what lets a paperback go to a niece or a charity shop. Subsection (d) withholds that privilege from anyone who holds the copy by rental, lease, loan or another arrangement that doesn't make them its owner.</p>\n<p>Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=201014950\">Kindle Store Terms of Use</a>, last updated on 21 October 2026, say that Kindle content is \"licensed, not sold\" to you, and forbid handing any rights in it to a third party by sale, rental, lease, sublicence or assignment. Apple's and Google's terms carry bans of their own, and the post on resale quotes all three.</p>\n<h2 id=\"what-an-executor-can-get\">What an executor can get</h2>\n<p>A will can name an executor, but that doesn't hand them the account. Most US states have a law on digital assets based on the Revised Uniform Fiduciary Access to Digital Assets Act, a model act that the Uniform Law Commission <a rel=\"external\" href=\"https://www.uniformlaws.org/HigherLogic/System/DownloadDocumentFile.ashx?DocumentFileKey=47cdcf85-e35a-995f-09ee-8030d01148a9\">approved in July 2015</a>. The map in the Commission's <a rel=\"external\" href=\"https://www.uniformlaws.org/HigherLogic/System/DownloadDocumentFile.ashx?DocumentFileKey=00d8a069-472b-d54b-edfb-c170261c9c1b\">fact sheet of October 2020</a> colours every state but California, Louisiana, Massachusetts and Oklahoma as having enacted it.</p>\n<p>Washington's version, <a rel=\"external\" href=\"https://app.leg.wa.gov/RCW/default.aspx?cite=11.120&amp;full=true\">chapter 11.120 RCW</a>, passed in 2016, shows how it works. The personal representative sends the company a written request with a certified death certificate and the letter of appointment. The company then decides, at its own discretion, whether to give full access to the account, partial access, or a copy of the assets. If the user set a direction in the company's own tool, and the tool let them change it at any time, that direction beats a contrary one in the will. Either kind of direction beats a clause in the terms of service, unless the terms asked the user for a separate, affirmative step beyond accepting them. Section 11.120.050 limits all of this: the chapter doesn't change or impair any right of the company or the user under the terms of service, and it gives the fiduciary no new or expanded rights. So an executor may be allowed to look inside a Kindle account, but the licence that came with each book stays as it was.</p>\n<h2 id=\"sharing-inside-a-family\">Sharing inside a family</h2>\n<p>Amazon Household is now Amazon Family, as a <a rel=\"external\" href=\"https://www.aboutamazon.com/news/amazon-prime/how-does-amazon-household-work\">note of 7 March 2025</a> on Amazon's old page says. <a rel=\"external\" href=\"https://www.aboutamazon.com/news/retail/how-does-amazon-family-work\">Amazon's page on it</a> says that a family holds you, one other adult for the Prime benefits, and up to four children, and that one adult must have Prime to share those benefits. The adults choose, person by person, which kinds of content in the Family Library to share: Kindle ebooks, audiobooks, apps and games. A member who leaves loses the shared Family Library content and can switch to a new Amazon Family only once every 12 months. Amazon's page doesn't say what happens to shared books when the account that bought them is closed after a death. Apple's <a rel=\"external\" href=\"https://www.apple.com/legal/internet-services/itunes/us/terms.html\">Media Services terms</a> cap Family Sharing, and the sharing of purchases, at six members.</p>\n<h2 id=\"files-without-drm\">Files without DRM</h2>\n<p>A DRM-free EPUB or PDF is a file on a disk, so whoever inherits the computer can open it in any reader. Being able to open it says nothing about the licence. Leanpub, one of the <a href=\"/blog/where-to-buy-drm-free-ebooks/\">stores that sell without DRM</a>, states in its <a rel=\"external\" href=\"https://leanpub.com/terms\">terms</a> that it uses no copy control at all, yet the same terms (updated 19 October 2026) say the reader's licence is non-transferable and the book may not be sold or its rights transferred. Public-domain books need no one's permission. <a rel=\"external\" href=\"https://standardebooks.org/about\">Standard Ebooks</a> dedicates its own work on each file to the public domain through CC0, and its texts are believed to be in the US public domain, so in the US its files can go to anyone, inheritance included.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/can-you-resell-ebooks/","url":"https://readwyn.com/blog/can-you-resell-ebooks/","title":"Can you resell an ebook?","summary":"Why you can't legally sell an ebook second-hand in the US or the EU: what the store terms say, and the ReDigi and Tom Kabinet rulings.","content_html":"<p>When you finish a paperback, you can take it to a second-hand shop. When you finish an ebook on your phone or e-reader, you can't do the same with the file, at least not lawfully in the US or the EU. Amazon, Apple and Google forbid resale in their store terms, and courts on both sides of the Atlantic have ruled that the rule which lets you resell a printed book doesn't cover a download.</p>\n<p>Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=201014950\">Kindle Store Terms of Use</a>, updated on 21 October 2026, say you may not sell, rent, lease or sublicense Kindle content, or assign any rights in it to a third party. Apple's media services terms, <a rel=\"external\" href=\"https://www.apple.com/legal/internet-services/itunes/us/terms.html\">last updated on 14 September 2026</a>, cover Apple Books and say you agree not to \"rent, loan, sell, share, or distribute\" the content. <a rel=\"external\" href=\"https://play.google.com/about/play-terms/\">Google Play's terms of 14 October 2026</a> let you keep and display copies on your own devices for private use, and bar you from selling, renting, leasing or transferring content to anyone else unless the terms expressly allow it. Whether you own the book at all is <a href=\"/blog/do-you-own-your-kindle-books/\">a separate question</a>.</p>\n<p>In the US, resale of a paper book rests on the first-sale doctrine in <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/109\">section 109 of the Copyright Act</a>, which lets the owner of a lawfully made copy sell or give away that copy. The snag is that a file can't change hands without being copied. ReDigi tried to get round that for music: its Music Manager software checked that a track had been bought lawfully, then moved it to ReDigi's server in small blocks so that it never existed on two devices at once. The Court of Appeals for the Second Circuit <a rel=\"external\" href=\"https://www.copyright.gov/fair-use/summaries/capitol-records-llc-redigi-inc-no.16-2321-2nd-cir.dec.12.2018.pdf\">ruled on 12 December 2018</a> that this still made unauthorised reproductions, which section 109 doesn't excuse, and that it wasn't fair use. By then the district court had entered a stipulated judgment of $3.5 million in damages and barred ReDigi from operating. The case was about songs, but the reasoning turned on the copy that every file transfer makes.</p>\n<p>The Copyright Office had already turned the idea down. Its <a rel=\"external\" href=\"https://www.copyright.gov/reports/studies/dmca/sec-104-report-vol-1.pdf\">DMCA Section 104 report of August 2001</a> recommended no change to section 109, pointing out that digital copies don't wear out and that it was unclear whether software to delete the seller's copy worked, or whether the market would pay for it.</p>\n<p>Tom Kabinet, a Dutch company, sold second-hand ebooks through a reading club on its website, and two associations of Dutch publishers asked the District Court of The Hague to stop it. The EU's highest court got to the same answer as the Second Circuit by another route. On <a rel=\"external\" href=\"https://curia.europa.eu/jcms/upload/docs/application/pdf/2019-12/cp190159en.pdf\">19 December 2019 the Court of Justice ruled</a> in case C-263/18 that supplying an ebook by download for permanent use is a communication to the public under Directive 2001/29, not a distribution, so the exhaustion rule that frees the resale of a paper book doesn't apply. Ebooks don't deteriorate, the court said, so a used copy is a perfect substitute for a new one. And since an ebook's licence usually lets the buyer read it only on their own equipment, the club was reaching a new public.</p>\n<p>A downloaded program, by contrast, can be resold. In <a rel=\"external\" href=\"https://curia.europa.eu/jcms/upload/docs/application/pdf/2012-07/cp120094en.pdf\">UsedSoft v Oracle on 3 July 2012</a>, the same court held that selling a downloaded program with a licence of unlimited length exhausts the distribution right under the software directive, so the buyer can resell that copy even if the licence forbids transfer. Ebooks come under Directive 2001/29 instead, where the court found in Tom Kabinet that exhaustion was meant only for tangible objects such as printed books.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/check-epub-with-epubcheck/","url":"https://readwyn.com/blog/check-epub-with-epubcheck/","title":"How to check an EPUB for errors","summary":"Run the free EPUBCheck tool on your EPUB to find the errors that make stores reject it. How to install it, and what its errors, warnings and codes mean.","content_html":"<p>Run EPUBCheck, the free checker that <a rel=\"external\" href=\"https://itunespartner.apple.com/books/support/9-prepare-book\">Apple Books</a> requires every submitted EPUB to pass. It's a <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/\">W3C project</a> that the DAISY Consortium maintains, and by its <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/docs/getting-started/\">own account</a> most retailers won't take a book that fails it. Download the zip from the project's <a rel=\"external\" href=\"https://github.com/w3c/epubcheck/releases\">releases page</a>, unzip it, open Terminal in that folder and type <code>java -jar epubcheck.jar yourbook.epub</code>.</p>\n<h2 id=\"installing-it\">Installing it</h2>\n<p>The official download holds <code>epubcheck.jar</code> and a <code>lib/</code> folder, and the <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/docs/installation/\">installation guide</a> says the two must stay together, so move the whole folder, not the jar alone. You also need Java. That guide still names Java 7, but the <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/releases/\">release notes</a> made Java 8 or newer a requirement from the 5.0 series on. Type <code>java -version</code> to see what you have.</p>\n<p>With Homebrew on a Mac, <code>brew install epubcheck</code> gives you an <code>epubcheck</code> command that works in any folder, and the <a rel=\"external\" href=\"https://formulae.brew.sh/formula/epubcheck\">formula</a> carries the current release. The Ubuntu and Debian package is another matter. <code>apt-get install epubcheck</code> works, but at the start of November 2026 the oldstable, stable, testing and unstable suites of <a rel=\"external\" href=\"https://tracker.debian.org/pkg/epubcheck\">Debian</a> and every <a rel=\"external\" href=\"https://packages.ubuntu.com/search?keywords=epubcheck&amp;searchon=names&amp;exact=1\">Ubuntu</a> release up to 26.10 still ship 4.2.6 (Debian's oldoldstable has the older 4.2.4). Version 4.2.6 is a build from <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/releases/\">June 2021</a> that checks the EPUB 3.2 rules. Volunteers look after those packages, not the project, as the installation guide points out.</p>\n<p>The current release is 5.4.0, published on 15 September 2026; its <a rel=\"external\" href=\"https://github.com/w3c/epubcheck/releases/tag/v5.4.0\">release notes</a> describe it as checking against EPUB 3.4, which in November 2026 was still a W3C <a rel=\"external\" href=\"https://www.w3.org/TR/epub-34/\">Candidate Recommendation</a>, not a finished standard.</p>\n<p>There's no updater: you upgrade by replacing the folder with the new zip. If it crashes with a <code>StackOverflowError</code>, the <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/docs/troubleshooting/\">troubleshooting page</a> suggests a bigger Java thread stack, for example <code>java -Xss1024k -jar epubcheck.jar book.epub</code>.</p>\n<p>You don't have to touch Terminal at all. The project's <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/docs/apps-and-tools/\">apps and tools page</a> lists graphical front ends that run EPUBCheck for you, among them pagina EPUB-Checker for Windows, macOS and Linux, and ECheck for macOS, plus EPUBCheck plugins for calibre and Sigil. It also lists epubcheck.net, a website you upload the book to, and epubcheck-standalone, an experimental build compiled to JavaScript that checks the file inside your browser without uploading it.</p>\n<h2 id=\"the-commands-you-ll-use\">The commands you'll use</h2>\n<pre><code>epubcheck book.epub\nepubcheck --mode exp book-folder/ --save\nepubcheck -e book.epub\nepubcheck --json report.json book.epub\n</code></pre>\n<p>The first line checks a finished file. A clean book ends with a count of 0 fatals, 0 errors, 0 warnings and 0 infos; anything else is listed with a code, the file inside the EPUB that has the problem, and usually a line and a column. The second line checks an unzipped book folder, which suits a book you're still editing, and <code>--save</code> then packs the folder into an <code>.epub</code>. With <code>-e</code> you see only errors and fatal errors; <code>-u</code> goes the other way and adds the minor “usage” notes that are hidden by default. The JSON (<code>--json</code>) and XML (<code>-o</code>) reports keep at most 25 copies of any one message unless you add <code>--maxOfEachMessage unlimited</code>; there's also an XMP report (<code>-x</code>). The <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/docs/cli/\">command line reference</a> lists every option.</p>\n<p>In a script, watch the exit status. EPUBCheck returns 1 only when it reports an error and 0 when it finds nothing worse than warnings, so add <code>--failonwarnings</code> if a warning should stop your build.</p>\n<h2 id=\"errors-warnings-and-the-rest\">Errors, warnings and the rest</h2>\n<p>Every message starts with its severity, and the <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/docs/messages/\">messages reference</a> ties the two that matter to the wording of the EPUB specification. An error means the book breaks a requirement written as MUST, MUST NOT or REQUIRED. A warning means it ignores a recommendation, written as SHOULD, SHOULD NOT or RECOMMENDED. A fatal error means EPUBCheck couldn't read any further; the reference puts that down to a bug in the checker or to a condition it didn't foresee, such as an I/O error. Info and usage messages are notes about the book, not faults.</p>\n<p>The project's <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/docs/running/\">guide to running it</a> takes a line like this one apart: <code>ERROR(RSC-001): ./book.epub(-1,-1): File \"EPUB/missing.xhtml\" could not be found.</code> Look the code up in the messages reference. The numbers in brackets are the line and column, and -1,-1 means the problem has no single place in a file, as with a file that isn't there at all.</p>\n<p>A pass says nothing about the writing. EPUBCheck's <a rel=\"external\" href=\"https://www.w3.org/publishing/epubcheck/docs/getting-started/\">introduction</a> lists what it inspects (the container, the package file and its metadata, the markup, and the links between files) and says it won't catch typos or the wrong picture.</p>\n<h2 id=\"some-codes-and-their-fixes\">Some codes and their fixes</h2>\n<p><code>PKG-006</code> (the messages reference writes this and some other codes with an underscore, <code>PKG_006</code>) says the <code>mimetype</code> file is missing or isn't the first entry in the zip. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/\">EPUB 3.3 specification</a> requires that file to come first, uncompressed, containing nothing but <code>application/epub+zip</code> with no spaces or line break around it. If you repack a book folder by hand, let EPUBCheck do the packing with <code>--mode exp --save</code>. (<a href=\"/blog/what-is-inside-an-epub-file/\">What is inside an EPUB file</a> covers the rest of the container.)</p>\n<p><code>RSC-005</code> reads “error while parsing file” and is followed by the real complaint from the schema, such as an attribute that isn't allowed on an element. The text after the colon is the part to fix.</p>\n<p><code>RSC-007</code> means a page refers to an image, stylesheet or chapter that isn't in the EPUB. Its neighbour <code>RSC-008</code> means the file is there but missing from the manifest in the package file. <code>RSC-012</code> is a link to an <code>#anchor</code> that no element in the target page carries, and <code>NCX-001</code> means the identifier in the NCX (the older table of contents file) differs from the one in the package file.</p>\n<p>File names get smaller flags. <code>PKG-010</code> is a warning about a space in a name, and <code>PKG-012</code>, a usage note, is about letters outside ASCII; the messages reference says either can trip up older reading systems.</p>\n<h2 id=\"which-stores-ask-for-a-pass\">Which stores ask for a pass</h2>\n<p>Apple's <a rel=\"external\" href=\"https://itunespartner.apple.com/books/support/9-prepare-book\">publishing guide</a> asks for the most recent version of EPUBCheck, which at the time of writing means 5.4.0, not the 4.2.6 that <code>apt-get</code> installs or a zip you unpacked a few years ago. Google's <a rel=\"external\" href=\"https://support.google.com/books/partner/answer/3316879?hl=en-GB\">Play Books partner help</a> tells publishers to run EPUBCheck before uploading and shows processing errors under the book's Content tab in Partner Centre; it takes EPUB 2 and EPUB 3, and prefers 3.3.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/colour-e-ink-for-reading/","url":"https://readwyn.com/blog/colour-e-ink-for-reading/","title":"Is colour e-ink any good for reading?","summary":"Colour e-ink helps with comics, covers, and coloured highlights, but it shows colour at half the sharpness and makes plain text a little greyer.","content_html":"<p>Partly. Colour e-ink is good for comics, picture books, maps, covers, and highlights you can tell apart by colour, but it costs something on plain text. The colour layer in the Kobo and Kindle colour models sits over an ordinary black-and-white screen, so the page comes out a shade darker and text loses a little contrast. Colour itself is shown at half the sharpness of text, and it looks muted next to a tablet. If you mostly read novels, a black-and-white e-reader still gives you the cleaner page.</p>\n<h2 id=\"how-kaleido-makes-colour\">How Kaleido makes colour</h2>\n<p>The colour readers from Kobo and Amazon use E Ink's Kaleido 3. E Ink's <a rel=\"external\" href=\"https://www.eink.com/news/detail/e_ink_launches_e%20ink%20kaleido_3_the_next_generation_of_print_color_epaper_technology\">launch announcement of April 2022</a> describes it as a colour filter array laid over its Carta black-and-white ink film. The ink underneath is the same <a href=\"/blog/how-e-ink-works/\">black-and-white particles</a> as on a monochrome reader; the filter tints what you see.</p>\n<p>That design is why there are two resolutions. E Ink's figures for Kaleido 3 are 300 pixels per inch for black and white and 150 for colour, with 16 levels of grey and 4,096 colours. The company says it raised colour saturation by 30 per cent over Kaleido Plus, whose colour resolution was 100 ppi (<a rel=\"external\" href=\"https://www.eink.com/brand/detail/Kaleido3\">E Ink's Kaleido 3 page</a>). Halving the resolution in each direction means one colour dot covers the area of four black-and-white pixels. Black text is drawn at the full 300 ppi; a coloured illustration gets a quarter as many dots. The Carta ink under the filter has <a href=\"/blog/e-ink-carta-pearl-kaleido/\">version numbers of its own</a>.</p>\n<p>Kobo's first colour readers, the Libra Colour and the Clara Colour, both use Kaleido 3; <a rel=\"external\" href=\"https://displaydaily.com/?p=380743\">Display Daily reported</a> pre-orders opening in April 2024, with shipping from 30 April. Amazon announced its first colour model, the <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/kindle-color-specs-price-announce\">Kindle Colorsoft Signature Edition</a>, on 16 October 2024 (<a rel=\"external\" href=\"https://www.macrumors.com/2024/10/16/amazon-kindle-color-display/\">MacRumors</a>). Amazon's page talks of an oxide backplane and a light guide with micro-deflectors but names no panel. <a rel=\"external\" href=\"https://www.engadget.com/mobile/tablets/kindle-colorsoft-review-the-missing-link-in-amazons-ereader-lineup-131529685.html\">Engadget's review</a> says the screen is based on Kaleido 3 with Amazon's own changes to the stack, and lists the same 300 and 150 ppi figures.</p>\n<h2 id=\"what-the-colour-layer-does-to-text\">What the colour layer does to text</h2>\n<p>The filter is always there, even on a page with no colour on it. When Kobo launched its colour models, it told <a rel=\"external\" href=\"https://goodereader.com/blog/electronic-readers/everything-you-need-to-know-about-the-new-kobo-e-readers\">Good e-Reader</a> that Kaleido 3 makes the grey background darker, and suggested keeping the front light at about 10 per cent.</p>\n<p>Engadget's reviewer, after reading black-and-white text on the Colorsoft beside an older Paperwhite, put a slight loss of sharpness and contrast among the cons and gave the Paperwhite a small lead. On a long novel, though, it's a cost with nothing on the other side of the scale, because the book has no colour to show.</p>\n<h2 id=\"where-colour-earns-its-place\">Where colour earns its place</h2>\n<p>Covers are the first thing you notice: the library and the store show in colour instead of grey. Highlights are the second. The Colorsoft lets you highlight in yellow, orange, blue, or pink and search your highlights by colour, according to <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/kindle-color-specs-price-announce\">Amazon's announcement</a>. On the Kobo Libra Colour that Readwyn's Kobo research used, the four colours are yellow, pink, blue, and green, stored in the device's database as the numbers 0 to 3. Readwyn keeps that colour when it syncs highlights in either direction (more in <a href=\"/blog/kobo-highlight-colours/\">what Kobo highlight colours mean</a>).</p>\n<p>Then there are books that were printed in colour: comics, graphic novels, children's books, cookbooks, charts, and maps. Engadget's verdict was that the Colorsoft suits readers who mostly read comics and graphic novels, and that for standard reading the Paperwhite is the better buy. On a 6 or 7 inch screen a full comic page is also small, and a <a href=\"/blog/double-page-spreads-in-comics/\">double-page spread</a> is smaller still.</p>\n<p>Don't expect a tablet's colours. A long <a rel=\"external\" href=\"https://myereader.substack.com/p/kobo-libra-colour-review-e-ink-colour\">review of the Libra Colour</a> on the My eReader blog calls them muted and less saturated than on an LCD tablet. The Colorsoft also has two colour styles, Standard and Vivid; <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">Amazon's user guide</a> says Vivid makes colours punchier and can lose detail in places. In an informal poll for Engadget's original review, everyone at the site preferred the Kobo's screen, and the reviewer thought its slightly warmer tone made the colours look a little more saturated. On a replacement Colorsoft that Amazon sent at the end of 2024, she found the difference in warmth much smaller.</p>\n<h2 id=\"gallery-3-the-slower-alternative\">Gallery 3, the slower alternative</h2>\n<p>E Ink makes a second kind of colour screen that puts colour into the ink itself. <a rel=\"external\" href=\"https://www.eink.com/news/detail/E%20Ink_Launches_E%20Ink_Gallery_3_Color_ePaper_for_Sustainable_Digital_Reading\">Gallery 3</a>, announced on 25 April 2022, uses four particles (cyan, magenta, yellow, and white) and can show colour at every pixel, at 300 ppi. There's <a rel=\"external\" href=\"https://www.eink.com/tech/detail/How_it_works\">no filter over the page</a>: E Ink puts all the coloured particles in every pixel.</p>\n<p>The price is speed. E Ink's own figures for Gallery 3 are 350 ms for a black-and-white page update, 500 ms in its fast colour mode, 750 to 1,000 ms in standard colour, and 1,500 ms for its best colour. It also works only between 0 and 50 °C. One device built on it is the reMarkable Paper Pro, a tablet for writing notes; in September 2024 <a rel=\"external\" href=\"https://blog.eink.com/new-remarkable-paper-pro\">E Ink described</a> its 11.8 inch Gallery 3 screen as 2160 by 1620 pixels, which works out at 229 ppi rather than 300.</p>\n<h2 id=\"what-colour-costs\">What colour costs</h2>\n<p>On Kobo's US store in November 2026, the <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-clara-colour\">Clara Colour</a> is $179.99 and the <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-clara-bw\">Clara BW</a>, with a black-and-white Carta 1300 screen, is $159.99. The gap was also $20 at launch, when Display Daily listed the two at $149.99 and $129.99. The Kindle Colorsoft Signature Edition went on pre-order at $279.99 in October 2024.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/commonplace-book/","url":"https://readwyn.com/blog/commonplace-book/","title":"What is a commonplace book, and are ebook highlights one?","summary":"A commonplace book is a notebook of copied passages filed under headings. What Erasmus advised, how John Locke indexed his, and how ebook highlights compare.","content_html":"<p>Partly. A commonplace book is a notebook into which a reader copies passages worth keeping (a line of verse, a fact, a proverb, an argument) and files each one under a heading, so that everything on one subject ends up in one place, whichever book it came from. Highlights on an e-reader do the collecting. They don't do the filing. They stay sorted by book and by position in the book, and nobody decides which subject a passage belongs to, which was the step that made the old notebooks useful.</p>\n<p>The name is a translation of the Latin <em>locus communis</em>, itself taken from the Greek <em>koinòs tópos</em>: a <a rel=\"external\" href=\"https://en.wiktionary.org/wiki/commonplace\">passage of general use</a> that a writer or speaker could apply to many subjects. Erasmus wrote <em>De copia</em>, a book on rhetoric, for John Colet to teach at St Paul's School in London, and Josse Bade printed the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Copia:_Foundations_of_the_Abundant_Style\">first edition in Paris in 1512</a>. Its advice on gathering examples has the student set up broad headings such as Piety and Impiety, with narrower ones like Gratitude beneath them, and enter what he reads under the right one, as the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Commonplace_book\">Wikipedia article on commonplace books</a> summarises it.</p>\n<h2 id=\"locke-s-index\">Locke's index</h2>\n<p>John Locke worked out his own system and handed it to Le Clerc, who printed it in French in 1686 in the second volume of the <em>Bibliothèque universelle</em>. An English translation came out in London in 1706 as <a rel=\"external\" href=\"https://archive.org/details/gu_newmethodmaki00lock\"><em>A New Method of Making Common-Place-Books</em></a>. After a dedication and Le Clerc's praise of the method comes Locke's letter to Monsieur Toinard, which promises a method \"an exact Index of which may be made in Two Pages\".</p>\n<p>The index lists classes of headings rather than the headings themselves. Locke ruled the first two pages of a blank book into a grid with one cell for each initial letter paired with each vowel. He dropped K, Y and W, which C, I and U could stand in for, and put Q (always followed by U) in the fifth cell of Z. In the alphabetical order of his day, I and J still counted as <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/J\">one letter</a>, as they did in English dictionaries until about 1820, and U and V as <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/U\">another</a>, so the grid had 20 letters and five vowels: 100 classes. To file something, you choose a heading whose first word carries the subject and note its first letter and the vowel after it. Epistle is E I. If that cell is empty, you take the next blank left-hand page, write its number in the cell, and give that page and the one facing it to every E I heading. Locke's examples for that spread are Epicurus, Ebionites, Epigram and Edict, all visible at once when the book lies open. When a spread fills, a number at the foot of the page sends you to the continuation. He had tried a finer grid of 500 classes with a second vowel and preferred the plain one, and he wrote that in 25 years of use he never once met a heading that began Z U.</p>\n<h2 id=\"how-highlights-compare\">How highlights compare</h2>\n<table><thead><tr><th></th><th>Commonplace book (Locke, 1706)</th><th>Ebook highlights (Kobo, 2026)</th></tr></thead><tbody>\n<tr><td>How a passage gets in</td><td>Copied out by hand</td><td>Selected on the screen and saved</td></tr>\n<tr><td>What decides where it goes</td><td>A heading the reader chooses</td><td>The book and the place in it</td></tr>\n<tr><td>How you find it again</td><td>A two-page index by letter and vowel</td><td>A list of notes inside each book</td></tr>\n<tr><td>What it holds</td><td>Passages from any source, and remarks</td><td>Passages from that book, with an optional note</td></tr>\n<tr><td>Taking it elsewhere</td><td>It is already a book of its own</td><td>An export file from Kobo.com</td></tr>\n</tbody></table>\n<p>Readwyn's Notebook gathers the highlights of every book in a library, including those synced from a Kobo, in one place. Each book's highlights appear in reading order, grouped by chapter, with a chip for each colour used and a search field. It's still the right-hand column of the table, arranged by book rather than by subject.</p>\n<p>On a Kobo you press and hold a word, then drag the circles at each end over the passage. The Clara Colour and the Libra Colour also let you change a highlight's colour, and PDFs can't be highlighted at all, according to <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017481334-Highlight-text-on-your-Kobo-eReader\">Kobo's help page on highlighting</a>. The collected highlights for a book sit behind the Notes icon of the Reading Menu, mixed in with bookmarks and notes (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017640053-View-your-highlights-on-your-Kobo-eReader\">Kobo help</a>). To get them out, Kobo.com offers Export Annotations under My Books, as PDF, HTML, plain text or Markdown. In October 2026 its <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/29991333812631-Export-annotations-from-your-books\">export page</a> says this works only for books bought from Kobo or read through Kobo Plus, may depend on the author's approval, and keeps each generated file for 30 days. Kindle highlights have a separate route, covered in <a href=\"/blog/export-kindle-highlights/\">where your Kindle highlights are</a>.</p>\n<p>Nothing in the format of a digital highlight rules out a heading. The W3C <a rel=\"external\" href=\"https://www.w3.org/TR/annotation-model/\">Web Annotation Data Model</a>, a Recommendation since 23 February 2017, can record a highlight as a copy of the exact text plus a snippet of what comes before and after it. Its list of motivations has \"highlighting\", and beside it \"tagging\" and \"classifying\", the second for when a reader means to \"classify the Target as something\", which is the job Locke's headings did.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/controlled-digital-lending/","url":"https://readwyn.com/blog/controlled-digital-lending/","title":"What is controlled digital lending, and why did a court stop it?","summary":"Controlled digital lending lets a library lend one scan for each print copy it owns. How it began, and why US courts ruled against the Internet Archive's version.","content_html":"<p>On 4 September 2024 a US appeals court ruled that the Internet Archive could not scan in-copyright print books and lend the scans online, even when it lent only one scan for each print copy it kept in storage. That practice is called controlled digital lending, or CDL. A library digitises a book it owns, keeps the print copy out of circulation, and lends the scan to one reader at a time for a fixed period, so the number of copies on loan never exceeds the number it bought or was given. Its supporters argue that this is ordinary library lending in a new format. In <a rel=\"external\" href=\"https://storage.courtlistener.com/recap/gov.uscourts.ca2.60988/gov.uscourts.ca2.60988.306.1.pdf\">Hachette v. Internet Archive</a>, the Second Circuit disagreed: for books that publishers sell or license as ebooks, it held that CDL isn't fair use.</p>\n<h2 id=\"where-the-idea-came-from\">Where the idea came from</h2>\n<p>The Archive began lending scans through Open Library in June 2010. Its <a rel=\"external\" href=\"https://blog.archive.org/2010/06/28/digital-lending-library/\">announcement</a> offered two-week loans of books that weren't commercially available, provided nobody else had the copy out, and sent readers to OverDrive for the more than 70,000 current titles that libraries already licensed.</p>\n<p>The legal theory came later. A <a rel=\"external\" href=\"https://cdn.prod.website-files.com/5f308d0beb0dcb479f2d2b7c/67dc506812266e84ff926d62_FINAL-White-Paper-on-Controlled-Digital-Lending-of-Library-Books2.pdf\">2018 white paper</a> credits Michelle Wu of Georgetown Law with first exploring it, in a 2011 article in Law Library Journal. In 2018 a group of librarians and lawyers that included Wu and the Archive's policy counsel, Lila Bailey, published the <a rel=\"external\" href=\"https://controlleddigitallending.org/position-statement-on-controlled-digital-lending-by-libraries\">Position Statement on Controlled Digital Lending by Libraries</a>. Its example is a library with three print copies that digitises one: it can lend three scans, or two scans and a print book, but never four copies in total.</p>\n<p>The white paper, by David Hansen of Duke and Kyle Courtney of Harvard, sets out the statement's six conditions. The library must have acquired the book lawfully and must own it rather than license it. It can't circulate more copies, in any format, than it owns. Each scan goes to one user at a time, for a loan period like a print loan, and DRM stops readers from copying it. The argument rests on the first-sale doctrine (section 109 of the US Copyright Act, which lets the owner of a copy lend it) and on fair use.</p>\n<p>Also in 2018, the Archive started the Open Libraries Project: for each scanned title that a partner library owned and kept out of circulation, the Archive allowed one more loan at a time. <a rel=\"external\" href=\"https://storage.courtlistener.com/recap/gov.uscourts.ca2.60988/gov.uscourts.ca2.60988.306.1.pdf\">According to the court</a>, 62 libraries had joined by December 2021.</p>\n<h2 id=\"the-emergency-library-and-the-lawsuit\">The emergency library and the lawsuit</h2>\n<p>On 24 March 2020, with schools and libraries closed by COVID-19, the Archive <a rel=\"external\" href=\"https://blog.archive.org/2020/03/24/announcing-a-national-emergency-library-to-provide-digitized-books-to-students-and-the-public/\">opened a National Emergency Library</a>. It dropped the waiting lists on the 1.4 million books in its lending library until 30 June or the end of the US national emergency, whichever came later. The appeal court found that up to 10,000 people could then borrow one scan at once, whatever the number of print copies behind it. The Archive itself called this a departure from CDL.</p>\n<p>Hachette, HarperCollins, Wiley, and Penguin Random House sued on 1 June 2020 in the Southern District of New York. The complaint named <a rel=\"external\" href=\"https://storage.courtlistener.com/recap/gov.uscourts.ca2.60988/gov.uscourts.ca2.60988.306.1.pdf\">127 books</a>, from The Bell Jar to Wind Power for Dummies. Nine days later the Archive <a rel=\"external\" href=\"https://blog.archive.org/2020/06/10/temporary-national-emergency-library-to-close-2-weeks-early-returning-to-traditional-controlled-digital-lending/\">said</a> it would close the emergency library on 16 June, two weeks early, and return to \"traditional controlled digital lending\". The suit carried on, because it challenged ordinary CDL too.</p>\n<p>Judge John Koeltl <a rel=\"external\" href=\"https://www.eff.org/files/2023/03/29/188_opinion.pdf\">granted summary judgment</a> to the publishers on 24 March 2023, three years to the day after the emergency library opened, and found that all four fair-use factors favoured them. On 11 August 2023 the parties filed a <a rel=\"external\" href=\"https://publishers.org/news/publishers-and-internet-archive-submit-negotiated-judgment-with-permanent-injunction-to-district-court-in-hachette-book-group-et-al-v-internet-archive/\">negotiated judgment</a>: a permanent injunction covering the four publishers' whole catalogues of books available as ebooks, not just the 127, and a confidential sum that the Archive would pay only if it lost on appeal.</p>\n<h2 id=\"what-the-appeal-court-decided\">What the appeal court decided</h2>\n<p>Judges Menashi, Robinson, and Kahn heard the appeal on 28 June 2024. Their <a rel=\"external\" href=\"https://storage.courtlistener.com/recap/gov.uscourts.ca2.60988/gov.uscourts.ca2.60988.306.1.pdf\">opinion</a> corrected Koeltl on one point. The Archive is a nonprofit and charges nothing for loans, so its use was non-commercial, even with a Donate link on its site and a share of sales made through its links to Better World Books. That didn't rescue it. The scans did the same job as the books and added nothing new, and they competed with the ebook licences that publishers sell to libraries through distributors such as OverDrive. The opinion lists those terms: Hachette sells two-year licences, and HarperCollins sells licences that expire after 26 loans, which is one reason <a href=\"/blog/why-library-ebooks-have-waitlists/\">library ebooks have waiting lists</a>.</p>\n<p>The court also went back to its own <a href=\"/blog/google-books-lawsuit/\">Google Books decision</a>. Snippets in search results were fair use there, but the 2015 ruling had said that authors would have a strong claim against putting whole digitised books before the public, which is what the Archive did. In a footnote it refused the Archive's fallback request to keep lending scans of books it owned itself, outside the partner scheme: the analysis, it said, would not be substantially different.</p>\n<p>On 4 December 2024 the Archive <a rel=\"external\" href=\"https://blog.archive.org/2024/12/04/end-of-hachette-v-internet-archive/\">announced</a> that it wouldn't ask the Supreme Court to hear the case. Its <a rel=\"external\" href=\"https://help.archive.org/help/why-are-so-many-books-listed-as-borrow-unavailable-at-the-internet-archive/\">help page</a> puts the books it can no longer lend at more than 500,000, a count that includes titles pulled at the demand of other publishers in the Association of American Publishers. The Archive had <a rel=\"external\" href=\"https://blog.archive.org/2023/08/17/what-the-hachette-v-internet-archive-decision-means-for-our-library/\">agreed in August 2023</a> to take down any AAP member's ebook titles by the same process, in return for the association not backing further suits over CDL. The judgment reaches only books that are available as ebooks, and the same post said that out-of-print books would stay on loan, along with interlibrary loan and accessible copies for readers with print disabilities.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/dark-mode-for-reading-research/","url":"https://readwyn.com/blog/dark-mode-for-reading-research/","title":"Is dark mode better for reading?","summary":"What studies of light and dark pages found about reading accuracy, small type, room lighting, low vision, myopia, and battery life on OLED screens.","content_html":"<p>Turn on dark mode in a reading app and the page goes black with pale letters. For readers with normal sight, that page is slightly harder to read accurately than black text on white. Researchers call this the positive polarity advantage, and lab studies keep finding it. It shows up in speed and accuracy, not in how tired people say their eyes feel. A dark page can still make sense for some eyes, and on a phone with an OLED screen it saves battery at high brightness.</p>\n<h2 id=\"what-the-lab-studies-found\">What the lab studies found</h2>\n<p>Dark text on a light background is positive polarity; light text on dark is negative polarity. Much of the research comes from the experimental psychology institute of Heinrich Heine University Düsseldorf. In 2007 Axel Buchner and Nadine Baumgartner had people proofread under both polarities, in a dark room and under office lighting, and <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/17510822/\">dark text on light won both times</a>. Heart rate, breathing, skin conductance, and self-reported eyestrain and fatigue didn't change with the polarity, so neither page took more effort that they could measure.</p>\n<p>Older readers were the open question, because light scatters in an ageing lens and that might favour a dark page. The light page still won an acuity test and a proofreading task for younger adults and for older adults alike (<a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/23654206/\">Piepenbrock and colleagues, 2013</a>). The younger group was 18 to 33 and the older 60 to 85, according to Nielsen Norman Group's <a rel=\"external\" href=\"https://www.nngroup.com/articles/dark-mode/\">account of the study</a>. A <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/25141597/\">follow-up in Human Factors</a> set the text at 8, 10, 12, and 14 pt, and the light page's lead grew as the type got smaller.</p>\n<p>The likely cause is the pupil. A bright page makes it contract, and a smaller pupil gives a sharper image on the retina, as a smaller aperture does in a camera. A 2014 study by the Düsseldorf group <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/25135324/\">measured smaller pupils</a> and better proofreading with the light page.</p>\n<p>Many people switch to dark mode at night, and that's where one study found it did worst. Researchers at the MIT AgeLab <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/28166901/\">timed how long people needed to read a word</a> at two type sizes, in near darkness and in light like daylight. Light text on dark in a dark room needed the most time. In bright light, both polarities did well, and so did dark text on light in the dark room. The authors took the result as support for the pupil explanation.</p>\n<h2 id=\"battery-life\">Battery life</h2>\n<p>A dark page saves power only on a screen whose pixels make their own light. An LCD lights every pixel from a backlight, and an OLED has none. Purdue University researchers <a rel=\"external\" href=\"https://www.purdue.edu/newsroom/releases/2021/Q3/dark-mode-may-not-save-your-phones-battery-life-as-much-as-you-think,-but-there-are-a-few-silver-linings.html\">measured six popular Android apps</a>, among them Google Maps and YouTube, on the Pixel 2, Pixel 4, Pixel 5, and Moto Z3 in 2021. At 30 to 50% brightness, dark mode saved 3 to 9% of the power on average; at full brightness, 39 to 47%. Indoors, automatic brightness tends to sit at about 30 to 40%, so the usual saving is at the small end.</p>\n<h2 id=\"eyes-that-may-prefer-a-dark-page\">Eyes that may prefer a dark page</h2>\n<p>Some readers do better with white on black. Contrast polarity was one of the variables in Gordon Legge's 1985 study of 16 low-vision readers (<a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/4013092/\">abstract</a>), and <a rel=\"external\" href=\"https://www.nngroup.com/articles/dark-mode/\">NN/g's summary</a> reports that the seven of them with cloudy ocular media, such as cataracts, read faster with the dark page. Android's <a rel=\"external\" href=\"https://developer.android.com/develop/ui/views/theming/darktheme\">developer guide to dark theme</a> also says a dark theme helps people with low vision and people who are sensitive to bright light. The page doesn't have to follow the rest of the app: in Readwyn, the Text panel of the reader sets it to App, Light, or Dark on its own, and <a href=\"/blog/dark-mode-for-ebooks-on-mac/\">another post gives the steps</a>.</p>\n<p>Astigmatism can push the other way. <a rel=\"external\" href=\"https://www.allaboutvision.com/conditions/computer-vision-syndrome/digital-eye-strain/is-dark-mode-better-for-eyes/\">All About Vision</a>, in a page reviewed by an optometrist, says people with astigmatism or myopia can see halation, a haze where light letters spread into the black around them, and that it's worse in dark mode.</p>\n<p>A small study from the University of Tübingen looked at myopia. <a rel=\"external\" href=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC6052140/\">Seven adults aged 23 to 29</a> read for an hour while the researchers measured the choroid, the layer of blood vessels behind the retina. It thinned by about 16 µm with black text on white and thickened by about 10 µm with white on black. Other studies tie a thinner choroid to the development of myopia, so the authors propose that dark text on light may encourage it. The study measured a change over one hour, not myopia itself.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/distributed-proofreaders/","url":"https://readwyn.com/blog/distributed-proofreaders/","title":"How volunteers turn scanned books into free ebooks","summary":"Distributed Proofreaders makes most new Project Gutenberg books: volunteers check each scanned page in five rounds before one person builds the ebook.","content_html":"<p>A volunteer at <a rel=\"external\" href=\"https://www.pgdp.net/c/\">Distributed Proofreaders</a> gets one scanned page and the text that OCR software read from it on a single web page, corrects the text and sends the page back. The next volunteer gets the same image with the corrected text. A page usually goes through three such proofreading rounds and two formatting rounds, and then one person joins all the pages into the finished ebook. That's where most new <a href=\"/blog/what-is-project-gutenberg/\">Project Gutenberg</a> books come from: Gutenberg's home page names the site as its main source, and so does the site itself.</p>\n<p>Charles Franks started it in 2000 to help Project Gutenberg, the site's <a rel=\"external\" href=\"https://www.pgdp.net/c/\">history note</a> says. It became an official Project Gutenberg site in 2002 and a separate legal entity in May 2006. On 1 November 2026 the counters on the same page showed 51,469 finished titles, 106 of them posted in October, with 638 books still in the proofreading rounds, 2,213 in their final checks before assembly, and 1,247 volunteers active in the past 30 days.</p>\n<h2 id=\"from-a-printed-book-to-an-ebook\">From a printed book to an ebook</h2>\n<p>The site's <a rel=\"external\" href=\"https://www.pgdp.net/d/walkthrough/en\">walkthrough for new volunteers</a> and Project Gutenberg's own pages give this order:</p>\n<ol>\n<li>Copyright clearance: Project Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/help/copyright.html\">copyright page</a> asks for scans of the title page and the back of it (the verso) and promises an answer within a few days. Only a book that is out of copyright in the United States gets through.</li>\n<li>Content: a volunteer supplies the page images and the text that OCR software read from them.</li>\n<li>The rounds, each done by a different volunteer with the scan on the same screen as the text. In proofreading round 1 (P1) the raw OCR text gets corrected; P2 runs the spellcheck and compares again, and P3 looks for the small errors that two rounds missed. Formatting round 1 (F1) adds the markup, some of it set by rules for that one project, and F2 removes the formatting errors that remain.</li>\n<li>Post-processing: one volunteer assembles every page into a single e-text.</li>\n<li>“Smooth Reading”, which is optional: anyone may download the book and read it as a reader would, without the scans, and mark what looks wrong.</li>\n<li>Post-processing verification: another volunteer checks the post-processor's file before it goes to Project Gutenberg.</li>\n</ol>\n<h2 id=\"the-rules-of-a-page\">The rules of a page</h2>\n<p>The <a rel=\"external\" href=\"https://www.pgdp.net/d/walkthrough/en/02_P1Frame.htm\">rules shown to P1 proofreaders</a> keep the book as it was printed. A proofreader keeps each line ending where it falls in the scan, but joins any word that was split across two lines. Old spellings stay as printed. Each paragraph starts after a blank line with no indent, and stray spaces that OCR put around punctuation come out. If a proofreader isn't sure, the page should match the original, with a note in the form <code>[** note]</code> for the next volunteer or the project manager.</p>\n<p>The walkthrough sends a new volunteer to P1 or to “Smooth Reading”, and its <a rel=\"external\" href=\"https://www.pgdp.net/d/walkthrough/en/10_ConclusionFrame.htm\">last page</a> promises that at least two more experienced proofreaders check every P1 page again before it's done. Newcomers are asked to start on a project marked BEGINNERS ONLY and to do only a few pages of it, five or so; a mentor reviews those pages and sends feedback after a few days, though the <a rel=\"external\" href=\"https://www.pgdp.net/d/walkthrough/en/03_ProjFrame.htm\">sample project's comments</a> warn that it can take a week. After 5 to 15 pages on such projects, the <a rel=\"external\" href=\"https://www.pgdp.net/d/walkthrough/en/02_P1Frame.htm\">P1 page</a> asks them to leave those projects to newer people and move to one marked EASY. The <a rel=\"external\" href=\"https://www.pgdp.net/c/\">home page</a> suggests a page a day but asks for no commitment beyond doing your best.</p>\n<h2 id=\"after-the-rounds\">After the rounds</h2>\n<p>Once the post-processing verifier signs off, the book goes to Project Gutenberg, whose <a rel=\"external\" href=\"https://www.gutenberg.org/help/volunteers_faq.html\">volunteer FAQ</a>, in its advice to producers, says that most of them make the HTML version first and derive the plain text from it. Gutenberg's ebookmaker software makes the downloadable ebooks from the HTML files that volunteers submit.</p>\n<p>The 50,000th book to come out of this process was itself a book about printing: William Savage's <em>A Dictionary of the Art of Printing</em>, a printer's manual published in London in 1841, which <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/77410\">Gutenberg released on 6 December 2025</a>. The site's blog <a rel=\"external\" href=\"https://blog.pgdp.net/2025/12/07/celebrating-50000-titles/\">marked the milestone</a> the next day.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/do-dyslexia-fonts-work/","url":"https://readwyn.com/blog/do-dyslexia-fonts-work/","title":"Do dyslexia fonts actually help?","summary":"What the studies of OpenDyslexic and Dyslexie found about reading speed and accuracy, and why letter spacing helped where the letter shapes did not.","content_html":"<p>Not in the tests so far. In a Dutch study published in 2017, 170 children with dyslexia read a text set in Dyslexie and the same text in Arial, and they were neither faster nor more accurate in Dyslexie. A second experiment in the <a rel=\"external\" href=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC5934461/\">same paper by Kuster and colleagues</a> gave word lists to 102 children with dyslexia and 45 without, in Dyslexie, Arial and Times New Roman, with the same result. Most of the children preferred Arial or Times New Roman, and the font a child liked didn't predict how well they read in it. In the authors' words, the font \"neither benefits nor impedes the reading process\".</p>\n<p>OpenDyslexic, the free font whose letters are drawn heavier at the bottom, fared no better. In a study published in 2016, <a rel=\"external\" href=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC5629233/\">Wery and Diliberto</a> tested it with 12 pupils in grades 3 to 6 at a school for children with dyslexia in the south-east of the United States. Against Arial and Times New Roman, in letter naming, word reading and nonsense words, it improved neither speed nor accuracy, for any single pupil or for the group. Some pupils called the font new or different; none said they'd rather read in it. With 12 pupils the study is small, and, as the authors note, it didn't measure comprehension. The font's own <a rel=\"external\" href=\"https://opendyslexic.org/\">website</a> claims little more than that it is designed against some common symptoms of dyslexia, and that if you like reading it, it's for you.</p>\n<h2 id=\"what-did-make-a-difference\">What did make a difference</h2>\n<p>In a <a rel=\"external\" href=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC3396504/\">2012 study in PNAS</a>, Zorzi and colleagues gave 74 Italian and French children with dyslexia, aged 8 to 14, text in 14 pt Times with an extra 2.5 pt between letters, plus wider word and line spacing. The gap between an \"i\" and an \"l\" went from 2.7 pt to 5.2 pt, and the children made about half as many errors. A second experiment, with 20 other Italian children, gave the normal text the same doubled line spacing, so that only the spacing of letters and words differed. Errors still fell, from 11.8 to 6, and reading speed rose from 1.64 to 1.87 syllables a second. Two weeks later, with ordinary line spacing in the normal text again, errors went from 13.5 to 4.95. Younger children without dyslexia who read at the same level got no significant gain, which fits the authors' explanation: readers with dyslexia are unusually affected by crowding, where neighbouring letters make each other harder to recognise.</p>\n<p>Spacing also explains a study in which Dyslexie did come out ahead. <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/27194598/\">Marinus and colleagues</a> found in 2016 that 39 struggling readers of English read 7% more words a minute in Dyslexie than in Arial. Once they gave Arial the same spacing between letters and between words as Dyslexie, the difference was no longer significant. The authors concluded that the gain came from the spacing, not from the letter shapes that Christian Boer had designed in 2008. That spacing doesn't need a special font. Amazon's April 2025 <a rel=\"external\" href=\"https://www.aboutamazon.com/news/books-and-authors/kindle-accessibility-features-for-all-readers\">guide to the Kindle's accessibility features</a> lists the space between words and between characters as settings of their own, right beside the OpenDyslexic font.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/do-you-own-your-kindle-books/","url":"https://readwyn.com/blog/do-you-own-your-kindle-books/","title":"Do you own the Kindle books you buy?","summary":"Amazon's terms say Kindle books are licensed, not sold: what that licence allows, how it differs from a paperback, and what California's 2024 law changed.","content_html":"<p>No. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=201014950\">Kindle Store Terms of Use</a> say that Kindle Content is \"licensed, not sold\" to you by the \"Content Provider\", which the terms define as whoever offers the book in the Kindle Store, Amazon itself or a third party. The money pays for a non-exclusive right to read it on the conditions of those terms. In July 2009 Amazon showed what that can mean, when it <a href=\"/blog/amazon-deleted-1984-from-kindles/\">deleted copies of Nineteen Eighty-Four</a> from customers' Kindles because the seller had no rights to the book.</p>\n<h2 id=\"what-the-licence-allows\">What the licence allows</h2>\n<p>In the version of the terms dated 21 October 2026, the right covers reading through Kindle Software (or as the service otherwise permits), on the number of devices allowed for that book, for personal and non-commercial use. Unless a book states otherwise, you can't sell, rent, lease, distribute, broadcast or sublicense it, and the terms forbid any attempt to get round the DRM. Your rights end automatically if you break any term, and Amazon can then revoke access without a refund. It may also change, suspend or discontinue the service without notice.</p>\n<p>Risk of loss, says a smaller clause, passes to you as soon as you download or access the book. Another caps Amazon's liability to you, outside the Kindle software itself, at what you paid for the book or for the part of the service that your claim concerns.</p>\n<h2 id=\"paper-and-kindle-side-by-side\">Paper and Kindle side by side</h2>\n<table><thead><tr><th></th><th>A printed book bought in the US</th><th>A Kindle book</th></tr></thead><tbody>\n<tr><td>What you get</td><td>Ownership of that copy</td><td>A licence to read it</td></tr>\n<tr><td>Selling or giving it away</td><td>Allowed without the copyright owner's permission</td><td>Not allowed unless the book says so</td></tr>\n<tr><td>Where you can read it</td><td>Anywhere</td><td>Kindle apps and devices, up to the book's device limit</td></tr>\n<tr><td>If you break the seller's terms</td><td>No terms come with the copy</td><td>Your rights end, and access can go with no refund</td></tr>\n<tr><td>If the seller changes its service</td><td>The copy stays on your shelf</td><td>Amazon may change or end the service without notice</td></tr>\n</tbody></table>\n<p>The paper column rests on the first-sale rule in <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/109\">17 U.S.C. § 109</a>: the owner of a lawfully made copy may sell it or otherwise dispose of it without the copyright owner's consent. Subsection (d) of the same section withholds that privilege from someone who got a copy by rental, lease or loan without acquiring ownership of it.</p>\n<p>European law reached a similar place by another route. Two associations of Dutch publishers went to court in The Hague against Tom Kabinet, a company that sold \"second-hand\" ebooks to members of a reading club on its website. The Dutch court asked the Court of Justice of the EU for a ruling. On 19 December 2019, in Case C-263/18, its Grand Chamber held that supplying an ebook by download for permanent use is a communication to the public, not a distribution (<a rel=\"external\" href=\"https://curia.europa.eu/jcms/upload/docs/application/pdf/2019-12/cp190159en.pdf\">press release</a>). So the exhaustion rule that frees a printed book for resale doesn't apply. Part of its reasoning was that a digital copy doesn't wear out, which makes a used one a perfect substitute for a new one.</p>\n<h2 id=\"what-california-changed\">What California changed</h2>\n<p>California's <a rel=\"external\" href=\"https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202320240AB2426\">AB 2426</a>, approved on 24 September 2024 as Chapter 513 of that year's statutes, added section 17500.6 to the Business and Professions Code. It took effect on 1 January 2025, according to the <a rel=\"external\" href=\"https://irwin.asmdc.org/press-releases/20240925-governor-newsom-signs-ab-2426-law-increases-transparency-surrounding\">press release</a> of its author, Assemblymember Jacqui Irwin. Ebooks are covered: the law lists digital books among its digital goods. A seller that uses \"buy\", \"purchase\" or any word a reasonable person would read as unrestricted ownership must do one of two things. It can ask the buyer, at each sale, to confirm that they are getting a licence. That confirmation lists every restriction and, where it applies, warns that access can be revoked if the seller loses the rights. Or it can show, before each sale, a plain statement that buying is a licence, set apart from the other terms, with a link or QR code to the full licence.</p>\n<p>A shop may still sell only a licence, as long as it says so before the sale. The law also exempts three cases: subscriptions that give access only while you pay, goods given away free, and goods whose access the seller can't revoke after the sale. That last group includes a file offered at the time of purchase for permanent offline download to your own storage, usable without an internet connection, which describes the shops that sell <a href=\"/blog/where-to-buy-drm-free-ebooks/\">DRM-free ebooks</a> as downloads.</p>\n<h2 id=\"the-copy-on-your-computer\">The copy on your computer</h2>\n<p>Until February 2025, the Content Library on Amazon's website had a \"Download &amp; transfer via USB\" option that saved a purchased book to a computer, one title at a time. That month Amazon warned that the option would go on 26 February, <a rel=\"external\" href=\"https://www.theverge.com/news/612898/amazon-removing-kindle-book-download-transfer-option-usb\">The Verge reported</a>, and a spokesperson said books already on a Kindle would stay readable (<a href=\"/blog/kindle-download-and-transfer-removed/\">more on that change</a>).</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/does-bionic-reading-work/","url":"https://readwyn.com/blog/does-bionic-reading-work/","title":"Does Bionic Reading make you read faster?","summary":"Four independent tests found that bolding the start of each word, as Bionic Reading does, leaves reading speed unchanged. The name is a registered trademark of a Swiss company.","content_html":"<p>No. The published tests of skilled adult readers have found that Bionic Reading, which prints the opening letters of each word in bold, leaves people reading at the same pace as plain text. The most direct is a 2024 paper by Joshua Snell in Acta Psychologica, titled “No, Bionic Reading does not work”. His participants read 100 paragraphs in one format or the other, and <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/38723450/\">the analysis found no significant difference in reading times</a>.</p>\n<p>Here is the first sentence of chapter one of <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/145\">Middlemarch</a> with the first half of each word in bold, which is how Snell describes the format: <strong>Mi</strong>ss <strong>Bro</strong>oke <strong>h</strong>ad <strong>th</strong>at <strong>ki</strong>nd <strong>o</strong>f <strong>bea</strong>uty <strong>wh</strong>ich <strong>se</strong>ems <strong>t</strong>o <strong>b</strong>e <strong>thr</strong>own <strong>in</strong>to <strong>rel</strong>ief <strong>b</strong>y <strong>po</strong>or <strong>dr</strong>ess. The <a rel=\"external\" href=\"https://bionic-reading.com/br-method/\">maker's method page</a> calls the bold letters artificial fixation points: your eye is meant to hop from one to the next while your brain completes each word. Its apps adjust the effect with three settings named Fixation, Saccade and Opacity. Snell notes in his abstract that theory might predict a cost rather than a benefit.</p>\n<p>The name belongs to a company in Chur, Switzerland. Its <a rel=\"external\" href=\"https://bionic-reading.com/patent-trademark/\">patent and trademark page</a> names the Swiss typographic designer Renato Casutt as the inventor, gives BRCG Casutt GmbH as the holder of the rights, with Bionic Reading AG at the same address as the partner that sells the products, and says only products licensed by Bionic Reading AG may use the name. It lists trademark registrations in nine places, from Switzerland and the European Union to Japan and New Zealand, and a French patent for a system and method of fast reading, filed in 2017 and granted in 2021. The <a rel=\"external\" href=\"https://tsdr.uspto.gov/statusview/rn5557651\">USPTO record for BIONIC READING</a>, registration number 5557651, was registered on 11 September 2018 and is still live as of November 2026; the word “Reading” on its own is disclaimed.</p>\n<p>The biggest test came from Readwise, which put an open experiment on Hacker News, Reddit and Twitter in June 2022 and <a rel=\"external\" href=\"https://blog.readwise.io/bionic-reading-results/\">posted the results on 19 July</a>. By then 2,074 people had finished it. Each read two essays by Paul Graham of about 1,000 words, cut into halves, with the typeface alternating between plain Literata and Literata with Bionic Reading. After duplicates and rushed attempts were removed, 1,916 readers were left. Their mean speed was 325.3 words per minute with the bold starts and 327.9 without, so on average people were 2.6 words per minute slower, a gap the authors found wasn't statistically significant (a two-tailed p of 0.055). Comprehension came out at 88% for both. Volunteers from the internet aren't a lab sample, but with nearly two thousand of them the two averages still sit within 3 words per minute of each other.</p>\n<p>A lab study with an eye tracker followed. Adam Parker, Haibei Wang, Hyunjun Lee and Holly Joseph published it in Royal Society Open Science on 2 September 2026, <a rel=\"external\" href=\"https://doi.org/10.1098/rsos.250177\">as a registered report with 90 skilled adult readers</a>. They compared three versions of each sentence: the start of words bolded, the middle of words bolded, and no bold at all, with a target word of four or eight letters in every sentence. On longer words, the bold moved the point where readers' eyes first landed closer to the start of the word. That shift bought nothing, though: word skipping, the time spent on each word, sentence reading times and comprehension showed no difference.</p>\n<p>A second eye-tracking study found the same. With 53 adult readers at the University of the Free State in South Africa, T. R. Beelders measured <a rel=\"external\" href=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC12565662/\">no significant difference</a> in reading speed, or in the number and length of fixations (Journal of Eye Movement Research, October 2025).</p>\n<p>One published study does report a gain, with a different design. Hossein Khazaee and two colleagues in Iran <a rel=\"external\" href=\"https://journals.razi.ac.ir/article_3914.html\">split 60 intermediate learners of English into two groups</a> (Technology Assisted Language Education, September 2025). A SwiftRead test chose learners of similar comprehension and speed, around 250 words a minute. One group then read 32 texts in the Bionic Reading app over eight teaching sessions, the other read the same texts in plain type, and the same test at the end found the Bionic group faster and better at comprehension. That is a training study of second-language learners, and the authors themselves ask for more research.</p>\n<p>The study that the company cites doesn't show a speed gain either, although its <a rel=\"external\" href=\"https://bionic-reading.com/br-method/\">method page</a> still tells readers they'll “Learn new things faster”. The <a rel=\"external\" href=\"https://bionic-reading.com/br-study/\">study page on its site</a> reports a 2022 pilot at ZHAW Zurich University of Applied Sciences, in which twelve linguistics students read at 191.28 words per minute with Bionic Reading and 189.14 without, no significant difference. The page says outright that faster reading isn't what the measurement shows. Its case rests on retention instead: 11.75 points against 9.83 on questions about the texts, eight for each text, answered aloud. Twelve people took part, and the larger tests by Readwise and by Parker's group found no comprehension gain.</p>\n<p>One question stays open. The same page says readers with ADHD or dyslexia report that the format helps, and it labels that as feedback from practice rather than a result of the study. None of the tests above was built around such readers: the ZHAW students had no reading difficulties, Parker's group tested skilled adult readers, and Beelders tested fluent adult readers of English.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/does-highlighting-help-you-remember/","url":"https://readwyn.com/blog/does-highlighting-help-you-remember/","title":"Does highlighting help you remember what you read?","summary":"What the research on highlighting finds: a small gain on memory tests for college students, none for understanding, and a low rating in a large 2013 review.","content_html":"<p>A little for memory, and hardly at all for understanding. A large 2013 review of study techniques rated highlighting low utility. In a 1992 study by Sarah Peterson, students who underlined as they read did worse on the comprehension questions that needed an inference linking ideas from different parts of the text. John Dunlosky, who <a rel=\"external\" href=\"https://www.aft.org/ae/fall2013/dunlosky\">described the study in 2013</a>, suggests that the underliners may have spent their attention on individual ideas and less of it on how those ideas connected.</p>\n<p><a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/26173288/\">Dunlosky and four colleagues' 2013 paper</a> in <em>Psychological Science in the Public Interest</em> rated ten techniques by how well the evidence for each held up across different students, materials and kinds of test. Practice testing and spreading study out over time came top. Highlighting landed in the bottom group with rereading, the keyword mnemonic, mental imagery and summarising. Their complaint about highlighting and rereading was that most students use them and neither reliably improves results. In the version he wrote for teachers, Dunlosky lists Air Force trainees, children and undergraduates among the groups that did no better with a highlighter than with plain reading. He also admits that he owns a favourite highlighter himself, and says he wouldn't take students' away from them.</p>\n<p>Writing a summary in your own words sits in the same low group. The review found that it helps some students on some tasks, undergraduates among them, and Dunlosky notes that younger students who can't yet write a good summary need a lot of teaching before it does them any good.</p>\n<p>A larger pool of studies gives highlighting slightly more credit. Héctor Ponce, Richard Mayer and Ester Méndez pooled <a rel=\"external\" href=\"https://link.springer.com/article/10.1007/s10648-021-09654-1\">36 published articles from 1938 to 2019</a>, 85 effect sizes in all. Highlighting your own text improved scores on memory tests (an average effect size of 0.36) but not on comprehension tests (0.20). It helped college students and not school students. The odd result is that readers given text with the important material already highlighted did better than readers who marked their own: an average effect of 0.44 on memory and on comprehension alike, and a gain for both age groups (0.41 for college students, 0.48 for school pupils).</p>\n<p>The gap between the two suggests that the choosing is where readers go wrong. A 2025 analysis of the <a rel=\"external\" href=\"https://link.springer.com/article/10.1186/s40536-025-00241-2\">digital PIRLS 2021 reading test</a> looked at how fourth-grade pupils in five countries used the highlighter built into the test. Marking the text that a question needed went with a correct answer in only one of the five countries. The same paper's review of earlier work notes that school pupils pick less relevant text than adults, and cites a 2018 study in which highlighting helped adults reading on paper but not adults reading on a screen.</p>\n<p>Timing changes the picture too. Carole Yue, Benjamin Storm, Nate Kornell and Elizabeth Bjork had students <a rel=\"external\" href=\"https://link.springer.com/article/10.1007/s10648-014-9277-z\">read a passage twice, with or without a highlighter</a>, and tested them a week later. Those who highlighted did better, mostly when their two readings came back to back rather than spread apart. Highlighting didn't cost them anything on the sentences they left unmarked. And the students who doubted that highlighting worked gained more from it than the ones who believed in it.</p>\n<p>Dunlosky's advice on the same page is that highlighting is only the beginning: once students have read and marked a text, they should go back over it with one of the techniques higher up his list, and practice testing sits at the top. Peterson's underliners lost track of the links between ideas, and a highlight read on its own, months later, hides those links as well. Readwyn's Notebook has a \"Show context\" button for that: it swaps the highlight for the paragraph before it, its own paragraph and the paragraph after. Read with its surroundings, a highlight works as a prompt. Cover the line, recall what it said and why it mattered, then check; the post on <a href=\"/blog/why-you-forget-what-you-read/\">why you forget what you read</a> covers that kind of retrieval.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/does-speed-reading-work/","url":"https://readwyn.com/blog/does-speed-reading-work/","title":"Does speed reading work?","summary":"Reading scientists find a trade between speed and comprehension. Famous fast readers were probably skimming, one-word-at-a-time apps cost understanding, and skimming has one real use.","content_html":"<p>You can read faster, but you'll understand less. Going from around 250 words a minute to 500 or 750 while keeping the same grasp of the text is unlikely, according to a 2016 review of the evidence by five reading scientists led by Keith Rayner in <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/26769745/\">Psychological Science in the Public Interest</a>. Speed and accuracy trade against each other. The review's advice for getting through text quicker without losing it is plain: read a lot and build your vocabulary, because language skill is what sets reading speed. (Your own pace is a separate question, covered in <a href=\"/blog/reading-speed-words-per-minute/\">reading speed in words per minute</a>.)</p>\n<h2 id=\"famous-fast-readers\">Famous fast readers</h2>\n<p>Evelyn Wood, whose Reading Dynamics course <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/26769745/\">started in 1959</a>, claimed a professor who read more than 2,500 words a minute with excellent recall, and John F. Kennedy was said to manage 1,200. When <a rel=\"external\" href=\"https://slate.com/news-and-politics/2000/02/the-1000-word-dash.html\">Slate checked these figures in February 2000</a>, Ronald Carver of the University of Missouri-Kansas City guessed that Kennedy read at 500 to 600 and could perhaps skim at 1,000. Carver called all three claims bunk. The Slate writer, Timothy Noah, added that Wood's professor, if he existed, was probably skimming too.</p>\n<p>Carver had tested this himself, <a rel=\"external\" href=\"https://slate.com/news-and-politics/2000/02/the-1000-word-dash.html\">Slate reported</a>, in a 1985 study in Reading Research Quarterly. None of his 16 subjects could read faster than 600 words a minute while keeping more than 75% of the content. He also threw out an earlier report of a reader at 17,040 words a minute, since the times, the word counts and the comprehension tests had all been self-reported and self-graded.</p>\n<h2 id=\"apps-that-keep-your-eyes-still\">Apps that keep your eyes still</h2>\n<p>Some speed-reading apps flash one word at a time in the same spot, so your eyes never have to hop along the line. The technique is called rapid serial visual presentation, or RSVP. In November 2026 the <a rel=\"external\" href=\"https://spritz.com/\">Spritz website</a> promises “Read 1000+ Words Per Minute” and says retention is better than with ordinary reading, without citing a study.</p>\n<p>Eyes don't only move forward, though. Readers jump back to words they've already passed, and those regressions help them understand. Schotter, Tran and Rayner showed this in 2014 by <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/24747167/\">masking each word as soon as the reader's eyes had moved past it</a>, so nothing could be reread, much as in RSVP. Comprehension dropped, and not only on sentences built to be ambiguous. A 2017 study by Acklin and Papesh in the American Journal of Psychology then <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/29461715/\">tested RSVP itself</a> at 700 and 1,000 words a minute against ordinary static text, using passages written at 6th-grade and 12th-grade level. Static text gave the best comprehension. Between the two RSVP speeds the slower one helped verbatim recall, and the faster one did better on questions that needed an inference. <a href=\"/blog/does-bionic-reading-work/\">Bionic Reading</a>, which bolds the start of each word, didn't speed up skilled adult readers in tests either, although one <a rel=\"external\" href=\"https://journals.razi.ac.ir/article_3914.html\">training study of 60 Iranian learners of English</a>, published in 2025, reported faster reading after eight sessions with the app.</p>\n<h2 id=\"the-voice-in-your-head\">The voice in your head</h2>\n<p>Speed-reading teachers go after the inner voice. Iris Reading, which sells courses, warns that if you pronounce every word \"you will only read as fast as you talk\", and its page on <a rel=\"external\" href=\"https://irisreading.com/how-to-read-faster-by-reducing-subvocalization/\">reducing subvocalization</a> says that programs which tell you to eliminate the voice completely are wrong, and teaches you to say only some of the words in your head instead. Most readers have one to lose. In a <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/28582587/\">2017 survey of 570 people</a> by Vilhauer at New York University, 80.7% heard an inner voice at least sometimes when reading silently, and a third of those heard it every time. Most gave it particular qualities, such as a gender, an accent or a pitch, and some heard other people's voices rather than their own. The remaining 19.3% said they understood the words without hearing anything.</p>\n<p>Researchers can block inner speech with a technique called articulatory suppression. In a 1990 Oxford study of braille readers it <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/2364246/\">significantly cut comprehension of difficult texts</a>, though speed held up. With easy texts the best readers weren't affected at all and beginners lost both speed and comprehension. Readers in between lost some comprehension and, if anything, read a little faster.</p>\n<h2 id=\"where-skimming-pays-off\">Where skimming pays off</h2>\n<p>Rayner's review doesn't call fast reading useless. If you don't need a thorough understanding, it says, skimming gets you through a text faster with moderate comprehension.</p>\n<p>The loss of comprehension can be small when the questions are easy. Fitzsimmons and colleagues had 32 young adults read 40 Wikipedia articles, half carefully and half by skimming, for a <a rel=\"external\" href=\"https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0239134\">2020 PLOS ONE study</a>. The skimmers skipped more words and fixated for less time, and still answered 86% of yes/no questions correctly, against 91% when they read with care.</p>\n<p>What skimming buys depends on what you're asked afterwards. In 2009 Duggan and Payne at the University of Bath gave readers <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/19751073/\">only enough time to read half of each text</a>. Skimming the whole text beat reading half of it for memory of the important ideas, but it did nothing for the less important details or for inferences, and in a second experiment it was no better than reading the first or second half of every paragraph. The eye tracker in that experiment showed where skimmers go: early in each paragraph, near the top of each page, and to the earlier pages of the document.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/does-your-e-reader-track-what-you-read/","url":"https://readwyn.com/blog/does-your-e-reader-track-what-you-read/","title":"Does your e-reader track what you read?","summary":"What Kindle, Kobo and Libby record about your reading, from the furthest page read to session times, and which settings turn some of it off.","content_html":"<p>Yes. Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">user guide for the Kindle Colorsoft</a> names four things that its Whispersync for Books setting backs up to Amazon's servers: your annotations, the most recent page you read, the furthest page you read, and your collections. The switch is under All Settings, then Device Options, then Whispersync, and the guide describes it as turning that automatic backup on or off. Your highlights land in a file on the device called My Clippings either way.</p>\n<p>Pages are also counted for money. Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G201541130\">KDP help page</a> says authors in Kindle+ (the subscription <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/what-is-kindle-plus\">renamed from Kindle Unlimited</a> in October 2026) are paid from a shared fund the first time a subscriber reads a given page of their book, measured in Kindle Edition Normalized Pages. A return to a page already read counts as rereading and earns nothing extra. Paying that way needs a record of which pages of which book each subscriber has reached; <a href=\"/blog/how-kindle-unlimited-pays-authors/\">how Kindle+ pays authors</a> goes through the sums.</p>\n<p>When several Kindle users have highlighted the same passage, your Kindle underlines it, so other people's reading reaches your screen too. You can hide the underlines under Aa, More, Popular Highlights. The guide doesn't say whether hiding them keeps your own highlights out of the count. Time to Read, the estimate of how long a chapter or book will take, works from your past and current reading speed, and that speed, the guide says, is kept only on the Kindle and never reaches Amazon's servers. For your own files, notes, highlights, bookmarks and the last page read sync only for personal documents archived in Kindle format, which the guide ties to Send to Kindle and its Personal Document Archiving setting.</p>\n<p>Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/13058637908631-Adjust-your-privacy-settings-on-Kobo-com\">help page on privacy</a> opens by saying that Kobo tracks \"your data, activity, and preferences\". If you live in Canada, the United States, the United Kingdom, Europe or Brazil, the controls are on the website at kobo.com/privacy. Everywhere else they're on the device: More, Settings, Energy saving and privacy, and at the very bottom a checkbox labelled \"Automatically share data about the features you use\". It covers that one eReader only, so a second device needs its own, and a reset or a sign-out means setting it again. A Wi-Fi sync also <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017513334-Sync-your-Kobo-eReader-over-Wi-Fi\">updates your bookmarks and annotations</a> with your account.</p>\n<p>Kobo's link to StoryGraph shows how fine the record is. The <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/41397329904919-About-Storygraph\">help page on that integration</a> lists what Kobo passes on: the books you're reading, your percentage, session timestamps and finish dates. If you mark a book finished at under 90 percent, StoryGraph files it as Did Not Finish. Sideloaded books aren't included; the sync covers books bought from the Kobo store or read through Kobo Plus.</p>\n<p>OverDrive, which makes the Libby app for library ebooks, lists lending history, holds, reading progress, bookmarks, highlights and notes among what it may collect, next to your device ID and library card number, in its <a rel=\"external\" href=\"https://www.overdrive.com/policies/privacy-policy.htm/\">privacy policy</a> (last updated November 2025). The same policy says OverDrive doesn't sell your information, that you can hide your lending history in the app, and that it keeps data for as long as it considers necessary to run the service.</p>\n<p>One case of an app overreaching is on record in detail: Adobe Digital Editions 4, the desktop app that libraries encouraged their patrons to use. On 7 October 2014 the EFF <a rel=\"external\" href=\"https://www.eff.org/deeplinks/2014/10/adobe-spyware-reveals-again-price-drm-your-privacy-and-security\">wrote about reports</a> that it was sending reading logs back to Adobe in the clear. Adobe's answer was that it collected data only on the book currently open, including how long you'd been reading and how much you'd read, and it didn't deny the lack of encryption. When the EFF's Cooper Quintin <a rel=\"external\" href=\"https://www.eff.org/deeplinks/2014/10/what-we-can-learn-adobe-e-reader-mess\">watched the traffic</a>, version 4.0.0 sent the titles, the pages read and the time spent reading at every start, and it also reported books added to an attached Sony Reader PRS-600 that had never been opened in the app. Version 4.0.1, released later that month, moved to HTTPS. In the EFF's tests it no longer sent anything about what was read; data went to Adobe only the first time a book with DRM was opened.</p>\n<p>Readwyn has \"Read privately\" (⌥⌘R on a Mac) in the \"…\" menu of a book's screen. It opens the book with nothing saved: no progress, no dates, no time read, no change of status and no sync to a reading service, and the top bar shows \"Private reading · nothing is saved\" the whole time.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/download-public-domain-books-from-hathitrust/","url":"https://readwyn.com/blog/download-public-domain-books-from-hathitrust/","title":"How to get public-domain scans from HathiTrust and Google Books","summary":"Why HathiTrust lets you save only one page of many public-domain books, who can download the whole PDF, and how to get the same scan from Google Books.","content_html":"<p>Open a public-domain book on HathiTrust without logging in, and the download options in the sidebar will often let you save the page in front of you and nothing more. That happens when Google scanned the book. HathiTrust's page on <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2387247137\">what you can download</a> says anyone may download a whole public-domain book that has no download restriction (one digitised by the Internet Archive, say, or opened under a Creative Commons licence), but a Google scan only one page at a time. For the whole book, follow the link to Google Books, which offers its own copy as a free PDF or EPUB when it shows the book in full view.</p>\n<p>The page limit isn't HathiTrust's own choice. The libraries' agreements with Google require it to prevent bulk downloading of the books Google digitised, so it gives full PDFs of them only to people who log in through a member library, where it can work with the library on how the files are used. A guest account, made with a Google, LinkedIn or Microsoft login or a University of Michigan “friend” account, adds only saved collections, according to the <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2386919473\">Logging In page</a>. The same page warns that someone at a member library that uses Google Apps gets only guest access if they sign in with Google instead of picking the institution from the menu. To tell which kind of book you have, <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2386591873\">look for Google's watermark</a> on the page scans.</p>\n<p>Members who log in can take the whole book as a PDF, with the page images and the OCR text behind them, or as an <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2387247137\">EPUB</a>, plain text, or page images. According to the <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2387247130\">formats page</a>, the text comes as one file or as a zip with a file for each page, and the JPEG and TIFF images are lower in resolution than the library's own scans, with a watermark at the bottom; TIFF comes ten pages at a time. How sharp they get depends on what the library sent: black and white TIFFs are usually 600 ppi, JPEG2000 files 300 or 400. All of the text is automatic OCR, which the same page warns can contain misspellings and gibberish; correcting that page by page is the work that <a href=\"/blog/distributed-proofreaders/\">Distributed Proofreaders</a> volunteers do. Anyone can try the whole-book download on the help page's own example of an unrestricted book, <a rel=\"external\" href=\"https://hdl.handle.net/2027/gri.ark:/13960/t8pc40n7d\">A Handbook of Greek Lace Making</a>.</p>\n<p>HathiTrust decides automatically which books are public domain, at first from the date and place of publication in the catalogue record. As of 2026 the <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2387345450\">rules</a> are these: a book published in the US in 1930 or earlier, one published elsewhere in 1900 or earlier, and a US federal government document are all full view. A book published outside the US from 1901 to 1930 is full view only from a US IP address, because <a href=\"/blog/public-domain-differs-by-country/\">public domain differs by country</a>; anywhere else it shows as “Limited (search only)”. Anything from 1931 on is limited for everyone, members included, unless a rights holder has opened it with a Creative Commons licence or HathiTrust's copyright reviewers have cleared it. The reviewers choose their own books, and the team says it can no longer take requests for single titles.</p>\n<p>Limited books are <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2386919535\">hidden from a search</a> until you set “Item Viewability” to “All Items”. They're still useful: a <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2386591860\">search inside one</a> lists each page with a hit, by scan number and printed page number, with the count of hits.</p>\n<p>For the Google Books route, open the book in HathiTrust's viewer, expand “Get This Item” in the sidebar, and click the Google Books link. HathiTrust's help calls it <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2387345450\">“Download at Google Books”</a> on a limited book and <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2387247137\">“Find in Google Books”</a> on one in full view; if there's no such link, Google doesn't have that copy. On a book in full view, Google's <a rel=\"external\" href=\"https://support.google.com/books/answer/43729?hl=en-GB\">help for classic Google Books</a> says to hover over “eBook - Read at no cost” and click PDF or EPUB next to “Download”. To search only these books, set the Search menu of <a rel=\"external\" href=\"https://books.google.com/advanced_book_search\">Advanced Book Search</a> to “Full view only”. Google makes its own copyright decisions, so a book that HathiTrust opens may not download there, and the same help page lets you ask Google to review a book you think is public domain in your country.</p>\n<p>What you may do with the file differs by scanner. Google's scans <a rel=\"external\" href=\"https://hathitrust.atlassian.net/servicedesk/customer/kb/view/2386591873\">may not be hosted again</a>, redistributed or used commercially, but text you transcribe from them, paraphrase or translate carries no such restriction. Books that other organisations scanned have no Google terms. One caution on the same page covers both kinds: some of HathiTrust's books are public domain in the US but not worldwide.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/e-ink-carta-pearl-kaleido/","url":"https://readwyn.com/blog/e-ink-carta-pearl-kaleido/","title":"What the E Ink screen names Carta, Pearl and Kaleido mean","summary":"E Ink Carta vs Pearl: when each screen generation came out, what E Ink claimed for it, what Carta 1200 and 1300 add, and where colour Kaleido fits.","content_html":"<p>E Ink launched Pearl in July 2010, in the graphite Kindle DX, and Carta in September 2013, in the Kindle Paperwhite, so Pearl is the older of the two. Both are black-and-white inks from the same company, and both were sold on contrast. E Ink said Pearl had a contrast ratio about 50 per cent higher than its earlier screens (<a rel=\"external\" href=\"https://www.engadget.com/2010/07/01/e-ink-explains-the-new-pearl-display-used-in-the-updated-kindle/\">Engadget, 1 July 2010</a>). Three years later it claimed another 50 per cent for Carta over previous generations, along with over 20 per cent more reflectance, which you see as a whiter page (<a rel=\"external\" href=\"https://www.eink.com/news/detail/E-Ink-Introduces-Carta-the-Next-Generation-of-Electronic-Ink-Display-Technology\">E Ink press release, 4 September 2013</a>). Kaleido, the screen in Kobo's colour readers, is a colour filter laid over Carta ink.</p>\n<h2 id=\"before-pearl\">Before Pearl</h2>\n<p>Before Pearl, E Ink's reading screens used Vizplex. The company announced it in May 2007, saying it refreshed about twice as fast as the company's previous film and gave a 20 per cent brighter image (<a rel=\"external\" href=\"https://www.engadget.com/2007/05/10/e-ink-corp-announces-vizplex-tech-to-speed-brighten-displays/\">Engadget, 10 May 2007</a>). The <a rel=\"external\" href=\"https://www.engadget.com/2010/07/01/e-ink-explains-the-new-pearl-display-used-in-the-updated-kindle/\">Pearl announcement of 2010</a> described the new ink as building on the Vizplex designs of the day, and gave it 16 levels of grey. That's still the figure E Ink <a rel=\"external\" href=\"https://www.eink.com/brand/detail/Carta1000\">quotes for Carta</a> today.</p>\n<p>Neither Vizplex nor Pearl appears among the brands on <a rel=\"external\" href=\"https://www.eink.com/brand\">E Ink's website</a> any more. Carta is the black-and-white brand it sells now.</p>\n<h2 id=\"the-numbers-after-carta\">The numbers after Carta</h2>\n<p>Carta became a numbered family. E Ink's page for <a rel=\"external\" href=\"https://www.eink.com/brand/detail/Carta1000\">Carta 1000</a> repeats the 50 per cent contrast gain and says it brings the screen close to the contrast of a paperback. Carta 1100 and 1200 came next, and their <a rel=\"external\" href=\"https://www.eink.com/brand/detail/carta1200\">product page</a> credits them with a 20 per cent faster response than Carta 1000 and a 15 per cent better contrast ratio; the faster response is aimed at smoother handwriting and animation.</p>\n<p>Carta 1300 has no page of its own on E Ink's site as of November 2026, though a 7-inch Carta 1300 module is listed under Carta 1000. Kobo names it in its specs. The spec sheet of its <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-clara-bw\">Clara BW</a> gives the display as a \"6” glare-free E Ink Carta 1300 HD touchscreen\" at 1448 by 1072 pixels, and the page promises faster page turns and deeper contrast. A maker's spec sheet is the place to check which panel a reader has; if you don't know the exact model, start with <a href=\"/blog/which-kobo-do-i-have/\">which Kobo do I have</a> or <a href=\"/blog/which-kindle-do-i-have/\">which Kindle do I have</a>.</p>\n<h2 id=\"how-much-the-differences-show\">How much the differences show</h2>\n<p>Less than the percentages suggest. When Pearl was new, The eBook Reader blog set a Sony PRS-350 and a Kindle 3, both with Pearl, beside a new Vizplex reader, the Pixelar MReader, and <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2010/09/21/pearl-vs-vizplex-are-the-new-high-contrast-pearl-screens-overrated/\">found the gap small</a>. With the Sony set to the same bold font as the MReader, the Pearl text was darker, but not by much, and the author thought Amazon's own waveform and font technology accounted for as much of the Kindle's contrast as the panel. In 2024 the same blog <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2024/05/03/e-ink-carta-1300-is-not-a-worthwhile-upgrade/\">compared the Clara BW with the Clara 2E</a>, which has Carta 1200. Page turns came out roughly equal, the new screen was only milliseconds quicker, and the older one had the slightly whiter background.</p>\n<h2 id=\"kaleido-and-gallery\">Kaleido and Gallery</h2>\n<p>Kaleido isn't a new ink. E Ink's <a rel=\"external\" href=\"https://www.eink.com/brand/detail/Kaleido\">Kaleido page</a> lists the module's layers as a backplane, an ink layer, a colour filter layer and a protective sheet. The ink under the filter is second-generation Carta. The filter stays over the whole page, even a page with only text on it, and Kobo <a rel=\"external\" href=\"https://goodereader.com/blog/electronic-readers/everything-you-need-to-know-about-the-new-kobo-e-readers\">told Good e-Reader</a> that Kaleido 3 makes the grey background darker. So a page of <a href=\"/blog/colour-e-ink-for-reading/\">plain text on a colour reader</a> comes out a little darker than on a black-and-white one. Kobo's first colour readers, the Libra Colour and Clara Colour, used Kaleido 3 when they were <a rel=\"external\" href=\"https://www.trustedreviews.com/news/kobo-announces-libra-colour-and-clara-colour-e-readers-4515476\">announced in April 2024</a>. Gallery is the other colour brand: the colour is in the ink itself, with no filter over it.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/e-reader-pixels-per-inch/","url":"https://readwyn.com/blog/e-reader-pixels-per-inch/","title":"How many pixels per inch does an e-reader need?","summary":"Is 300 ppi enough on an e-reader? What 167, 212 and 300 ppi change for small type and for Chinese and Japanese text, with the specs of current models.","content_html":"<p>In September 2022 Amazon took its cheapest Kindle from 167 to 300 pixels per inch, <a rel=\"external\" href=\"https://www.engadget.com/amazon-kindle-2022-upgrade-300-ppi-display-usb-c-16-gb-storage-130030447.html\">the same density as the Paperwhite</a>. That jump matters because at 300 ppi, held a foot from your eyes, a screen sits at about the limit of 20/20 vision, so you won't pick out single pixels in the letters. Below that density, each pixel is bigger than normal sight can resolve at that distance: about 1.35 times too big at 212 ppi, and 1.7 times at 167 ppi.</p>\n<p>The arithmetic behind \"enough\" comes from an argument about a phone. When Apple launched the iPhone 4 in June 2010, it said the screen's <a rel=\"external\" href=\"https://www.apple.com/newsroom/2010/06/07Apple-Presents-iPhone-4/\">326 pixels per inch</a> were too dense for the eye to tell apart at a normal distance. Raymond Soneira of DisplayMate disagreed and put the limit <a rel=\"external\" href=\"https://www.engadget.com/2010/06/10/iphone-4s-retina-display-claim-put-under-the-math-microscope\">near 477</a>. The astronomer Phil Plait then <a rel=\"external\" href=\"https://www.discovermagazine.com/the-sciences/resolving-the-iphone-resolution\">worked it through</a>: 20/20 vision resolves about 1 arcminute, which at 12 inches comes to roughly 286 ppi, and perfect eyes resolve about 0.6 arcminutes, or roughly 477 ppi. Apple and Soneira were both right, for different eyes. A 300 ppi e-reader at that distance is a little past the 20/20 threshold; bring it closer, or have better than average sight, and the pixels come back within reach.</p>\n<p>In practice, density decides how many pixels draw each character. Type at 10 point has an em of 3.5 mm. That's about 23 pixels tall at 167 ppi, 29 at 212 ppi (the density of the <a rel=\"external\" href=\"https://press.aboutamazon.com/2012/9/introducing-the-new-kindle-paperwhite-the-most-advanced-e-reader-ever-constructed\">first Kindle Paperwhite in 2012</a>), and 42 at 300 ppi, which is also the density Amazon's print service asks of the images on a paperback cover (<a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G201953020\">a minimum of 300 DPI</a>). In Latin text, fewer pixels mainly blunt the curves and serifs. Chinese and Japanese suffer more, because each character fills the whole em square with strokes: 鬱, which Unihan glosses as dense, thick or moody, has <a rel=\"external\" href=\"https://www.unicode.org/cgi-bin/GetUnihanData.pl?codepoint=9B31\">29 strokes in Unicode's Unihan database</a>. Squeeze 29 strokes into a square 23 pixels across and the parallel strokes in its busiest parts end up barely a pixel apart, so some of them run together. At 42 pixels there's nearly twice the room. For more on how e-readers set those scripts, see <a href=\"/blog/vertical-text-in-ebooks/\">vertical text in ebooks</a>.</p>\n<p>On 1 November 2026 Kobo's US site listed the 6-inch <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-clara-bw\">Clara BW at 1448 by 1072 pixels</a>, which works out to 300 ppi, and the 10.3-inch <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-elipsa-2e\">Elipsa 2E at only 227 ppi</a>. Colour screens carry two figures: the <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-libra-colour\">Libra Colour</a>, with an E Ink Kaleido 3 panel, is 300 ppi for black-and-white content and 150 ppi for colour, so a colour picture on it is drawn at half the density of the text around it (the colour filter that halves it is described in <a href=\"/blog/e-ink-carta-pearl-kaleido/\">Carta, Pearl and Kaleido</a>).</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/e-reader-screen-size/","url":"https://readwyn.com/blog/e-reader-screen-size/","title":"Six, seven, or ten inches: which e-reader size suits you?","summary":"What size e-reader to get: how much more page a 7-inch or 10-inch screen shows than a 6-inch one, what it weighs, and how small an A4 PDF looks on each.","content_html":"<p>For novels, a 6-inch or 7-inch e-reader is enough, and the 7-inch shows about a third more text per page for roughly 25 g more. Get a 10-inch screen if you read PDFs, papers, or A4 documents: on anything smaller, a whole A4 page shrinks to less than half its printed size. The cost is weight, about twice that of a 6-inch model.</p>\n<p>Screen sizes are diagonals, and that hides how fast the area grows. Three of Kobo's current models make the point, with the figures of their spec sheets as they stood on 1 November 2026. The 6-inch <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-clara-bw\">Clara BW</a> has 1448 by 1072 pixels, a screen of about 17 square inches, and weighs 174 g. The 7-inch <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-libra-colour\">Libra Colour</a> has 1264 by 1680 pixels: about 24 square inches, 37 per cent more, at 199.5 g. The 10.3-inch <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-elipsa-2e\">Elipsa 2E</a> has 1404 by 1872 pixels, about 51 square inches. That's three times the page of the Clara BW, and at 390 g it weighs 2.2 times as much. It also spreads its pixels thinner, 227 to the inch against 300 on the two smaller Kobos, which matters for <a href=\"/blog/e-reader-pixels-per-inch/\">small type</a>.</p>\n<h2 id=\"six-inches-from-the-start\">Six inches from the start</h2>\n<p>The 6-inch size goes back to Sony's Librié of 2004, sold only in Japan. The Sony Reader PRS-500 of 2006, which Engadget remembers as <a rel=\"external\" href=\"https://www.engadget.com/2014-04-06-gadget-rewind-2006-sony-reader-prs-500.html\">a 6-inch device</a>, used the same E Ink display. A 6-inch screen at a 3:4 ratio is about 91 by 122 mm, roomy enough for a page of a novel in a comfortable type size.</p>\n<p>A PDF page was another matter. In February 2007 Scott Hanselman wrote that most A4 and Letter PDFs had text that was <a rel=\"external\" href=\"https://www.hanselman.com/blog/hands-on-sony-eink-reader-prs500-reviewed\">very small on the 6-inch screen</a> of his PRS-500, and that he could either resize the text with a PDF editor or turn the device sideways and read each page in two halves. An A4 page is <a rel=\"external\" href=\"https://www.cl.cam.ac.uk/~mgk25/iso-paper.html\">210 by 297 mm</a>, so fitting all of it on a 6-inch screen shrinks it to about 41 per cent, and 10-point body text comes out near 4 points. On a 7-inch screen the page shrinks to about 48 per cent, and the same text to under 5 points.</p>\n<p>Amazon's first Kindle, launched on 19 November 2007, weighed <a rel=\"external\" href=\"https://press.aboutamazon.com/2007/11/introducing-amazon-kindle\">10.3 ounces</a>, about 292 g. By 2009 Amazon's press releases spoke of <a rel=\"external\" href=\"https://press.aboutamazon.com/2009/5/introducing-kindle-dx-amazons-large-screen-addition-to-the-kindle-family-of-wireless-reading-devices\">“Kindle’s 6-inch display”</a> as the standard against which a bigger screen was measured.</p>\n<h2 id=\"amazon-s-first-large-screen-for-pdfs\">Amazon's first large screen, for PDFs</h2>\n<p>That bigger screen was the Kindle DX, announced on 6 May 2009. It had a <a rel=\"external\" href=\"https://press.aboutamazon.com/2009/5/introducing-kindle-dx-amazons-large-screen-addition-to-the-kindle-family-of-wireless-reading-devices\">9.7-inch display</a> with 2.5 times the surface area of the 6-inch one, a built-in PDF reader, and a screen that rotated with the device, so you could hold it upright or sideways. Amazon named documents, newspapers, magazines, and textbooks as what the extra area was for, and Jeff Bezos added cookbooks and computer books.</p>\n<p>Kobo's large size, 10.3 inches, has a handy property. The Elipsa 2E's screen works out to about 157 by 210 mm: as tall as A5 and a little wider. Showing a whole A4 page on it is the same reduction as an A4 to A5 photocopy, <a rel=\"external\" href=\"https://www.cl.cam.ac.uk/~mgk25/iso-paper.html\">71 per cent</a>, so 10-point text becomes about 7 points. That's small but readable, and if you crop the white margins of the PDF, the text gets closer to its printed size. A US Letter page fits at about 73 per cent. Books with a fixed page, such as textbooks and comics, gain the same way; <a href=\"/blog/fixed-layout-vs-reflowable/\">fixed layout and reflowable ebooks</a> explains why an EPUB novel doesn't need the extra room and a PDF does.</p>\n<h2 id=\"amazon-at-seven-inches-since-2017\">Amazon at seven inches, since 2017</h2>\n<p>On 11 October 2017 Amazon moved the Kindle Oasis to a 7-inch, 300 ppi screen and said it fitted <a rel=\"external\" href=\"https://press.aboutamazon.com/2017/10/meet-the-all-new-kindle-oasis-featuring-a-7-inch-300-ppi-display-waterproof-design-and-audible\">over 30 per cent more words per page</a>, so you turn the page less often. The same release gave its weight as 194 g, 10 g less than the 6-inch Paperwhite of the time, so a bigger diagonal doesn't always mean a heavier device. Read the weight on the spec sheet of each model.</p>\n<p>Amazon's own large screen came back with the Kindle Scribe in September 2022: a <a rel=\"external\" href=\"https://press.aboutamazon.com/2022/9/introducing-amazon-kindle-scribe-the-first-kindle-for-reading-and-writing\">10.2-inch, 300 ppi</a> Paperwhite display (Amazon's name for its front-lit E Ink screen), which the release called the world's first of that size and pitched, among other things, for writing on PDF documents. The Paperwhite itself grew to <a rel=\"external\" href=\"https://press.aboutamazon.com/2024/10/amazon-launches-entirely-new-kindle-lineup-including-reimagined-kindle-scribe-and-first-ever-color-kindle\">7 inches</a> in October 2024, a first for that line, and the Scribe of September 2025 went to <a rel=\"external\" href=\"https://9to5toys.com/2025/09/30/amazon-reveals-new-kindle-scribe-colorsoft/\">11 inches at 400 g</a>: about the weight of the Elipsa 2E, for a page a little larger again.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ebook-cover-size/","url":"https://readwyn.com/blog/ebook-cover-size/","title":"What size should an ebook cover be?","summary":"The cover sizes that Amazon KDP, Apple Books, Kobo Writing Life and Google Play Books ask for, and one master size that suits all four with a crop for Kobo.","content_html":"<p>Make it 1,600 pixels wide and 2,560 pixels tall. That's the size Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G6GTK3T3NUHKLEFX\">cover image guidelines for Kindle Direct Publishing</a> recommend, a height 1.6 times the width, and the same file clears the minimums of Apple Books and Google Play Books as well. Kobo is the one store that asks for a different shape, so a cover made for all four needs a little thought about cropping.</p>\n<h2 id=\"what-each-store-asks-for\">What each store asks for</h2>\n<p>The four stores state their rules in different terms: Amazon counts the longest side, Apple the shortest, and Kobo gives a ratio and leaves the pixel count to a partner store. This is how their help pages read on 1 November 2026 (Kobo's was last changed in July 2026):</p>\n<table><thead><tr><th>Store</th><th>Smallest accepted</th><th>Recommended</th><th>Shape</th><th>Files</th></tr></thead><tbody>\n<tr><td>Amazon KDP</td><td>500 px on the longest side</td><td>1,600 × 2,560 px</td><td>Not fixed; aspect ratio up to 500:1</td><td>JPEG (preferred), PNG, GIF, TIFF; up to 50 MB and 10,000 px a side</td></tr>\n<tr><td>Apple Books</td><td>1,400 px on the shorter side</td><td>No size given</td><td>Not stated</td><td>JPEG or PNG, RGB</td></tr>\n<tr><td>Kobo Writing Life</td><td>Not given for Kobo's own store; 500 × 500 px for its partner Bol</td><td>Width three quarters of the height</td><td>3:4, portrait</td><td>JPG or PNG; up to 5 MB</td></tr>\n<tr><td>Google Play Books</td><td>640 px</td><td>No size given</td><td>Not stated</td><td>JPEG, PNG, TIFF or PDF; up to 7,200 px a side</td></tr>\n</tbody></table>\n<p>Amazon's limits are loose: KDP takes anything from 500 pixels up, but it shows a warning at upload for a cover smaller than 2,560 × 1,600 and asks you not to stretch a small image to reach that size. Apple's <a rel=\"external\" href=\"https://help.apple.com/itc/booksassetguide/en.lproj/itc1bda991ba.html\">asset guide</a> says the same about enlarging, and adds that it rejects images that are excessively blurry or pixelated. Google's figures come from its <a rel=\"external\" href=\"https://support.google.com/books/partner/answer/3424254?hl=en-GB\">book file guidelines</a>; the same page gives audiobook covers a different rule (at least 1,024 pixels and preferably square), which is easy to mix up with the ebook one.</p>\n<p>Kobo's <a rel=\"external\" href=\"https://kobowritinglife.zendesk.com/hc/en-us/articles/360059385711\">cover image tips</a> explain its ratio by the hardware: the screens of its devices are 3:4, so the cover should be too. At 1,600 pixels wide, that's 2,133 tall, which is 427 pixels shorter than the Amazon size. The practical answer is one master at 1,600 × 2,560 for Amazon, Apple and Google, and a copy for Kobo with about 213 pixels trimmed from the top and from the bottom. If the title and the author's name stay clear of those two bands, the Kobo copy loses only background. Both sizes also meet the only pixel limits on Kobo's page, which are for its partner store Bol: 500 to 6,000 pixels a side.</p>\n<h2 id=\"dpi-and-colour\">DPI and colour</h2>\n<p>Three of the stores also give a figure in dots per inch, a measure that matters for print far more than for a screen. It's 300 in each case: the <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G6GTK3T3NUHKLEFX\">KDP recommendation</a>, Apple's rule of thumb, and Kobo's best, though Kobo takes files as low as 72 DPI too. (Google sets a 72 DPI minimum only for audiobook covers.) The DPI figure is a note stored in the file; how sharp the cover looks on a screen depends on its pixel count. On colour, Apple requires RGB, and Amazon accepts CMYK but converts it to RGB itself.</p>\n<h2 id=\"the-cover-inside-the-book\">The cover inside the book</h2>\n<p>The image you upload to the store is the marketing cover, the one on the product page. The book file carries its own copy. In an EPUB 3 file, the cover is the one image in the package that has the <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-cover-image\"><code>cover-image</code> property</a>; the standard recommends setting it and allows only one. Google won't take an EPUB <a rel=\"external\" href=\"https://support.google.com/books/partner/answer/3424254?hl=en-GB\">without the front cover image inside it</a>. <a rel=\"external\" href=\"https://help.apple.com/itc/booksassetguide/en.lproj/itc1bda991ba.html\">Apple's limits</a> for the inner copy are tighter than for the marketing one: interior images, the in-book cover included, may have at most 5.6 million pixels. A 1,600 × 2,560 cover has about 4.1 million and fits; a 2,000 × 3,200 cover has 6.4 million and doesn't. A book whose internal cover is too small fails <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G6GTK3T3NUHKLEFX\">Amazon's quality checks</a>; the image should fill at least half of the space it gets. (For more on where the cover sits in the file, see <a href=\"/blog/what-is-inside-an-epub-file/\">what is inside an EPUB file</a>, and for Amazon's other upload rules, <a href=\"/blog/kindle-direct-publishing-file-formats/\">Kindle Direct Publishing file formats</a>.)</p>\n<h2 id=\"covers-as-thumbnails\">Covers as thumbnails</h2>\n<p>Stores also show every cover as a small thumbnail beside the title in a list, and that's where a cover made for a full page can go wrong. Pictures of 3D box sets suffer on Kobo's site: the blank space around the boxes makes the cover look smaller, so its <a rel=\"external\" href=\"https://kobowritinglife.zendesk.com/hc/en-us/articles/360059385711\">tips</a> ask for a flat 2D image instead. Google lets you upload the cover as a separate file even when the EPUB already contains one, and it <a rel=\"external\" href=\"https://support.google.com/books/partner/answer/3424254?hl=en-GB\">draws the thumbnail</a> from that file, so the two can differ. A white or very pale cover seems to vanish against the white page around it. KDP's guidelines suggest <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G6GTK3T3NUHKLEFX\">a grey border of 3 or 4 pixels</a> for one, so the cover keeps an edge.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ebook-has-no-table-of-contents/","url":"https://readwyn.com/blog/ebook-has-no-table-of-contents/","title":"Why does my ebook have no table of contents?","summary":"Why an ebook's contents menu can be empty or list file names, where an EPUB keeps its table of contents, and why a converted book can lose it.","content_html":"<p>You open the contents menu on your e-reader and it's empty, or it lists file names such as <code>text/ch07.xhtml</code> where the chapters should be, even though the book has a contents page near the front. That page is only text with links. The menu is filled from a separate, machine-readable table of contents inside the file, and your book either lacks one or has the kind your reader ignores.</p>\n<p>An EPUB can carry three tables of contents. One is the page you can read. Another is the NCX, an XML file that the <a rel=\"external\" href=\"https://idpf.org/epub/20/spec/OPF_2.0.1_draft.htm\">EPUB 2 package specification</a> made compulsory: an EPUB 2 book lists it in its manifest and names it in the <code>toc</code> attribute of the spine. The third is the EPUB 3 navigation document, an XHTML file holding a <code>&lt;nav epub:type=\"toc\"&gt;</code> list of links. <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/\">EPUB 3.3</a> requires exactly one such list in every book and files the NCX under legacy features, there for EPUB 2 reading systems. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-rs-33/\">reading system specification</a> obliges an EPUB 3 reader to show you its links.</p>\n<p>Reading apps differ in what they do when that list is missing. Kobo spells out its rules in the <a rel=\"external\" href=\"https://github.com/kobolabs/epub-spec\">EPUB guidelines</a> it keeps on GitHub, which I read in November 2026: for an EPUB 2 the menu comes from the NCX, and with no NCX a Kobo lists the files of the spine; a file the NCX leaves out gets an entry named after the file or the opening words of the section. For an EPUB 3, Kobo reads the <code>toc</code> nav alone and ignores any NCX. With no <code>toc</code> nav, Kobo e-readers and the desktop app fall back to file names, the iOS app to the <a href=\"/blog/ebook-page-numbers-and-locations/\">page list</a>, and the Android app to the page list or the landmarks. If the Android app finds neither, its menu is empty.</p>\n<p>Amazon insists on the machine-readable kind. Its KDP help page on <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G201605710\">the table of contents</a> (November 2026) calls a logical table of contents, built from a <code>toc</code> nav or an NCX, “Required for all Kindle books by Amazon”, because it fills the Go To menu. The visible contents page is only strongly recommended.</p>\n<h2 id=\"how-a-book-loses-it\">How a book loses it</h2>\n<p>A converter can keep only a table of contents that it knows how to read, and the file you start from may not have one at all. calibre's <a rel=\"external\" href=\"https://manual.calibre-ebook.com/conversion.html\">conversion manual</a> describes the fallback: without a table of contents in the input's metadata, calibre detects chapters (by default, <code>h1</code> and <code>h2</code> headings containing words such as “chapter” or “part”, or elements with the class <code>chapter</code>) and adds links from the text when it finds fewer chapters than a set threshold. Chapter titles set as plain bold, centred paragraphs match none of those rules, so the menu comes out thin or empty. Readwyn's PDF conversion takes the contents from the PDF's outline, or from its headings when the PDF has no outline.</p>\n<p>An EPUB can be repaired. calibre's editor has Tools › Table of Contents › Edit Table of Contents, which its <a rel=\"external\" href=\"https://manual.calibre-ebook.com/edit.html\">manual</a> says can generate entries from the headings, from links, or from the individual files, and lets you aim any entry by clicking in a preview.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ebook-page-numbers-and-locations/","url":"https://readwyn.com/blog/ebook-page-numbers-and-locations/","title":"Why ebooks have locations instead of page numbers","summary":"Ebook text reflows, so a page depends on your screen and font. Kindle locations, print page numbers and the EPUB page list stand in for it.","content_html":"<p>An ebook has no pages of its own because its text reflows: a page exists only once an app lays the words out on one screen, in one font, at one size. Make the type bigger and the same chapter takes more screens. A location belongs to a fixed block of the text instead, so it stays the same whatever you do to the font. Two Kindle readers can meet at location 2,140, but someone holding the hardback can't.</p>\n<p>That mismatch is why Amazon, on 7 February 2011, <a rel=\"external\" href=\"https://phys.org/news/2011-02-kindle-readers-friends-page.html\">released a preview of software</a> that gave Kindle books page numbers taken from the print edition; until then they had only locations. Customers wanted numbers that matched print so they could cite passages and keep up with a book club or a class. Amazon added that earlier page numbers in ebooks had not matched any printed book. According to the <a rel=\"external\" href=\"https://www.csmonitor.com/Innovation/Horizons/2011/0208/Amazon-Kindle-gets-real-page-numbers\">Christian Science Monitor</a>, the Kindle would show both kinds of number, and David Pogue of the New York Times called the underlying problem “insoluble”. The new page numbers only existed in books that someone had mapped to a printed edition, which at launch meant tens of thousands of titles, <a rel=\"external\" href=\"https://phys.org/news/2011-02-kindle-readers-friends-page.html\">AFP reported</a>.</p>\n<p>The mapping is still done book by book. The publisher puts the page numbers into the EPUB, matched to the print edition closest to the ebook, and gives that edition's ISBN in the metadata, as Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GH4DRT75GWWAGBTU\">Kindle publishing guidelines</a> asked in November 2026. Labels may only be Roman or Arabic numerals, with no word “Page” in them, since the Kindle adds that itself. You can't see the result in Kindle Previewer or on a sideloaded copy; the numbers show up only once the book is on sale, and its product page mentions them. The same guidelines tell publishers to strip the page numbers out of a printed table of contents, so an entry such as “Chapter 1 ... P. 36” becomes a plain link to Chapter 1.</p>\n<p>In the EPUB standard the mechanism is called a page list. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-nav-pagelist\">EPUB 3.3 specification</a> defines it as a list of links to fixed page boundaries, which can come from a print edition or be made for the ebook alone, and a book may have only one. The spot in the text where each page starts can be marked with the <code>pagebreak</code> term, and a <code>dc:source</code> entry in the metadata, marked as the <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-source-of\">source of the pagination</a>, names the printed edition by its ISBN. EPUB 2 books do the same with a <code>pageList</code> in the NCX file, which the <a rel=\"external\" href=\"https://idpf.org/epub/20/spec/OPF_2.0.1_draft.htm\">EPUB 2 package specification</a> requires “if the publication is designed to allow the user to navigate to pages”; the same file holds their <a href=\"/blog/ebook-has-no-table-of-contents/\">table of contents</a>. The <a rel=\"external\" href=\"https://kb.daisy.org/publishing/docs/navigation/pagelist.html\">DAISY Knowledge Base</a> points out an odd case: nothing says what order the list must follow, and when a publisher moves the front matter to the back of the ebook, some ebook makers order the page list to match. The preface's pages then come after the last chapter.</p>\n<p>Without a page list, any page number you see is the reading app's own invention. Adobe's reading engine, which the <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Adobe_Digital_Editions\">MobileRead wiki</a> says ran on more than 50 devices, gives each file of the book one page per 1,024 bytes of its compressed size and spreads the breaks evenly through the characters. A small screen rarely holds that much text, so one Adobe page can last several screens; the wiki describes Sony Readers showing “3-7” for the seventh screen of page 3.</p>\n<p>Readwyn estimates too: a highlight synced from a Kobo is shown with its printed page if the EPUB has a page list or page break markers, and otherwise with a page worked out from the highlight's position in the text and the book's page count, which by default comes from the <a href=\"/blog/ebook-page-count-from-word-count/\">word count</a>.</p>\n<p>Kobo itself lets you choose. Its <a rel=\"external\" href=\"https://api.kobobooks.com/1.0/ReleaseNotes/121\">release notes</a> say the header and footer show both chapter progress and book progress, and that under Reading settings › Reading progress you can pick page numbers, percent read or time remaining, an option that Good e-Reader <a rel=\"external\" href=\"https://goodereader.com/blog/?p=206937\">dates</a> to firmware 4.19.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ebook-shows-boxes-instead-of-letters/","url":"https://readwyn.com/blog/ebook-shows-boxes-instead-of-letters/","title":"Why does my ebook show boxes instead of letters?","summary":"Empty boxes in an ebook mean the font in use has no glyph for those characters. The fix is a font, in the book or on the device, that covers the script.","content_html":"<p>The font your reading app draws with has no shape (a glyph) for those characters, and the app found no other font on the device that has one. The text in the file is usually intact. Type professionals call the small empty rectangles <a rel=\"external\" href=\"https://www.npr.org/sections/codeswitch/2014/08/03/337168933/-no-tofu-doesn-t-equate-to-no-problem-for-google-universal-typeface\">tofu</a>, as NPR's Code Switch blog explained in August 2014. The fix is a font that covers the script: the one the publisher embedded in the book, if there is one, or any font on the device that has those characters.</p>\n<h2 id=\"numbers-for-characters-shapes-from-fonts\">Numbers for characters, shapes from fonts</h2>\n<p>The <a rel=\"external\" href=\"https://www.unicode.org/history/publicationdates.html\">first version of the Unicode Standard</a> came out in October 1991. Unicode gives every character a number, its code point, and leaves the look of it to fonts. A file can therefore hold perfectly good Greek or Tamil that a given font can't draw. Even Amazon's own store hedges on this: as of November 2026, Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G202124400\">KDP page on Indian languages</a> lists Hindi, Tamil, Malayalam, Marathi and Gujarati for Kindle ebooks, but recommends typing the title and author name in English, because these languages may not display correctly on every Amazon and Kindle website, app and email.</p>\n<p>When a font is asked for a character it lacks, the software falls back on glyph 0. The <a rel=\"external\" href=\"https://learn.microsoft.com/en-us/typography/opentype/spec/recom\">OpenType recommendations</a> say glyph 0 must be the <code>.notdef</code> glyph, and suggest drawing it as an empty rectangle, a rectangle with a question mark inside, or one with an X. They warn against leaving it blank, because a blank <code>.notdef</code> looks like a space and you'd never know anything was missing. Since every font draws its own <code>.notdef</code>, the boxes in one app can look different from those in another.</p>\n<p>Some fonts draw a box that tells you more. The <a rel=\"external\" href=\"https://github.com/unicode-org/last-resort-font\">Last Resort font</a>, drawn by Apple, Michael Everson of Evertype and the Unicode Consortium, carries a copyright notice that starts in 1998. Each of its glyphs stands for a whole Unicode block: a sample character of the block inside a square frame, with the block's name and code point range written along the frame, legible at large sizes. A frame around a small Devanagari letter tells you that the passage is in Devanagari, not that the file is damaged.</p>\n<h2 id=\"noto-named-after-the-boxes\">Noto, named after the boxes</h2>\n<p>Google released its first Noto fonts in 2012, <a rel=\"external\" href=\"https://www.npr.org/sections/codeswitch/2014/08/03/337168933/-no-tofu-doesn-t-equate-to-no-problem-for-google-universal-typeface\">NPR reported</a>, and the name is short for \"no tofu\". By August 2014 the family covered 100 scripts and 100,000 characters, which Jungshik Shin, an engineer on Google's text and font team, said took in more than 600 written languages. The project's <a rel=\"external\" href=\"https://github.com/notofonts/noto-fonts/blob/main/FAQ.md\">FAQ</a> still gives the aim as \"no more tofu\".</p>\n<p>The same FAQ explains why Noto comes as many files and not one: it has far more glyphs than a single font can hold, and the Chinese, Japanese and Korean set on its own is already as big as a font can be. The ceiling is in the file format. OpenType keeps the glyph count in a 16-bit field (<code>numGlyphs</code> in the <a rel=\"external\" href=\"https://learn.microsoft.com/en-us/typography/opentype/spec/maxp\"><code>maxp</code> table</a>), so a font can't have more than 65,535 glyphs. <a rel=\"external\" href=\"https://www.unicode.org/versions/latest/\">Unicode 18.0</a>, released on 16 September 2026, defines 172,808 characters.</p>\n<p>No single font can cover Unicode, so apps try one font after another. Ebook apps lay out pages with CSS, and <a rel=\"external\" href=\"https://www.w3.org/TR/css-fonts-4/\">CSS Fonts Level 4</a> spells out the order: each font the text names, then the fonts installed on the device, and only then a missing glyph symbol, from a Last Resort font or the like. One kind of character never reaches the device fonts. For a code point in Unicode's Private Use Area, the app must look only in the fonts the text names. A publisher who put a decorative initial or a symbol there is betting on their embedded font; if it doesn't load, that character is a box every time.</p>\n<h2 id=\"where-the-boxes-come-from-in-an-ebook\">Where the boxes come from in an ebook</h2>\n<p>A book doesn't have to rely on the fonts of the device, because EPUB lets it bring its own. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-rs-33/\">EPUB 3.3 rules for reading systems</a> say they must support TrueType, OpenType, WOFF and WOFF2 fonts that the book's style sheet loads with <code>@font-face</code>. When the publisher embedded a font with the right script and the app uses it, the text displays as the publisher saw it. Boxes come when the book embeds no such font and the device has none, when your settings replace the book's fonts, or when an embedded font fails to load.</p>\n<p>Your own font setting is the easiest of these to check. On a Kobo, the Font Face list (tap the middle of the page, then the Text icon) swaps in a font of your choice, and <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017639913-Adjust-font-size-and-change-the-font-style-on-your-Kobo-eReader\">Kobo's help page</a> says the eReader remembers that choice and applies it to the other books you read. A font picked for an English novel may have no Greek or Cyrillic at all. Kobo eReaders accept extra <code>.otf</code> and <code>.ttf</code> files in a folder called <code>fonts</code> on the device, <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/13009477876631-Load-fonts-onto-your-Kobo-eReader\">Kobo says</a>, so you can add a font with wider coverage (see <a href=\"/blog/add-fonts-to-kobo-and-kindle/\">adding fonts to a Kobo or a Kindle</a> and <a href=\"/blog/why-ebook-fonts-look-different/\">why an ebook ignores your font setting</a>).</p>\n<p>An embedded font can also fail to load. <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/\">EPUB 3.3</a> lets a publisher obfuscate an embedded font so that it can't be installed or reused straight from the book, but the reading system rules say only that apps should undo the obfuscation, not that they must. They also note that EPUB 3 allowed obfuscation before version 3.0.1 without fixing whether it came before or after compression, so an app can end up with an invalid font from an older file. The text then falls back to the device's fonts, which may lack the script.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/epub-metadata-dublin-core/","url":"https://readwyn.com/blog/epub-metadata-dublin-core/","title":"What metadata is inside an EPUB?","summary":"The Dublin Core fields in an EPUB's package file: the four that are required, how roles, sort names, dates and series are written, and how EPUB 2 differs.","content_html":"<p>An EPUB 3 file needs only four pieces of metadata: <code>dc:identifier</code>, <code>dc:title</code>, <code>dc:language</code>, and exactly one <code>dcterms:modified</code> timestamp. <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-pkg-metadata\">EPUB 3.3</a> makes everything else optional, the author included. Its own minimal example is Norwegian Wood with an identifier, a title, the language <code>en</code> and a timestamp, and no Murakami anywhere. The record sits in the <code>metadata</code> section of the package document, the <code>.opf</code> file described in <a href=\"/blog/what-is-inside-an-epub-file/\">what is inside an EPUB file</a>.</p>\n<h2 id=\"where-the-dc-names-come-from\">Where the dc: names come from</h2>\n<p>The <code>dc:</code> prefix stands for Dublin Core, and the Dublin is the one in Ohio. OCLC and the National Center for Supercomputing Applications held <a rel=\"external\" href=\"https://www.dublincore.org/about/history/\">a workshop there in March 1995</a>, where more than 50 people worked on a common set of terms for describing things on the web, and named the result after the town. Its <a rel=\"external\" href=\"https://www.dublincore.org/specifications/dublin-core/dces/\">fifteen elements</a> became a standard in 1998 as IETF RFC 2413, and were later published as ANSI/NISO Z39.85 and ISO 15836. EPUB takes all fifteen. Besides the three required ones, a package can hold creator, contributor, publisher, date, description, subject, rights, type, format, source, relation and coverage.</p>\n<h2 id=\"order-and-format\">Order and format</h2>\n<p>Position in the file carries meaning. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-opf-dcmes-required\">first <code>dc:title</code></a> is the main title that apps show, and the first <code>dc:language</code> is the primary language. Among authors, the first <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-opf-dccreator\"><code>dc:creator</code></a> is the primary one, and the order of the others sets their display priority. A second <code>dc:title</code> is allowed but discouraged, because apps handle it inconsistently: some drop the extra part, others join the parts in unexpected ways.</p>\n<p><code>dc:language</code> takes a BCP 47 tag such as <code>en-US</code>. The chapters don't inherit it, so each one has to declare its own language.</p>\n<p><code>dc:identifier</code> can be a UUID, a DOI or an ISBN. One of them must have an <code>id</code> that the <code>package</code> element names in its <code>unique-identifier</code> attribute, and publishers are asked to keep it when they only fix errata or correct the metadata. An ebook doesn't always need an ISBN of its own; <a href=\"/blog/isbn-for-ebooks/\">ISBNs for ebooks</a> explains when it does.</p>\n<p>EPUB 3 allows at most one <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-opf-dcdate\"><code>dc:date</code></a>, and it means the date of publication. The last change goes in <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-metadata-last-modified\"><code>dcterms:modified</code></a>, in UTC and in exactly the form <code>2016-01-01T00:00:01Z</code>. A value without the seconds or the final <code>Z</code> is invalid.</p>\n<h2 id=\"epub-2-and-epub-3-compared\">EPUB 2 and EPUB 3 compared</h2>\n<p>EPUB 2's package format, <a rel=\"external\" href=\"https://idpf.org/epub/20/spec/OPF_2.0.1_draft.htm\">OPF 2.0.1</a>, became a recommended specification on 4 September 2010. It wrote extra detail as <code>opf:</code> attributes on the Dublin Core elements. EPUB 3 moved it into separate <code>meta</code> elements that point at an element's <code>id</code> through a <code>refines</code> attribute.</p>\n<table><thead><tr><th>What is recorded</th><th>EPUB 2 (OPF 2.0.1)</th><th>EPUB 3 (EPUB 3.3)</th></tr></thead><tbody>\n<tr><td>Required</td><td>title, identifier, language</td><td>title, identifier, language, <code>dcterms:modified</code></td></tr>\n<tr><td>Creator's role</td><td><code>opf:role=\"aut\"</code></td><td><code>meta</code> with <code>property=\"role\"</code></td></tr>\n<tr><td>Name for sorting</td><td><code>opf:file-as</code></td><td><code>meta</code> with <code>property=\"file-as\"</code></td></tr>\n<tr><td>Kind of identifier</td><td><code>opf:scheme=\"ISBN\"</code></td><td><a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-identifier-type\"><code>identifier-type</code></a>, where <a rel=\"external\" href=\"https://ns.editeur.org/onix/en/5\">ONIX code <code>15</code></a> is a 13-digit ISBN</td></tr>\n<tr><td>Dates</td><td>several <code>dc:date</code>, told apart by <code>opf:event</code></td><td>one <code>dc:date</code>, plus <code>dcterms:modified</code></td></tr>\n<tr><td>Cover image</td><td>an OPF 2 <code>meta</code> element</td><td><code>cover-image</code> on the manifest item</td></tr>\n<tr><td>Series</td><td>not defined</td><td><a href=\"/blog/find-the-reading-order-of-a-series/\"><code>belongs-to-collection</code></a></td></tr>\n</tbody></table>\n<p>An EPUB 3 file may still carry the old <code>meta</code> element for <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-pkg-legacy\">EPUB 2 reading systems</a>, but EPUB 3 reading systems don't use it.</p>\n<h2 id=\"roles-and-sort-names\">Roles and sort names</h2>\n<p>Roles are usually codes from the Library of Congress's <a rel=\"external\" href=\"https://www.loc.gov/marc/relators/relaterm.html\">MARC relator list</a>. OPF 2 made that list the normative one. In EPUB 3 a role can come from any code list that a <code>scheme</code> attribute names, and the spec's examples use MARC. The list now has about 300 coded terms, from <code>aut</code> (author) and <code>trl</code> (translator) to <code>nrt</code> (narrator) and <code>cov</code> (cover designer). Someone with two jobs gets two <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-role\">role</a> entries: the spec's example is Maurice Sendak as both <code>aut</code> and <code>ill</code>. OPF 2 required the codes in lower case and allowed a home-made role only under the prefix <code>oth.</code>. The list has moved since: OPF 2 printed <code>aui</code> for the author of an introduction, and the Library of Congress now sends that term to <code>win</code>, <code>wfw</code> or <code>wpr</code>, for an introduction, a foreword or a preface.</p>\n<p>OPF 2 treats a contributor like a creator, except that their part in the book was secondary. An app with no better method was told to show every creator in the order given, and was free to leave contributors out (<a rel=\"external\" href=\"https://idpf.org/epub/20/spec/OPF_2.0.1_draft.htm#Section2.2.2\">OPF 2.0.1, sections 2.2.2 and 2.2.6</a>).</p>\n<p>The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-file-as\"><code>file-as</code></a> value is what lets an app shelve Lewis Carroll under C. OPF 2's own example files \"Rev. Dr. Martin Luther King Jr.\" as \"King, Martin Luther Jr.\". A sibling property, <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-alternate-script\"><code>alternate-script</code></a>, holds the name in another writing system, so Haruki Murakami can also appear as 村上 春樹 with <code>xml:lang=\"ja\"</code>.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/european-accessibility-act-ebooks/","url":"https://readwyn.com/blog/european-accessibility-act-ebooks/","title":"What the European Accessibility Act means for ebooks","summary":"Since 28 June 2025, ebooks, reading apps and e-readers sold in the EU must meet accessibility rules. What the Act asks of a book, and who is exempt.","content_html":"<p>If an ebook's page in the Kobo Store has a section on accessibility, covering things such as navigation, hazards and visual adjustments, that section is there because of the European Accessibility Act. Since 28 June 2025 the Act has covered ebooks sold to consumers in the EU, the apps made for reading them, and e-readers: the book files must work with assistive technology, and an e-reader put on sale since that date has to <a rel=\"external\" href=\"https://eur-lex.europa.eu/eli/dir/2019/882/oj\">offer text-to-speech</a> (what Kindles and Kobos actually do with that is in <a href=\"/blog/text-to-speech-on-e-readers/\">text-to-speech on e-readers</a>). Kobo told publishers in its <a rel=\"external\" href=\"https://kobocontentoperations.zendesk.com/hc/en-us/articles/22730102538395-Accessibility-FAQs\">accessibility FAQ</a>, last updated in July 2025, that it shows publishers' accessibility metadata on product pages, sorted under W3C headings like those, and offers a search filter for accessible books.</p>\n<p>The Act is <a rel=\"external\" href=\"https://eur-lex.europa.eu/eli/dir/2019/882/oj\">Directive (EU) 2019/882</a>, adopted on 17 April 2019. A directive binds governments rather than shops, and the <a rel=\"external\" href=\"https://commission.europa.eu/strategy-and-policy/policies/justice-and-fundamental-rights/disability/union-equality-strategy-rights-persons-disabilities-2021-2030/european-accessibility-act_en\">European Commission</a> says member states had until June 2022 to write it into national law. Ireland's version, <a rel=\"external\" href=\"https://www.irishstatutebook.ie/eli/2023/si/636/made/en/print\">S.I. No. 636 of 2023</a>, shows how far it reaches. The service it regulates is the book files together with any software, phone apps included, that exists to open and read them, and any publisher or other business that distributes ebooks counts as a provider of that service.</p>\n<p>For the books themselves, Annex I of the directive lists six requirements, which a <a rel=\"external\" href=\"https://www.w3.org/TR/epub-a11y-eaa-mapping/#x2-1-2-1-f-e-books\">W3C note quotes in full</a>. If a book has audio as well as text, the two must be synchronised. The files can't get in the way of assistive technology. You have to be able to reach the content, move around it and change how it's presented, and use it with a range of assistive tools. Those features must be described in metadata, and DRM can't switch any of this off.</p>\n<p>That note, dated 28 August 2025, shows which standard answers each line. The metadata and the DRM lines fall to <a rel=\"external\" href=\"https://www.w3.org/TR/epub-a11y-11/\">EPUB Accessibility 1.1</a>, a W3C Recommendation of 17 October 2024. WCAG 2 covers the two lines about assistive technology, and EPUB 3.3 handles the synchronised audio, which it calls media overlays. Under EPUB Accessibility 1.1 every EPUB must carry the schema.org fields <code>accessMode</code>, <code>accessibilityFeature</code> and <code>accessibilityHazard</code>, and should add an <code>accessibilitySummary</code> in plain words that also admits gaps, such as missing extended descriptions. The note concedes that <a href=\"/blog/fixed-layout-vs-reflowable/\">fixed-layout</a> books (comics, art books, illustrated children's books) can't meet every requirement, because their text can't be restyled.</p>\n<p>How a shop should show those fields is set out in a W3C community group's <a rel=\"external\" href=\"https://www.w3.org/publishing/a11y/metadata-display-guide/guidelines/\">display guide</a> (version 2.1, December 2025). It groups the answers under “Ways of reading”, with lines such as “Appearance can be modified”, or “No information about appearance modifiability is available” when the publisher sent nothing.</p>\n<p>A business with fewer than 10 staff and turnover or a balance sheet of no more than €2 million is a microenterprise, and <a rel=\"external\" href=\"https://www.irishstatutebook.ie/eli/2023/si/636/made/en/print\">the Irish regulations</a> exempt the services it provides. A larger one can argue that compliance would fundamentally alter its product or be a disproportionate burden, but only after a written assessment, which it keeps for five years. A <a rel=\"external\" href=\"https://www.w3.org/TR/epub-a11y-exemption/\">W3C note</a>, in its version of 4 February 2026, gives publishers a property, <code>a11y:exemption</code>, to record such a claim inside the EPUB, with the values <code>eaa-microenterprise</code>, <code>eaa-disproportionate-burden</code> and <code>eaa-fundamental-alteration</code>. The display guide's suggested wording for readers is “Claims an accessibility exemption in some jurisdictions”, though it advises shops that this line needn't appear on public pages.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/export-kindle-highlights/","url":"https://readwyn.com/blog/export-kindle-highlights/","title":"Where your Kindle highlights are, and how to get them out","summary":"Kindle highlights sit in My Clippings.txt on the device and in Amazon's online Notebook. How to copy them out, and the publisher limit that can cut them short.","content_html":"<p>A Kindle keeps your highlights in two places, and you can copy them out of either. On the e-reader, every highlight goes into a file called My Clippings (<a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">Kindle Colorsoft user guide</a>). On the drive it's a plain text file, My Clippings.txt, that you can copy to a computer over a USB cable. On a Mac, a Kindle Scribe or a Kindle released in 2024 or later no longer shows up on its own: <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TCUBEdEkbIhK07ysFu\">Amazon's help page on USB transfers</a> (November 2026) says these models need a separate app for a cable transfer, and points to its own Send to Kindle for Mac, whose USB manager moves files to and from the computer. Highlights in books bought from Amazon also sync to your account, where the Kindle app's Notebook and the web page at <a rel=\"external\" href=\"https://read.amazon.com/notebook\">read.amazon.com/notebook</a> list them book by book.</p>\n<p>The clippings file goes back to that first Kindle. Its <a rel=\"external\" href=\"http://g-ecx.images-amazon.com/images/G/01/digital/fiona/general/Kindle_User_Guide.pdf\">2007 user guide</a> puts My Clippings.txt in the documents folder and says it's kept as plain text so that you can bring it into other programs on your computer. You can't remove an entry from the file one at a time, and editing it on your computer leaves the highlights inside the books untouched. If you delete or move the file, the Kindle starts a fresh one at your next highlight. In 2007 Amazon backed up the annotations inside each book to its servers, but not the clippings file. Since every book writes into that one file, you'll have to pick out the entries of the book you want.</p>\n<p>In the Kindle app the list is called My Notebook. The <a rel=\"external\" href=\"https://apps.apple.com/us/app/amazon-kindle/id302584613\">app's App Store description</a> says it gathers all your notes in one place, and that highlights and notes sync along with your reading position. On a Mac you can export it: Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GFBT6Y7AFLJY947N\">keyboard shortcuts for Kindle for Mac</a> list \"Export Notes from Notebook\" under Command, Alt, E, for when the Notebook of a book is open. For a document you sent to the app yourself, <a rel=\"external\" href=\"https://docs.readwise.io/readwise/docs/importing-highlights/kindle\">Readwise's guide to Kindle imports</a> gives a way to email the highlights out: open the document, tap the Notebook icon at the top right, tap Share, and pick Email.</p>\n<p>In some books the copy comes out shorter than what you marked. <a rel=\"external\" href=\"https://docs.readwise.io/readwise/docs/importing-highlights/kindle\">Readwise</a>, a service that imports Kindle highlights, says publishers require Amazon to apply a \"copyright export limit\" to the highlighted text of each book. By its account the limit varies from book to book but usually sits around 10 to 20 per cent of the text, and highlights past it are truncated. It adds that the Notebook page shows a warning on a book that has hit the limit.</p>\n<p>Books that didn't come from the Kindle Store follow other rules, and the two sources don't quite agree on them. The Colorsoft guide says highlights in a document sent with Send to Kindle sync across your devices through Whispersync, as long as the document is archived in Kindle format. Readwise says only highlights from ebooks bought directly from Amazon reach the cloud, and its method for a document you sent yourself goes through the app's Notebook and email, not the web page. Neither names an export on the web page itself. A book copied over USB never passes through Amazon at all. Its highlights stay on the e-reader, with the book and in My Clippings.txt.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/find-the-reading-order-of-a-series/","url":"https://readwyn.com/blog/find-the-reading-order-of-a-series/","title":"How to find the reading order of a book series","summary":"Where to look up the reading order of a book series, why publication and story order can differ, and how an EPUB file records a book's place in a series.","content_html":"<p>Look the series up on the author's or the publisher's site, then check that list against a bibliography that gives the year each book first came out. If the years don't run in the same order as the numbers on the spines, the series has two orders, the one it was published in and the one the story happens in, and you'll have to pick.</p>\n<h2 id=\"when-a-series-has-two-orders\">When a series has two orders</h2>\n<p>Narnia shows how it happens. C. S. Lewis's seven books came out from 1950 to 1956, starting with <em>The Lion, the Witch and the Wardrobe</em>. <em>The Magician's Nephew</em> was the sixth to appear (1955), yet it tells how Narnia began, and the table in the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Chronicles_of_Narnia#Reading_order\">Wikipedia article on the series</a> puts <em>The Horse and His Boy</em> (1954) third in the story, in Narnian year 1014. The first editions weren't numbered. Macmillan, the first American publisher, numbered them in publication order, while some early British editions already gave the story order. When HarperCollins took over the rights in 1994 it chose the story order, <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Chronicles_of_Narnia#Publication_history\">at the suggestion of</a> Lewis's stepson, Douglas Gresham. Which book is \"1\" depends on the edition you hold.</p>\n<p>Gresham's case rests on a 1957 letter in which Lewis told an American boy that he agreed with the boy's chronological order more than with the boy's mother's. As <a rel=\"external\" href=\"https://web.archive.org/web/20080929080302/http://www.usatoday.com/life/movies/news/2005-12-01-narnia-side_x.htm\">USA Today reported in December 2005</a>, readers who prefer publication order think Lewis was being polite to a child. Andrew Adamson, who directed the first film, told the paper he sides with publication order, the way he read them as a child in New Zealand: the lamp post's origin works better once its mystery has built up. Readers who prefer publication order also point to the text, as <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/The_Chronicles_of_Narnia#Reading_order\">Wikipedia notes</a>. When Aslan is first named in <em>The Lion, the Witch and the Wardrobe</em>, the narrator tells the reader that the children knew no more about him than you do, which isn't true for anyone who began with <em>The Magician's Nephew</em>.</p>\n<h2 id=\"what-the-ebook-file-records\">What the ebook file records</h2>\n<p>EPUB 3 has fields for this in the package metadata. In the <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-belongs-to-collection\">EPUB 3.3 specification</a>, <code>belongs-to-collection</code> names the collection, and <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-collection-type\"><code>collection-type</code></a> says what kind it is: <code>series</code> for related works that are usually open-ended and come out one at a time, <code>set</code> for a finite group of works that make one whole and usually come out together. The spec's examples call both Harry Potter and The Lord of the Rings a set. Reading systems aren't required to support either value. The spec also asks the publisher to give the collection an identifier, so that two series with the same name, or two editions of one series, can be told apart.</p>\n<p>The number goes in <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-group-position\"><code>group-position</code></a>, which takes a whole number or a run of numbers separated by dots, such as 2.2.1. Each dot is a level: the spec's own example gives 98.4 for volume 98, issue 4 of the journal <em>Physical Review D</em>. A novella between books 2 and 3 numbered 2.5 would, by that reading, be part 5 of book 2. The number in a file is the one that edition's publisher chose, and for Narnia that depends on whose numbering it followed. If you keep your own library, set the number to the order you settled on; Readwyn's edit details sheet has a series field and a number for each book. The other metadata fields are in <a href=\"/blog/epub-metadata-dublin-core/\">EPUB metadata and Dublin Core</a>.</p>\n<h2 id=\"where-to-look\">Where to look</h2>\n<ol>\n<li>Check the author's, estate's, or publisher's website. Penguin's <a rel=\"external\" href=\"https://www.terrypratchettbooks.com/discworld-reading-order/\">Discworld reading order page</a> says Terry Pratchett's 41 Discworld novels, published over 32 years, can be read in any order, and it offers two lists to download: one by publication date, and one that follows a group of characters such as the City Watch or the Witches.</li>\n<li>For science fiction, fantasy, and horror, search the <a rel=\"external\" href=\"https://www.isfdb.org/\">Internet Speculative Fiction Database</a> (ISFDB). Volunteers have built it <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Internet_Speculative_Fiction_Database\">since 1995</a>, and its author bibliographies record each series.</li>\n<li>For any genre, search Kent District Library's <a rel=\"external\" href=\"https://ww2.kdl.org/\">What's Next database</a> by author, series, or title, with a filter for genre. The library counts any two or more books linked by character, setting, or other shared traits as a series.</li>\n<li>Open the series article on Wikipedia, if there is one, and look for a table of publication years; some, like the Narnia one, add the year inside the story.</li>\n<li>Look at the front matter of the ebook. An \"Also by\" page lists the other books, but in the order that edition's publisher chose.</li>\n</ol>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/first-e-reader-rocket-ebook/","url":"https://readwyn.com/blog/first-e-reader-rocket-ebook/","title":"What was the first e-reader?","summary":"The V&A calls Sony's Data Discman of 1990 the first e-reader; the Rocket eBook and SoftBook came in 1998, and Sony's Librie brought E Ink in 2004.","content_html":"<p>Sony's Data Discman went on sale in Japan in 1990, and the Victoria and Albert Museum, which keeps a DD-1EX in its collection, calls it the <a rel=\"external\" href=\"https://collections.vam.ac.uk/item/O1325440/sony-dd1-ex-electronic-book-electronic-book-reader-sony-corporation/\">world's first electronic book reader</a>. Sony called it a \"player\". If you mean a handheld that loads a book you bought and shows it one page at a time, the answer is 1998, <a rel=\"external\" href=\"https://web.archive.org/web/2000/http://www.cnn.com/TECH/computing/9812/18/ebooks.idg/index.html\">the year of the Rocket eBook and the SoftBook</a>. The first commercial reader with an E Ink screen was Sony's Librie, <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Sony_Librie\">sold only in Japan</a> from 2004.</p>\n<h2 id=\"from-a-disc\">From a disc</h2>\n<p>The Data Discman played 3.15-inch discs that held up to 100,000 pages of text or 32,000 pictures. Sets usually came with an encyclopaedia and a travel dictionary, and Sony aimed the machine at students and travellers. It weighed 705 g with its battery, and a full charge lasted up to three hours. Japan bought 90,000 in the first eight months. The US, where the player arrived in September 1991 at <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Data_Discman\">$550</a>, never took to it: few titles came out, and the price was beyond most students. You searched it by keyword, the way you look something up in a dictionary, and the <a rel=\"external\" href=\"https://collections.vam.ac.uk/item/O1325440/sony-dd1-ex-electronic-book-electronic-book-reader-sony-corporation/\">V&amp;A</a> counts only one literary disc for it, The Library of the Future.</p>\n<h2 id=\"down-a-serial-cable-or-down-the-phone-line\">Down a serial cable, or down the phone line</h2>\n<p>Martin Eberhard and Marc Tarpenning, who went on to found Tesla, started NuvoMedia on 15 April 1997, released the Rocket eBook in late 1998, and shipped 20,000 of them in 1999, <a rel=\"external\" href=\"https://www.businessinsider.com/tesla-the-origin-story-2014-10\">according to Business Insider</a>. The Rocket weighed 22 ounces and had a backlit black-and-white LCD of 480 by 320 pixels at 106 dpi. The standard model's 4 MB of flash held <a rel=\"external\" href=\"https://atpm.com/6.05/rocketebook.shtml\">roughly ten books</a>, a figure the reviewer called optimistic. You bought the book on the web from a Mac or a PC, for example from barnesandnoble.com, and the cradle that copied it across plugged into a serial port. In November 1998 <a rel=\"external\" href=\"https://web.archive.org/web/20000208053039/http://www.businessweek.com/1998/46/b3604010.htm\">around 400 titles were ready for download</a>, with 600 more due by Christmas.</p>\n<p>The SoftBook, from <a rel=\"external\" href=\"https://www.infotoday.com/IT/jun99/ab12.htm\">SoftBook Press of Menlo Park</a>, needed no computer at all. Behind a leather cover sat a 9.5-inch greyscale touch screen, and inside was a 33.6 kbps modem that dialled the SoftBook store over a phone line. In 2000 it <a rel=\"external\" href=\"http://www.atpm.com/6.05/softbookreader.shtml\">cost $599.95</a>, or $299.95 with a \"content package\" of $19.95 a month for 24 months. <a rel=\"external\" href=\"https://web.archive.org/web/20000208053039/http://www.businessweek.com/1998/46/b3604010.htm\">BusinessWeek reported in November 1998</a> that its battery lasted around four hours, against about nine for the Rocket. The same report said the new Open eBook format, the ancestor of <a href=\"/blog/history-of-epub/\">EPUB</a>, was \"based primarily on technology developed by SoftBook Press\".</p>\n<h2 id=\"under-new-owners\">Under new owners</h2>\n<p>In January 2000 Gemstar, whose chief executive Henry Yuen had invented the <a rel=\"external\" href=\"https://www.technologyreview.com/2005/12/29/229845/can-sony-make-e-books-succeed/\">VCR+ system</a> for programming video recorders, <a rel=\"external\" href=\"https://web.archive.org/web/20121025072227/http://news.cnet.com/Gemstar-buys-two-e-book-makers/2100-1017_3-235851.html\">bought both NuvoMedia and SoftBook</a> for stock. RCA then made the two designs as the <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/REB_1100\">REB 1100</a> and the <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/SoftBook_Reader\">REB 1200</a>. <a rel=\"external\" href=\"https://web.archive.org/web/20070716002335/http://reviews.cnet.com/e-book-readers/rca-reb1100-ebook/4505-3508_7-4744438.html\">CNET's review of the REB 1100</a> in 2001 liked the backlight and the 18-ounce body, but found the selection of books thin, and noted that copy protection tied each book to the one device: you couldn't read it on your PC or lend it to a friend.</p>\n<p>A journalist who ran NuvoMedia's website from November 1999 to April 2001 later <a rel=\"external\" href=\"https://www.technologyreview.com/2005/12/29/229845/can-sony-make-e-books-succeed/\">wrote in Technology Review</a> that the devices failed because of Gemstar's management and because most publishers charged hardcover prices for e-books. He says Gemstar never ran the $100 million marketing campaign it had promised, that nearly all the original staff of both companies had gone by mid-2001, and that shareholders forced Yuen out in 2002.</p>\n<h2 id=\"on-electronic-paper\">On electronic paper</h2>\n<p>Sony's Librie EBR-1000EP went on sale in Japan on 24 April 2004. Its 6-inch <a href=\"/blog/how-e-ink-works/\">E Ink</a> screen, <a rel=\"external\" href=\"https://www.sony.jp/CorporateCruise/Press/200403/04-0324/\">Sony's Japanese press release</a> says, had 800 by 600 pixels at about 170 ppi and four shades of grey. Because electronic paper draws power only when the page changes, Sony rated four AAA batteries for about 10,000 pages. The reader weighed about 190 g and held around 20 novels in its 10 MB of memory. Books came from the Timebook Town store in Sony's BBeB format, locked with Sony's OpenMG copy protection, and the Technology Review piece notes that they erased themselves after 60 days.</p>\n<p>Sony announced the Reader PRS-500 at CES in January 2006 and <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/PRS500\">released it that September</a>. It kept the 6-inch, 800 by 600 screen with four greys and weighed 8.8 ounces. By the time <a rel=\"external\" href=\"https://web.archive.org/web/20071217050355/http://reviews.cnet.com/e-book-readers/sony-prs-500-portable/4505-3508_7-31660696.html\">CNET reviewed it</a> in mid-2007 it cost $300, and Sony quoted 7,500 page turns on a charge. CNET's complaints echoed those about the first generation: books from Sony's Connect store worked on no other device, often cost as much as paper ones, and the Connect software didn't run on a Mac.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/first-kindle-2007/","url":"https://readwyn.com/blog/first-kindle-2007/","title":"Did the first Kindle come out in 2007, and what was it like?","summary":"Amazon's first Kindle went on sale on 19 November 2007 for $399 with a free mobile data link, and its first stock sold out in five and a half hours.","content_html":"<p>Yes. Amazon put the first Kindle on sale on 19 November 2007 for $399, and its <a rel=\"external\" href=\"https://press.aboutamazon.com/2007/11/introducing-amazon-kindle\">launch press release</a> said it could be ordered from amazon.com that same day. Its main selling point was the mobile data radio inside: you could buy a book on the device itself and be reading it in under a minute, with no computer, no Wi-Fi hotspot and no monthly bill.</p>\n<h2 id=\"the-device\">The device</h2>\n<p>According to the release, the Kindle weighed 10.3 ounces (about 290 g) and had a black and white <a href=\"/blog/how-e-ink-works/\">electronic paper</a> screen with no backlight. It also had a standard-layout keyboard, for searching and for typing notes against the text as you would in a margin. Long page-turn buttons ran down both edges, so a left-handed reader could hold it in one hand as easily as a right-handed one. The built-in memory held more than 200 titles and an SD card added more. There were six font sizes, The New Oxford American Dictionary was built in, and the Kindle could open Wikipedia. With the wireless on, Amazon said it needed charging about every other day; with it off, a charge lasted a week or more.</p>\n<p>The wireless delivery had its own name, Whispernet. Amazon's release only says it ran on EVDO, the data network of mobile phones; <a rel=\"external\" href=\"https://www.computerworld.com/article/1571212/amazon-unveils-e-book-reader-kindle.html\">Computerworld's launch report</a> names it as Sprint's EV-DO network, paid for by Amazon. Jeff Bezos said in the release that the team had been working on the Kindle for more than three years.</p>\n<h2 id=\"what-you-could-read-on-it\">What you could read on it</h2>\n<p>The Kindle Store opened with more than 90,000 books, including 101 of the 112 current New York Times best sellers and new releases, priced at $9.99 unless marked otherwise. Newspapers cost $5.99 to $14.99 a month and magazines $1.25 to $3.49. There were also over 300 blogs, Slashdot and TechCrunch among them, from $0.99 a month each, and the first chapter of most books could be downloaded free.</p>\n<p>Each Kindle had its own email address, the ancestor of today's <a href=\"/blog/send-epub-to-kindle/\">Send to Kindle</a>. Sending it a Word document or a picture cost $0.10, and the release lists the files it accepted: unprotected Word, HTML, TXT, JPEG, GIF, PNG and BMP. PDF wasn't on that list. Authors could upload their own books through a new self-publishing tool, the Digital Text Platform, launched the same day.</p>\n<h2 id=\"sold-out-the-first-day\">Sold out the first day</h2>\n<p>Two days after the launch, <a rel=\"external\" href=\"https://www.engadget.com/2007/11/21/kindle-sells-out-in-two-days/\">Engadget reported</a> that the Kindle had sold out in five and a half hours. Bezos gave the same figure on FOX Business, and the <a rel=\"external\" href=\"https://www.fool.com/investing/general/2007/11/20/the-kindle-is-so-hot-its-on-fire-sale.aspx\">Motley Fool</a>, quoting him, noted that later orders were given a shipping date of 29 November. Its writer wondered whether the first batch had been kept small on purpose. Amazon gave no unit numbers, and Computerworld says Steve Kessel, the executive in charge of digital media, did not disclose any.</p>\n<p><a rel=\"external\" href=\"https://press.aboutamazon.com/2009/2/kindle-2-ships-today\">When Kindle 2 shipped</a> on 23 February 2009, Amazon said the first units went to customers \"who were in line for the original Kindle\".</p>\n<h2 id=\"the-first-kindle-beside-kindle-2\">The first Kindle beside Kindle 2</h2>\n<p>Amazon announced Kindle 2 on 9 February 2009. Its <a rel=\"external\" href=\"https://press.aboutamazon.com/2009/2/introducing-amazon-kindle-2\">press release</a> compared the two machines, and the 2007 release gives the rest of the original's figures.</p>\n<table><thead><tr><th></th><th>Kindle (November 2007)</th><th>Kindle 2 (February 2009)</th></tr></thead><tbody>\n<tr><td>Price at launch</td><td>$399</td><td>$359</td></tr>\n<tr><td>Weight</td><td>10.3 oz</td><td>just over 10 oz</td></tr>\n<tr><td>Shades of grey</td><td>4</td><td>16</td></tr>\n<tr><td>Books in memory</td><td>more than 200</td><td>more than 1,500 (2 GB)</td></tr>\n<tr><td>Battery with wireless on</td><td>about every other day</td><td>four to five days</td></tr>\n<tr><td>Books in the Kindle Store</td><td>more than 90,000</td><td>more than 230,000</td></tr>\n</tbody></table>\n<p>Kindle 2 had a 6-inch screen of 600 by 800 pixels, turned pages about 20 percent faster than the original, and added an experimental text-to-speech feature. Amazon said the extra shades of grey gave crisper text and sharper pictures.</p>\n<h2 id=\"where-the-name-came-from\">Where the name came from</h2>\n<p>Lab126, an Amazon company, asked the San Francisco designer Michael Cronan to name the device around 2005: a <a rel=\"external\" href=\"https://www.printmag.com/daily-heller/who-named-the-kindle-and-why/\">piece in Print magazine</a> from December 2008 says it happened about three years earlier. Cronan also named TiVo. His wife and business partner, <a rel=\"external\" href=\"https://www.cronan.com/\">Karin Hibma</a>, told Print that they wanted a name people would remember that was neither technical nor trite, and that Bezos wanted to talk about the future of reading modestly. The name they chose is a verb that means to set something alight, which goes back to the Old Norse word for a candle, and Hibma tied it to a line of Voltaire's comparing what we learn from books to fire, something lit at home and passed on to others.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/fixed-layout-vs-reflowable/","url":"https://readwyn.com/blog/fixed-layout-vs-reflowable/","title":"Fixed layout or reflowable: why some ebooks won't let you change the font","summary":"A comic, a picture book or a PDF ignores your font settings because its pages are fixed in advance, while a reflowable ebook is re-set for each screen.","content_html":"<p>Open a PDF on a Kobo e-reader and you can't make the type any bigger. The help page Kobo keeps for PDFs (read in November 2026) lists what you <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017513414-Turn-pages-magnify-and-rotate-PDFs-using-your-Kobo-eReader\">can't do with one</a>: change the text size or style, select the text, look up a word. You zoom with a slider or a pinch and drag the page around instead. That's the mark of a fixed layout book. A reflowable ebook is a stream of text that the reader breaks into pages to fit your screen, font, and margins, so it can be re-set at any size. A fixed layout book (a PDF, or an EPUB marked as fixed) is a series of pages whose size and contents the publisher set in advance, and the reader can only scale the whole page.</p>\n<h2 id=\"how-an-epub-declares-it\">How an EPUB declares it</h2>\n<p>Every EPUB has a package file, the <code>.opf</code>, and one line in it decides the layout. Under the <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/\">EPUB 3.3 standard</a>, the <code>rendition:layout</code> property is either <code>reflowable</code>, the default, or <code>pre-paginated</code>, which means each content file in the reading order becomes exactly one page. A fixed XHTML page must state its size in a viewport tag (the standard's own example is <code>width=1200, height=600</code>), and everything on the page is positioned on that canvas as on a printed sheet. A book can also mix the two: EPUB 3.3 lets a publisher mark single files with <code>rendition:layout-pre-paginated</code>, so a reflowable novel can carry a few fixed pages, such as a map or a full-page illustration. The rules are older than EPUB 3.3. The International Digital Publishing Forum approved <a rel=\"external\" href=\"https://idpf.org/epub/fxl/\">EPUB 3 Fixed-Layout Documents</a> on 13 March 2012. Its authors came from Sony, Adobe, Barnes &amp; Noble, Google, Apple, Hachette and others, and the page now says the main EPUB specifications have taken its contents over.</p>\n<p>The fixed layout section of EPUB 3.3 also explains why the font won't change. It notes that reading apps usually block or limit your own style settings on pre-paginated pages, because changing the type would break a design that was drawn around it, and it asks publishers to weigh what that costs in usability and accessibility before choosing fixed layout.</p>\n<h2 id=\"which-books-use-it\">Which books use it</h2>\n<p>The W3C note on <a rel=\"external\" href=\"https://www.w3.org/TR/2025/NOTE-epub-fxl-a11y-20250925\">fixed layout accessibility</a>, edited by Wendy Reid (September 2025 version), lists children's picture books, comics, manga, and cookbooks as the usual cases. Amazon's KDP help tells publishers to make <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GY6SCWBEHVEGCLCN\">comic ebooks in fixed layout</a>, and it lists X-Ray, Word Wise, and text-to-speech among the features of reflowable books.</p>\n<p>The cost falls hardest on readers who need larger or spaced text. The same W3C note says fixed layout automatically fails two WCAG success criteria, 1.4.4 Resize Text and 1.4.12 Text Spacing, and can't meet 1.4.10 Reflow. It also points out that where the lettering is hand-drawn or part of the image, the text has to stay a picture, and then there's no text to resize at all.</p>\n<h2 id=\"on-a-6-inch-screen\">On a 6-inch screen</h2>\n<p>The <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-clara-bw\">Kobo Clara BW</a> has a 6-inch screen of 1448 by 1072 pixels, which works out to about 300 pixels per inch and an image area of roughly 9 by 12 cm. A fixed page has to shrink to fit that, whatever size it was designed at. If the book asks for two-page spreads in both orientations (<code>rendition:spread</code> set to <code>both</code>) and the reader honours that request, each page of the pair gets about 536 pixels of width in portrait, roughly 4.5 cm. With the screen turned sideways, each page of a spread gets up to 724 pixels, and Kobo lets you rotate a PDF by 90 degrees from its reading menu.</p>\n<p>A PDF imported into Readwyn gets a reflowable EPUB 3 copy, with the running headers and page numbers removed, so it can be <a href=\"/blog/read-pdf-as-reflowable-text-on-mac/\">read at any font size</a>. That needs a text layer and no encryption: a scanned or encrypted PDF stays as it is.</p>\n<h2 id=\"how-to-check-an-epub-yourself\">How to check an EPUB yourself</h2>\n<p>An EPUB is a ZIP archive, so you can read its package file with any archive tool:</p>\n<ol>\n<li>Make a copy of the <code>.epub</code> and change the copy's extension to <code>.zip</code>, then open it.</li>\n<li>Open <code>META-INF/container.xml</code>. The <code>full-path</code> of its <code>rootfile</code> element names the package file, for example <code>OEBPS/content.opf</code>.</li>\n<li>Open that <code>.opf</code> in a text editor and search for <code>rendition:layout</code>. If a <code>meta</code> element says <code>pre-paginated</code>, the book is fixed, apart from any <code>itemref</code> in the <code>spine</code> marked <code>rendition:layout-reflowable</code>. If the property is missing, the book is reflowable, apart from any pages that step 4 finds.</li>\n<li>Search the <code>spine</code> for <code>rendition:layout-pre-paginated</code>. Each <code>itemref</code> that carries it is a fixed page inside an otherwise reflowable book.</li>\n</ol>\n<p>The same archive holds the book's stylesheets and pages; <a href=\"/blog/what-is-inside-an-epub-file/\">what is inside an EPUB file</a> goes through the rest of its folders.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/front-light-and-warm-light/","url":"https://readwyn.com/blog/front-light-and-warm-light/","title":"How an e-reader's front light works, and why it isn't a backlight","summary":"What a front light on an e-reader is, how it differs from the backlight of a tablet, which models had it first, and how warm light changes its colour.","content_html":"<p>Drag the brightness slider up on a Kindle or a Kobo in a dark room and the page starts to glow, but none of that light comes from behind the text. A front light is a set of small LEDs in the edge of the screen that shine sideways into a thin clear sheet lying on top of the E Ink. The sheet spreads the light over the whole page and bends it down onto the ink, which bounces it back up to you, much as paper reflects a bedside lamp. E Ink's <a rel=\"external\" href=\"https://www.eink.com/tech/detail/Front_Light\">description of its front light</a> puts the LEDs in the sides of the display module and calls the sheet a light guide.</p>\n<h2 id=\"why-the-light-has-to-come-from-the-front\">Why the light has to come from the front</h2>\n<p>The image on an E Ink screen is made of opaque black and white pigment (<a href=\"/blog/how-e-ink-works/\">how E Ink works</a>), so light shone from behind would have nothing to pass through. When Sony added a light to its Reader in 2008, <a rel=\"external\" href=\"https://abcnews.com/Technology/story?id=5948383\">ABC News explained</a> that the LEDs lit the screen from the sides because the ink couldn't be lit from behind. A phone or tablet with an LCD works the other way round. Amazon's <a rel=\"external\" href=\"https://press.aboutamazon.com/2012/9/introducing-the-new-kindle-paperwhite-the-most-advanced-e-reader-ever-constructed\">announcement of the first Kindle Paperwhite</a> set the two side by side: an LCD's backlight sends light up through the display into your eyes, while the Paperwhite's guide sends it down towards the display, under an anti-glare layer.</p>\n<table><thead><tr><th></th><th>Front light (E Ink e-reader)</th><th>Backlight (LCD tablet or phone)</th></tr></thead><tbody>\n<tr><td>Where the LEDs sit</td><td>In the edges, above the image</td><td>Behind the image</td></tr>\n<tr><td>Which way the light goes</td><td>Down onto the ink, then reflected to you</td><td>Up through the display towards you</td></tr>\n<tr><td>With the light off</td><td>Readable by room light or sunlight</td><td>Nothing to see</td></tr>\n<tr><td>LEDs, as Amazon counted them in 2012</td><td>Four in the Paperwhite</td><td>Up to 50 in an LCD</td></tr>\n</tbody></table>\n<h2 id=\"the-first-lit-e-readers\">The first lit e-readers</h2>\n<p>The <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Front_light\">MobileRead wiki</a> names Sony's PRS-700 as the first E Ink device with a front light, with two brightness levels. <a rel=\"external\" href=\"https://www.informationweek.com/it-sectors/sony-revamps-e-book-reader-but-no-wireless\">InformationWeek</a> reported its unveiling on 3 October 2008, to go on sale the next month at about $400, with the LED light pitched for rooms too dim to read in. The same wiki page records owners who felt the light made the screen look fuzzy.</p>\n<p>The idea caught on in 2012. Barnes &amp; Noble announced the Nook Simple Touch with GlowLight on 12 April 2012 at $139, to ship in May. <a rel=\"external\" href=\"https://techcrunch.com/2012/04/12/nook-simple-touch-with-glowlight/\">TechCrunch reported</a> that the battery lasted about two months with the light off and one month with it on.</p>\n<p>Amazon followed on 6 September with the <a rel=\"external\" href=\"https://press.aboutamazon.com/2012/9/introducing-the-new-kindle-paperwhite-the-most-advanced-e-reader-ever-constructed\">Kindle Paperwhite at $119</a>, shipping on 1 October and rated for eight weeks with the light on all the time. At the launch, Jeff Bezos spoke of four years of research behind the light, and <a rel=\"external\" href=\"https://www.engadget.com/2012-09-30-amazon-kindle-paperwhite-review.html\">Engadget's review</a> described an optical fibre laid flat above the screen with a nano-imprinted light guide to spread its light. Even so, Engadget saw a few spots and some unevenness near the bottom of the screen on a white page with the light turned all the way up.</p>\n<h2 id=\"how-warm-light-is-made\">How warm light is made</h2>\n<p>Warm light comes from a second set of LEDs in a different colour. The 2019 Kindle Oasis has 25 LEDs, and <a rel=\"external\" href=\"https://www.expertreviews.co.uk/amazon/amazon-kindle-oasis\">Expert Reviews</a> counted them as 12 white and 13 amber, with a slider that sets how much of the amber shows. At its launch in June 2019, <a rel=\"external\" href=\"https://www.engadget.com/2019/06/19/the-new-kindle-oasis-adjusts-its-screen-color-to-the-time-of-day/\">Engadget wrote</a> that no other Kindle could change its colour temperature, and <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/read-anywhere-with-the-all-new-kindle-oasis\">Amazon's own announcement</a> added that the screen warmth could be scheduled to change with sunrise and sunset. The Paperwhite got adjustable warm light in the model <a rel=\"external\" href=\"https://press.aboutamazon.com/uk/2021/9/amazon-unveils-the-next-generation-kindle-paperwhite-and-new-kindle-paperwhite-signature-edition\">that went on pre-order on 21 September 2021</a>, and its Signature Edition added a sensor that sets the brightness from the room's light.</p>\n<p>Kobo was three years earlier. Its ComfortLight PRO came with the Kobo Aura ONE, <a rel=\"external\" href=\"https://global.rakuten.com/corp/news/press/2016/0818_01.html\">announced on 17 August 2016</a> at CA$249.99, which had an ambient light sensor for brightness and a bedtime setting that made the hue of the light follow the sun's progress through the day. On current Kobos you tap the Brightness icon at the top of the Home screen, then use the Auto toggle or drag the colour slider towards the sun (white) or the moon (orange). <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017481174\">Kobo's help page</a> warns that a manual change stops the automatic shift. Its list of models with the feature ran to twelve when I checked it in November 2026, from the Aura ONE to the Clara Colour and Libra Colour.</p>\n<p>Kobo says on that page that blue light can keep you alert and that cutting it at night may help you sleep. That's the maker's claim, not a study; what the research says about reading before bed is in <a href=\"/blog/reading-on-an-e-reader-before-bed/\">does reading on an e-reader before bed hurt your sleep</a>. E Ink has also redesigned the light itself: its ComfortGaze front light, according to <a rel=\"external\" href=\"https://www.eink.com/tech/detail/Front_Light\">the same E Ink page</a>, cuts what the company calls the Blue Light Ratio by up to 60% compared with its previous generation.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/google-books-lawsuit/","url":"https://readwyn.com/blog/google-books-lawsuit/","title":"Is Google Books legal? What the Authors Guild case decided","summary":"How Authors Guild v. Google ended: why US courts found Google's book scanning and snippet view to be fair use, and what the 2015 ruling left open.","content_html":"<p>Search Google Books for a word in a copyrighted book whose publisher hasn't licensed it, and you get “Snippet view”: a few thin strips of the page with your word in them, and nothing more. That's legal in the United States. On 16 October 2015 the <a rel=\"external\" href=\"https://www.govinfo.gov/content/pkg/USCOURTS-ca2-13-04829/pdf/USCOURTS-ca2-13-04829-0.pdf\">Second Circuit Court of Appeals held</a> that Google's scanning of whole books without permission, its full-text search and its snippets were all fair use, and that handing digital copies to the partner libraries was lawful too. The authors asked the Supreme Court to hear the case, and it <a rel=\"external\" href=\"https://www.supremecourt.gov/search.aspx?filename=/docketfiles/15-849.htm\">denied the petition on 18 April 2016</a>, so the ruling stands.</p>\n<h2 id=\"how-snippet-view-is-built\">How snippet view is built</h2>\n<p>The court's opinion describes the feature in more detail than Google's help pages do. A snippet is normally an eighth of a page: on a book with 24 lines a page, three lines. A search for one term shows at most three snippets, and always the same three, however often you repeat it and from however many computers. You can see more by searching for other words, but Google permanently withholds one snippet on every page and one whole page in every ten, which it calls “blacklisting”. The court put that at about 22% of a book's text. Snippet view is switched off entirely for dictionaries, cookbooks and books of short poems, where one snippet might be all a reader needs, and since 2005 a rights holder can remove a book from it with an online form.</p>\n<p>Snippets are the third of four levels. Google's <a rel=\"external\" href=\"https://support.google.com/websearch/answer/9690276?hl=en-GB\">Library Project help page</a> lists full view for books in the public domain, limited preview when a rights holder allows some pages, snippet view, and no preview for books Google knows of but hasn't scanned. You can download the full-view books; our post on <a href=\"/blog/download-public-domain-books-from-hathitrust/\">getting public-domain scans from Google Books and HathiTrust</a> shows how.</p>\n<h2 id=\"why-the-court-called-it-fair-use\">Why the court called it fair use</h2>\n<p>Judge Pierre Leval wrote for the panel. He found that the copying served a purpose the books themselves don't: it let anyone find out which of more than 20 million books used a term, and how often, without giving them the book. Google showed no advertising on the search results. The plaintiffs' lawyers had researchers run searches on the authors' books for weeks; they never reached 16% of any text, and what they got was scattered across the book in no order. The <a rel=\"external\" href=\"https://www.copyright.gov/fair-use/summaries/authorsguild-google-2dcir2015.pdf\">US Copyright Office's summary</a> gives the gist: the court accepted that the search might cost some sales, but found the snippets too fragmentary to be a substitute for buying the book.</p>\n<p>The ruling has limits. It covers snippet view only “as snippet view is presently constructed”, and Leval wrote that a way to read a coherent 16% of a book would raise a different question. It's a US court applying the US fair-use statute, section 107 of the Copyright Act. Which books get full view is a separate matter of public-domain status, and <a href=\"/blog/public-domain-differs-by-country/\">that differs by country</a>: <a rel=\"external\" href=\"https://support.google.com/books/answer/43729?hl=en-GB\">Google's help page</a> asks you to request a review if you think a book is in the public domain where you live. The court also ruled on the libraries' copies. Through Google's download system the partner libraries had taken at least 2.7 million digital copies of their own volumes, and the bare chance that a library might misuse them didn't make Google liable.</p>\n<h2 id=\"from-the-2005-lawsuit-to-the-appeal\">From the 2005 lawsuit to the appeal</h2>\n<p>Google's Library Project began in 2004. By the time of the lawsuit its partners included Michigan, Harvard, Stanford, Oxford and the New York Public Library, which picked books from their shelves for Google to scan. The Authors Guild and individual authors sued on 20 September 2005. The next month <a rel=\"external\" href=\"https://www.heraldnet.com/business/publishers-gang-up-on-google/\">five publishers</a> (Simon &amp; Schuster, McGraw-Hill, John Wiley &amp; Sons, Penguin Group USA and Pearson Education) filed a separate suit in the same Manhattan court, asking a judge to stop Google from resuming its scanning on 1 November. In October 2008 the parties announced a settlement under which <a rel=\"external\" href=\"https://www.google.com/intl/en/press/pressrel/20081027_booksearchagreement.html\">Google would pay $125 million</a> and a new Book Rights Registry would pass on money from book sales, institutional subscriptions and advertising to authors and publishers.</p>\n<p>Judge Denny Chin rejected the amended version on 22 March 2011. As <a rel=\"external\" href=\"http://laboratorium.net/archive/2011/03/22/inside_judge_chins_opinion\">James Grimmelmann's reading of the opinion</a> points out, the deciding point was that it released Google from liability for future acts, beyond what a class action may settle. Chin also urged the parties to move from an opt-out scheme to an opt-in one, and <a rel=\"external\" href=\"https://www.theguardian.com/books/2011/mar/23/google-books-settlement-ruling\">The Guardian reported</a> that some 6,500 authors had already opted out.</p>\n<p>The case went back to the original question, and Chin granted Google summary judgment on 14 November 2013. The authors who appealed were Jim Bouton, Betty Miles and Joseph Goulden; <a rel=\"external\" href=\"https://eff.org/files/2014/12/01/authors_guild_brief_and_special_appendix_for_appellants.pdf\">their brief</a> names the books at issue as <em>Ball Four</em>, <em>The Trouble with Thirteen</em> and <em>The Superlawyers</em>. The Guild itself stayed on the appeal, but the court noted that a 2014 ruling in its case against the HathiTrust Digital Library had already found it lacked standing to sue for its members' copyrights.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/guided-view-and-panel-view/","url":"https://readwyn.com/blog/guided-view-and-panel-view/","title":"Are Guided View and Panel View the same thing for comics?","summary":"Guided View and Kindle Panel View both zoom a comic one panel at a time. Where each came from, how they merged, and what a comic file must hold for them to work.","content_html":"<p>Partly. Both show a comic one panel at a time, enlarged to fill the screen, in the order the publisher chose, and on a Kindle today they run on the same data. They started out as rivals from two companies: Guided View belonged to comiXology, and Kindle Panel View was Amazon's version of the idea. Amazon has owned comiXology since 2014, and Amazon's publishing guidelines now treat Guided View as an animated layer on top of Panel View markup. Both depend on panel positions that someone marked and stored in the book file. Without them, a Kindle falls back to cutting each page into quarters.</p>\n<h2 id=\"guided-view-at-comixology\">Guided View at comiXology</h2>\n<p>Iconology Inc., the company behind comiXology, put its Comics app on Apple's App Store on <a rel=\"external\" href=\"https://gocollect.com/blog/comics-by-comixology-app-now-available-in-apple-s-app-store\">23 July 2009</a>, for the iPhone and iPod touch, with more than 100 titles from twenty publishers. Its pitch was a guided view that kept each page whole. Some competing apps cut a page into separate pictures that you swiped through as in a photo album.</p>\n<p>In November 2010 comiXology opened an invitation-only alpha of its <a rel=\"external\" href=\"https://www.bleedingcool.com/2010/11/16/comixology-to-give-guided-reading-to-all-publishers-and-creators/\">Guided View Authoring Tools</a>, so that publishers and independent creators could mark up the panels of their own books. Tokyopop and Devil's Due were among the first invited. The company's statement, as Bleeding Cool printed it, described reading “full screen or panel-to-panel”, and it called Guided View patent-pending.</p>\n<h2 id=\"panel-view-at-amazon\">Panel View at Amazon</h2>\n<p>Amazon announced Kindle Panel View as part of Kindle Format 8, alongside the colour Kindle Fire. When <a rel=\"external\" href=\"https://www.publishersweekly.com/pw/by-topic/booknews/comics/article/49417-how-many-digital-comics-formats-are-too-many.html\">Publishers Weekly</a> counted the competing comic formats in November 2011, it read Panel View as Amazon's take on Guided View and noted that Amazon held exclusive Kindle Fire rights to 100 DC graphic novels.</p>\n<p>Comics spread beyond the Fire in <a rel=\"external\" href=\"https://techcrunch.com/2012/06/14/amazon-updates-kindle-for-ios-android-and-cloud-reader-to-support-comics-childrens-books-with-improved-readability\">June 2012</a>, when the Kindle apps for iOS, Android and Cloud Reader gained them, with Panel View on titles that supported it. Creators could mark up their own panels with Kindle Comic Creator, a free tool that Amazon had released by <a rel=\"external\" href=\"https://dailycartoonist.com/?p=13961\">April 2013</a>. It took pages as JPEG, PDF, TIFF, PNG or PPM, suggested where the Panel View regions should go, and handled right-to-left page turns for manga. On E Ink, Nathan Groezinger's <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2013/11/03/kindle-paperwhite-2-comics-review-with-kindle-panel-view-video/\">review of the second Kindle Paperwhite</a> in November 2013 found Panel View working, though he complained that Amazon's store gave no way to tell which comics had it.</p>\n<h2 id=\"one-feature-on-the-kindle\">One feature on the Kindle</h2>\n<p>Amazon announced its purchase of comiXology on <a rel=\"external\" href=\"https://press.aboutamazon.com/news-releases/news-release-details/amazoncom-acquire-comixology\">10 April 2014</a>. The press release credited Guided View for the app's place among the ten top grossing iPad apps of 2011 and 2012. Guided View itself arrived in the Kindle app for iOS, Android and Fire tablets with a free update in <a rel=\"external\" href=\"https://smashpages.net/2017/05/09/amazon-adds-comixologys-guided-view-to-kindle\">May 2017</a>. In <a rel=\"external\" href=\"https://9to5google.com/2023/11/15/amazon-kill-comixology-app/\">November 2023</a> Amazon said that the comiXology app on iOS, Android and Fire OS would stop working on 4 December 2023, when it merged into the Kindle app, and that readers would find their comiXology books there.</p>\n<h2 id=\"what-the-file-holds\">What the file holds</h2>\n<p>A Kindle comic is a <a href=\"/blog/fixed-layout-vs-reflowable/\">fixed-layout</a> book in which each panel is a tap target. Amazon's guide to <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G9GSTY4LTRT39D4Z\">fixed-layout books with image pop-ups or Virtual Panels</a> gives the markup (November 2026): a <code>div</code> holding a link with the class <code>app-amzn-magnify</code>, whose <code>data-app-amzn-magnify</code> attribute carries a small JSON object. In it, <code>targetId</code> names the enlarged copy of the panel and <code>ordinal</code> gives its place in the reading order. A reader double-taps to enter Panel View. Amazon calls 150% the default and preferred zoom, which needs a page image of 2880 × 1800 pixels; 200% (3840 × 2400) is for small lettering, and 125% (2400 × 1500) only for a very large panel. Capital letters in a pop-up should be at least 2 mm high on a 7 inch screen.</p>\n<p>Guided View, in the same guide, adds an animated move from panel to panel with each swipe, panels centred at full screen size, and a custom masking colour around each one. Amazon applies it automatically to compatible comics, using the Panel View code that is already there.</p>\n<p>EPUB has its own standard for the same job. <a rel=\"external\" href=\"https://idpf.org/epub/renditions/region-nav/\">EPUB Region-Based Navigation 1.0</a>, which the IDPF made a recommended specification on 26 August 2015, puts the panels in a separate navigation document as a <code>nav</code> element with <code>epub:type=\"region-based\"</code>. Each panel is a link to a region of the page, a rectangle given in percentages of the page, or a circle or polygon in pixels or percentages, and a panel can hold nested regions for its speech balloons. A reading system that supports it must step forwards and backwards through the panels in list order, showing each one as large as the screen allows.</p>\n<p>A Kindle comic with no publisher panels at all gets Kindle Virtual Panels instead, Amazon's guide says: the Kindle cuts every page into four, in an order taken from the <code>primary-writing-mode</code> setting, which right-to-left manga must give a right-to-left value (<code>horizontal-rl</code>, or <code>vertical-rl</code> for Chinese, Japanese and Korean books). If even one page carries publisher panels, Virtual Panels are switched off for the whole book.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/history-of-epub/","url":"https://readwyn.com/blog/history-of-epub/","title":"Was EPUB created in 2007, and who looks after it now?","summary":"EPUB got its name in 2007, but it grew out of Open eBook from 1999. The W3C has maintained it since the IDPF combined with it in 2017.","content_html":"<p>Partly. The name EPUB dates from October 2007, but the format under it is eight years older: it grew out of the Open eBook Publication Structure (OEBPS), approved in 1999. The body behind both, the Open eBook Forum (renamed the International Digital Publishing Forum, or IDPF, in 2005), combined with the W3C in January 2017. The W3C has looked after EPUB since then, and its current version, EPUB 3.3, became a W3C Recommendation in 2023.</p>\n<h2 id=\"the-2007-date\">The 2007 date</h2>\n<p>The <a rel=\"external\" href=\"https://www.w3.org/publishing/epub32/epub-overview.html#sec-revision-history\">revision history</a> in the EPUB 3.2 overview puts OEBPS 1.0 in 1999, approved by the Open eBook Forum, with revisions in 2001 and 2002. According to the Library of Congress, <a rel=\"external\" href=\"https://www.loc.gov/preservation/digital/formats/fdd/fdd000171.shtml\">version 1.0 was written</a> during 1999 by an authoring group of more than 25 people, and the forum behind it was <a rel=\"external\" href=\"https://www.loc.gov/preservation/digital/formats/fdd/fdd000278.shtml\">incorporated in January 2000</a>. The forum took the name IDPF in mid 2005. OEBPS was an interchange format, and it had no single-file form for delivering a book to a reader.</p>\n<p>That single file came next. Work on a container started in late 2005, and the IDPF approved it as the OEBPS Container Format (OCF) in 2006. A 2.0 revision of OEBPS was written alongside it and approved in October 2007 under a new name, EPUB 2.0, as the <a rel=\"external\" href=\"https://www.w3.org/publishing/epub32/epub-overview.html#sec-revision-history\">same revision history</a> records: three specifications, one for the package, one for the content, and OCF for the container. EPUB 2.0.1, a round of corrections, followed in September 2010.</p>\n<p>The old name hasn't gone away. The media type of an EPUB's package file is <code>application/oebps-package+xml</code>, <a rel=\"external\" href=\"https://www.rfc-editor.org/rfc/rfc4839.txt\">registered in April 2007</a> for OEBPS package files and still defined in <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/\">EPUB 3.3</a>. Readwyn still opens a zipped OEB 1.x package today; <a href=\"/blog/oeb-package/\">a separate post</a> explains what one is and how to open it.</p>\n<h2 id=\"the-idpf\">The IDPF</h2>\n<p>The IDPF wrote EPUB 2 and EPUB 3. Its own site says the <a rel=\"external\" href=\"https://idpf.org/node/3449\">combination with the W3C was completed</a> on 30 January 2017, that the IDPF ceased operating as a membership organisation, and that idpf.org is now an archive. The W3C announced the merger on 1 February 2017 in a <a rel=\"external\" href=\"https://www.w3.org/press-releases/2017/idpf-w3c-combination\">press release</a> that also set up an EPUB 3.1 Community Group, free for anyone to join.</p>\n<h2 id=\"epub-3-since-2011\">EPUB 3 since 2011</h2>\n<p>EPUB 3.0 <a rel=\"external\" href=\"https://idpf.org/epub/30\">took effect on 11 October 2011</a>. It <a rel=\"external\" href=\"https://www.w3.org/publishing/epub32/epub-overview.html#sec-revision-history\">brought EPUB closer to HTML</a> and added <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-media-overlays\">Media Overlays</a>, markup that syncs the text with recorded audio. EPUB 3 has been revised several times since:</p>\n<ul>\n<li>EPUB 3.0.1 (2014) was mostly fixes, plus fixed-layout books. ISO <a rel=\"external\" href=\"https://idpf.org/epub/30\">republished the 3.0 specifications</a> word for word that year as ISO/IEC TS 30135, parts 1 to 7.</li>\n<li>EPUB 3.1 (2017) removed features such as content switching, triggers and bindings, and changed the version number in the package file.</li>\n<li>EPUB 3.2 (begun in 2018, <a rel=\"external\" href=\"https://www.w3.org/publishing/epub32/epub-spec.html\">final in May 2019</a>), from the EPUB 3 Community Group, put those features back as deprecated. Its <a rel=\"external\" href=\"https://www.w3.org/publishing/epub32/epub-overview.html#sec-revision-history\">history section</a> says that making everyone produce and read two versions of EPUB content would have cost more than 3.1 gained.</li>\n</ul>\n<p>By September 2014, <a rel=\"external\" href=\"https://www.loc.gov/preservation/digital/formats/fdd/fdd000278.shtml\">the Library of Congress says</a>, EPUB 2 was considered obsolete, though <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/\">EPUB 3.3</a> still keeps a few EPUB 2 features, labelled legacy, so that a book can work to some degree in a reading system that only knows EPUB 2 (<a href=\"/blog/epub-2-or-epub-3/\">what changes between the two inside a book</a>).</p>\n<p>A book made to EPUB 3.3 still says <code>version=\"3.0\"</code> in its package file: <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#attrdef-package-version\">the specification</a> requires that value and treats each 3.x update as a continuation of the same EPUB 3 format.</p>\n<h2 id=\"who-maintains-it-now\">Who maintains it now</h2>\n<p>EPUB 3.3 became a <a rel=\"external\" href=\"https://www.w3.org/standards/history/epub-33/\">W3C Recommendation on 25 May 2023</a> and has been republished as a Recommendation four times since, most recently on 13 January 2026. The Publishing Maintenance Working Group publishes it.</p>\n<p>The same group is working on <a rel=\"external\" href=\"https://www.w3.org/TR/epub-34/\">EPUB 3.4</a>, edited by Matt Garrish of the DAISY Consortium and Ivan Herman of the W3C. Its <a rel=\"external\" href=\"https://www.w3.org/standards/history/epub-34/\">publication history</a> starts with a First Public Working Draft on 27 March 2025 and reaches a Candidate Recommendation Snapshot on 21 July 2026. Since then it has come out again and again as a Candidate Recommendation Draft, six times in October 2026 alone, so the date at the top of the page moves from week to week. Comments on the draft go to the <code>w3c/epub-specs</code> repository on GitHub.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/history-of-kobo/","url":"https://readwyn.com/blog/history-of-kobo/","title":"Where Kobo came from, and who owns it","summary":"Kobo is owned by Rakuten of Japan, which agreed in 2011 to pay US$315 million for it, two years after the Canadian bookseller Indigo spun it off.","content_html":"<p>Kobo is owned by Rakuten, the Japanese e-commerce group. Rakuten <a rel=\"external\" href=\"https://global.rakuten.com/corp/news/press/2011/1109_01.html\">agreed in November 2011 to buy all of Kobo's shares</a> for US$315 million in cash, and the company trades as Rakuten Kobo Inc., which Rakuten still listed in November 2026 among its <a rel=\"external\" href=\"https://global.rakuten.com/corp/about/group.html\">major consolidated subsidiaries</a>.</p>\n<p>Kobo started in Canada as Shortcovers, the ebook division of Indigo Books &amp; Music, which Rakuten's <a rel=\"external\" href=\"https://global.rakuten.com/corp/news/press/2011/1109_01.html\">2011 announcement</a> calls the largest book, gift, and specialty toy retailer in Canada. In December 2009 Indigo renamed the service and spun it out as a company of its own. The new name was a coined word of four letters, an anagram of \"book\", as the <a rel=\"external\" href=\"https://www.csmonitor.com/Books/chapter-and-verse/2009/1217/Kobo-the-new-kid-on-the-e-reader-block\">Christian Science Monitor noted that month</a>. At first Kobo was a shop for ebooks, with no device of its own, and the Borders chain had already agreed to bring Kobo into its American shops.</p>\n<p>The hardware came a few months later. On 24 March 2010, <a rel=\"external\" href=\"https://www.cbc.ca/news/science/indigo-targets-amazon-with-kobo-e-reader-1.880917\">CBC reported</a> that Kobo had unveiled a $149 e-ink reader, due in Indigo stores in May and in American Borders shops that summer. It had no mobile data link like the Kindle's, so books went on over USB or Bluetooth, and it came with 100 public-domain classics already on it. At the time Indigo held 58 per cent of Kobo, and Borders, Australia's Red Group, and Cheung Kong Holdings of Hong Kong held stakes too. Kobo sold its books as <a href=\"/blog/history-of-epub/\">EPUB</a> files with DRM and offered its \"Powered by Kobo\" software to other e-reader makers. Its chief executive, Michael Serbinis, didn't expect the hardware itself to pay: \"Making an e-reader is going to get commoditized very fast,\" he told CBC.</p>\n<p>Rakuten announced the purchase on 9 November 2011. By then Kobo claimed a catalogue of more than 2.5 million titles, readers in over 100 countries, and the Kobo Touch and the new Kobo Vox on sale at Walmart, Best Buy, WHSmith, and FNAC as well as Indigo. The deal needed approval under the Investment Canada Act, and Rakuten promised that Kobo's headquarters, managers, and staff would stay in Toronto. It <a rel=\"external\" href=\"https://global.rakuten.com/corp/news/press/2012/0112_01.html\">closed on 12 January 2012</a>, once the Canadian regulators had approved it.</p>\n<p>For Rakuten the purchase was its way into ebooks. Its own <a rel=\"external\" href=\"https://global.rakuten.com/corp/about/history.html\">company history</a> files 2012 as the year it entered the business, by buying Kobo and opening a kobo ebook service in Japan.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-e-ink-works/","url":"https://readwyn.com/blog/how-e-ink-works/","title":"How does an E Ink screen work?","summary":"How an E Ink screen makes text from charged black-and-white pigment in tiny capsules, why it needs no power to hold a page, and who invented it.","content_html":"<p>An E Ink screen is a thin film holding millions of capsules, each about as wide as a human hair, filled with clear fluid and with white and black pigment particles that carry opposite electric charges. A field applied under a spot of the screen pulls one colour to the top of the capsules there and sends the other down, and the particles stay where they were put after the field is gone. E Ink calls the screen <a rel=\"external\" href=\"https://www.eink.com/tech/detail/How_it_works\">bistable</a>: the page stays visible, and the ink <a rel=\"external\" href=\"https://blog.eink.com/e-ink-wins-2025-retailtech-breakthrough-award\">draws no power</a>, until something changes it.</p>\n<h2 id=\"particles-where-an-lcd-has-molecules\">Particles where an LCD has molecules</h2>\n<p>An E Ink screen moves solid particles around, where a liquid-crystal display works on molecules. The 1998 paper in Nature that introduced the ink gave the particles of such displays a size of 0.1 to 5 micrometres and contrasted them with the \"molecular-scale properties\" that govern liquid crystals (<a rel=\"external\" href=\"https://www.nature.com/articles/28349\">Comiskey, Albert, Yoshizawa and Jacobson, Nature 394</a>). In the version that <a rel=\"external\" href=\"https://www.sciencefriday.com/articles/how-electronic-ink-was-invented/\">Science Friday described</a>, the white particles are titanium dioxide and the black ones carbon black, suspended in oil.</p>\n<p>Which colour carries which charge is less settled than you'd think. <a rel=\"external\" href=\"https://www.eink.com/tech/detail/How_it_works\">E Ink's own page</a> gives the white particles a negative charge when it describes its Carta ink, and a positive one in the FAQ further down. What matters is that the two charges are opposite, so one field moves the colours in opposite directions.</p>\n<p>Because you're looking at pigment, the screen is lit by the room, as paper is. Joseph Jacobson, who set the project going, wanted a screen that gave off no light and could be read at a tilt, according to Science Friday. The glow of a modern e-reader comes from a separate <a href=\"/blog/front-light-and-warm-light/\">front light</a>.</p>\n<h2 id=\"when-the-screen-uses-power\">When the screen uses power</h2>\n<p>An E Ink screen takes no power while a page sits still. The Nature paper called particle displays intrinsically bistable: black and white are both stable states, with no voltage needed to hold either. An <a rel=\"external\" href=\"https://blog.eink.com/e-ink-wins-2025-retailtech-breakthrough-award\">E Ink blog post on its 2025 RetailTech award</a> says the ink consumes power only while the image is changing, and that the shelf labels made with it run for years without new batteries.</p>\n<p>So an e-reader's battery pays for page turns and for the parts around the screen, such as the light, rather than for the page in front of you. Amazon's <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/new-kindle-color-scribe-paperwhite-entry\">announcement of its new Kindles on 16 October 2024</a> promised up to three months on one charge for the Paperwhite and up to eight weeks for the colour Colorsoft.</p>\n<h2 id=\"where-the-pixels-are\">Where the pixels are</h2>\n<p>E Ink <a rel=\"external\" href=\"https://www.eink.com/tech/detail/How_it_works\">laminates the ink to a plastic film and sticks the film to electronics</a>. A capsule isn't a pixel: the pixel grid belongs to those electronics, the backplane, which <a rel=\"external\" href=\"https://www.technologyreview.com/2002/01/22/235273/flexible-displays-gain-momentum/\">MIT Technology Review described in 2002</a> as the sheet that \"controls display pixels\".</p>\n<h2 id=\"how-fast-it-changes\">How fast it changes</h2>\n<p>E Ink's <a rel=\"external\" href=\"https://www.eink.com/tech/detail/How_it_works\">FAQ</a> says a full-screen update can take as little as 120 milliseconds, and that only the regions that change need updating. Its faster modes can reach animation, but constant updates cost power, and the company says its production modules don't support video rates. The quick partial updates also leave faint traces behind, which is <a href=\"/blog/why-e-ink-flashes/\">why an e-reader flashes black</a> every few pages.</p>\n<h2 id=\"who-invented-it\">Who invented it</h2>\n<p>The ink is a decade older than the Kindle. Jacobson, a professor at the MIT Media Lab, recruited two undergraduates, Barrett Comiskey and JD Albert. Science Friday reports that in a windowless basement lab, at two or three in the morning on 23 January 1997, they moved a particle inside a capsule with an outside field, after experts had told them two oppositely charged pigments couldn't share one capsule. E Ink Corporation was spun out of the Media Lab that year (<a rel=\"external\" href=\"https://blog.eink.com/whats-in-a-name-the-history-behind-e-ink\">E Ink, \"What's in a name\"</a>).</p>\n<p>The capsule was what made the July 1998 Nature paper matter. Earlier particle displays had short lives and were hard to manufacture; sealing the fluid into microcapsules solved the lifetime problem and meant a display could be made by printing alone. The first e-reader with the ink was Sony's Librié, <a rel=\"external\" href=\"https://phys.org/news/2004-04-first-generation-electronic-paper-philips-sony.html\">due on sale in Japan in late April 2004</a>, and <a href=\"/blog/first-kindle-2007/\">the Kindle followed in 2007</a>.</p>\n<p>The company changed hands in 2009. Prime View International, founded in 1992 as Taiwan's first maker of TFT LCDs, bought E Ink Corporation and took its name for the combined company. \"E Ink\" is a registered trademark, and the company calls the technology itself <a rel=\"external\" href=\"https://blog.eink.com/whats-in-a-name-the-history-behind-e-ink\">electrophoretic ink</a>.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-e-reader-dictionaries-work/","url":"https://readwyn.com/blog/how-e-reader-dictionaries-work/","title":"How the dictionary on a Kindle or Kobo works, and how to change it","summary":"Where to switch dictionaries on a Kindle or a Kobo, where the built-in ones come from, and how a dictionary file of your own gets onto the device.","content_html":"<p>Both a Kindle and a Kobo let you change the dictionary, and both will take a dictionary file you add yourself, though only Amazon documents how. A Kindle keeps one default dictionary for each language. A Kobo installs dictionaries one language at a time, and you switch to a translation from inside the definition box.</p>\n<p>On a Kindle the route starts from the Quick Actions menu, as Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_all_new_paperwhite_v2/v1/en-US/html/kug.html\">guide for the 10th-generation Kindle Paperwhite</a> gives it: All Settings, Language &amp; Dictionaries, then Dictionaries. Each language shows the dictionary it uses now, and an arrow beside it means there's more than one to pick from: tap the language, choose with the radio buttons, then tap OK. The same guide adds a shortcut: tap the dictionary's name on the definition card during a lookup. The dictionaries are books in your library, kept in a Dictionaries collection on the Home screen or in the cloud once the Kindle is registered. The first Kindle, <a rel=\"external\" href=\"https://press.aboutamazon.com/2007/11/introducing-amazon-kindle\">announced on 19 November 2007</a>, came with The New Oxford American Dictionary, and in 2014 Oxford University Press <a rel=\"external\" href=\"https://oxford.co.za/oxford-dictionaries-the-trusted-choice-for-the-worlds-largest-search-engines-apps-and-e-readers/\">said</a> Kindles were preloaded with the Oxford Dictionary of English as well.</p>\n<p>A Kobo keeps its list behind the More icon at the bottom of the Home screen: Settings, Language and dictionaries, then Edit under Dictionary, according to Kobo's page on <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017640093\">adding dictionaries</a>. A plus icon installs a dictionary, a check mark removes one, and Save keeps the change. To <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017639873\">translate a word</a>, hold it, lift your finger, tap the dropdown icon in the box and pick a language. Kobo's dictionaries draw on Oxford data as well: OUP's <a rel=\"external\" href=\"https://languages.oup.com/about-us/case-studies/kobo\">case study</a> names Spanish, Portuguese, German, Swedish, Turkish, Italian and English among the monolingual and bilingual sets it gave Kobo, the ones you'd use for <a href=\"/blog/read-in-a-foreign-language-on-an-e-reader/\">reading in a language you're learning</a>. One limit applies to all of them: Kobo <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017639893-Define-words-with-the-dictionary-on-your-Kobo-eReader\">warns</a> that lookup may not work in PDFs, in library loans, or in free books found online.</p>\n<h2 id=\"adding-a-dictionary-of-your-own\">Adding a dictionary of your own</h2>\n<p>Amazon's page on <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G2HXJS944GL88DNV\">creating dictionaries</a> treats a Kindle dictionary as an ebook with extra tags. Its metadata names the language of the books it serves and the language of its definitions, so an English dictionary names English twice, and Spanish to French and French to Spanish have to be two separate books. Every headword needs a full list of its inflected forms in a hidden index, so that a plural or a past tense opens the same entry. Kindles match accents loosely unless the dictionary asks for an exact match, which Amazon advises for languages where an accent marks a different word; its example is the Spanish \"uña\". To test one, Amazon says to export it from Kindle Previewer as a MOBI file, copy it over USB into the Kindle's Documents folder, and choose it under Dictionaries for its source language. Only the e-reader can make it the default for lookup; Previewer can't.</p>\n<p>The Kobo has done the same since 2020, though Kobo's help pages on dictionaries still say nothing about it. Firmware 4.24, released in September 2020, added support for custom dictionaries in a hidden <code>.kobo/custom-dict</code> folder, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2020/09/25/new-kobo-firmware-update-4-24-adds-new-dictionaries\">The eBook Reader reported</a>, which described the support as beta. The file is a zip named <code>dicthtml-xx.zip</code>, or <code>dicthtml-xx-yy.zip</code> for a translation, where the letters are language codes, according to <a rel=\"external\" href=\"https://pgaskin.net/dictutil/dicthtml/format.html\">Patrick Gaskin's format notes</a>. MobileRead users found that the Kobo has to <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=4041829\">restart</a> before it sees a new file, and that the dictionary then <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=4042311\">appears in the dictionary drop-down</a> with \"(Custom)\" after its name.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-kindle-unlimited-pays-authors/","url":"https://readwyn.com/blog/how-kindle-unlimited-pays-authors/","title":"How does Kindle Unlimited (now Kindle+) work, and how are authors paid?","summary":"Kindle Unlimited, now Kindle+, pays authors a share of a monthly fund by pages read for the first time. What counts as a page, and what exclusivity means.","content_html":"<p>By the page, and only the first time a subscriber reads that page. Readers pay Amazon a flat monthly fee, and authors split one shared pot, the KDP Select Global Fund, in proportion to the pages of their books that subscribers read. A borrow that nobody opens earns the author nothing.</p>\n<p>Kindle Unlimited became Kindle+ in October 2026; Amazon's own guide was <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/what-is-kindle-plus\">updated on 22 October 2026</a> with the new name. That page gives the US price as $11.99 a month plus tax, says a subscriber can hold up to 20 titles at once with no due dates, and says it isn't part of Prime (Prime members get a selection of the catalogue). KDP's help pages use the new name too, and an ebook enrolled in KDP Select now <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GD9PMU58BV24QFZ7\">takes its virtual-voice audiobook</a> into Kindle+ with it.</p>\n<h2 id=\"per-borrow-or-per-page\">Per borrow or per page</h2>\n<p>Until July 2015, Amazon paid authors a share of the fund whenever a Kindle Unlimited reader got past 10% of the book, as its <a rel=\"external\" href=\"https://web.archive.org/web/20150228031528/https://kdp.amazon.com/select\">KDP Select page</a> said in early 2015. From 1 July 2015, Amazon switched to <a rel=\"external\" href=\"https://www.engadget.com/2015/06/21/amazon-pays-kindle-authors-by-page-reads/\">paying by pages read</a>, so a 600-page novel read to the end now earns far more than a 30-page booklet opened once.</p>\n<h2 id=\"what-counts-as-a-page\">What counts as a page</h2>\n<p>The page isn't the one a reader sees on their Kindle, and it isn't the printed edition's page either. Amazon counts in <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G5HHMLVVEW48MQYW\">Kindle Edition Normalized Pages</a> (KENP), worked out with a standard font, line height and spacing, so a reader who picks a huge font doesn't turn one page into three. Counting runs from the Start Reading Location, which Amazon usually puts at chapter 1, to the end of the book, and images and charts count too. One subscriber can earn an author at most 3,000 KENP for a title.</p>\n<p>Each page pays once per subscriber. Going back to reread earns nothing, and since audiobooks joined the catalogue, Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G201541130\">royalty rules</a> treat listening to pages you've already read as rereading too. If someone reads and listens at the same time, the page is counted once, at whichever format's rate is higher.</p>\n<h2 id=\"the-rate-per-page\">The rate per page</h2>\n<p>There's no set price per page. The <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G5HHMLVVEW48MQYW\">KENP rate</a> for a marketplace is that month's fund for the marketplace divided by all the KENP read there, so it moves whenever the fund or the total reading changes. <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G201541130\">Amazon reviews the fund each month</a> and announces its size in the KDP community forum; the month's page counts are finalised near the 15th of the following month. Each country's slice of the fund depends on exchange rates, reading habits and local subscription prices. On 1 November 2026, the footer of the <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200798990\">KDP help pages</a> put total KDP Select author earnings for September 2026 at $69.3 million.</p>\n<p>On top of the pages, Amazon pays a discretionary <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G201623400\">KDP Select bonus</a> to some books each month, judged on customer signals beyond pages read and paid about 60 days after the month ends. Authors based in France are an odd case: they must set a price per page for each enrolled ebook, though <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G201541130\">Amazon says</a> the figure they enter doesn't change what they're paid.</p>\n<h2 id=\"what-exclusivity-covers\">What exclusivity covers</h2>\n<p>The <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GD9PMU58BV24QFZ7\">KDP Select terms</a> tie the ebook to Amazon for each 90-day term, along with its virtual-voice audiobook, if it has one. During the term the ebook may be distributed only through KDP and public libraries. The paperback, the hardback and other formats can be sold anywhere, and enrolling an ACX audiobook adds no exclusivity to it. Terms renew automatically unless the author turns renewal off, an author can cancel within three days of a term starting, and public-domain books can't be enrolled at all.</p>\n<p>Kobo Plus <a rel=\"external\" href=\"https://kobowritinglife.zendesk.com/hc/en-us/articles/360058975432-What-is-Kobo-Plus\">asks for no exclusivity</a>, and it pays for every minute read, rereads included, even when a reader gets through less than 10% of a book. Kobo <a rel=\"external\" href=\"https://kobowritinglife.zendesk.com/hc/en-us/articles/360059386091-How-does-Kobo-Plus-pay\">divides each month's subscription revenue</a> by the total minutes read to value a minute, then pays the author 60% of that value for each minute spent in their book.</p>\n<h2 id=\"when-the-author-leaves-or-the-reader-cancels\">When the author leaves, or the reader cancels</h2>\n<p>A borrowed book doesn't disappear when its author leaves KDP Select. <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G201541130\">Amazon says</a> a subscriber who has already downloaded the book can keep reading it as often as they like after the enrolment has expired, and the author still earns a share of the fund for those pages. New subscribers just can't borrow it any more.</p>\n<p>If the reader cancels, <a rel=\"external\" href=\"https://www.aboutamazon.com/news/devices/what-is-kindle-plus\">Amazon's guide</a> says the subscription stays active until the next billing date, after which the reader loses access to every Kindle+ title they hold, with no refund of fees already paid. Books bought outright aren't borrows, and they follow other rules, set out in <a href=\"/blog/do-you-own-your-kindle-books/\">do you own your Kindle books</a>.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-many-books-does-the-average-person-read/","url":"https://readwyn.com/blog/how-many-books-does-the-average-person-read/","title":"How many books does the average person read in a year?","summary":"Pew and Gallup put the mean at 12 to 14 books a year in 2021, and YouGov put it at eight in December 2025; the median is two to five. A few heavy readers pull the mean up.","content_html":"<p>In the last Pew and Gallup polls to give a mean, both from 2021, the average American adult read about 12 to 14 books a year. A YouGov poll of December 2025 put the mean at <a rel=\"external\" href=\"https://yougov.com/en-us/articles/53804-most-americans-didnt-read-many-books-in-2025\">eight</a> and the median at two. The typical adult reads far fewer: in the <a rel=\"external\" href=\"https://www.pewresearch.org/wp-content/uploads/sites/20/2026/04/S_26.04.09_books_topline.pdf\">most recent survey by the Pew Research Center</a>, just over half of adults read three books or fewer in the past year, and a quarter read none at all.</p>\n<h2 id=\"two-polls-two-counts\">Two polls, two counts</h2>\n<p>Both figures come from American polls. Pew first asked about books in 2011. In its survey of 1,502 adults in January and February 2021, Americans <a rel=\"external\" href=\"https://www.pewresearch.org/short-reads/2022/01/06/three-in-ten-americans-now-read-e-books/\">read a mean of roughly 14 books and a median of five</a>, and Pew noted that both figures were identical to 2011.</p>\n<p>Gallup has asked since 1990. Its December 2021 poll of 811 adults found an <a rel=\"external\" href=\"https://news.gallup.com/poll/388541/americans-reading-fewer-books-past.aspx\">average of 12.6 books</a>, the lowest it had ever recorded. The same question gave 15.6 in 2016 and 18.5 in 1999. Gallup put the margin of error on its mean at 1.9 books either way, so Pew's 14 from the same year falls inside the range of Gallup's own estimate.</p>\n<p>A book counts in both polls as soon as you start it. <a rel=\"external\" href=\"https://news.gallup.com/poll/388541/americans-reading-fewer-books-past.aspx\">Gallup's question</a> counts books read \"all or part of the way through\", and its interviewers are told to include ebooks and audiobooks. Pew's <a rel=\"external\" href=\"https://www.pewresearch.org/wp-content/uploads/sites/20/2026/04/S_26.04.09_books_topline.pdf\">questionnaire</a> uses nearly the same words and asks people to include print, electronic, and audiobooks. A novel dropped on page 40 counts as a book, and so does a book <a href=\"/blog/audiobooks-vs-reading/\">heard rather than read</a>. YouGov's <a rel=\"external\" href=\"https://yougov.com/en-us/articles/53804-most-americans-didnt-read-many-books-in-2025\">article on its 2025 poll</a> does not say whether a part-read book counted, although audiobooks did: 23% of Americans had listened to at least one that year.</p>\n<p>If you keep a count of your own, a list of finished books is the stricter measure: the Books finished figure on Readwyn's Stats screen, for example, counts only the books you finished in the range you choose.</p>\n<h2 id=\"why-the-mean-and-the-median-are-so-far-apart\">Why the mean and the median are so far apart</h2>\n<p>A mean adds up every book and divides by every person. One adult who reads 100 books a year lifts the mean of a group of 100 people by a whole book, while the median (the person in the middle of the line) doesn't move. Book reading is lopsided like that.</p>\n<p>Pew's October 2025 survey of 8,046 adults shows the spread. According to the <a rel=\"external\" href=\"https://www.pewresearch.org/wp-content/uploads/sites/20/2026/04/S_26.04.09_books_topline.pdf\">topline</a>, 25% read no books in the past 12 months, 8% read one, 19% read two or three, and 10% read four or five. Another 13% read six to 10, 10% read 11 to 20, and 14% read more than 20. Count up from the bottom and the 50th percentile lands in the band of two or three books. That top 14% does much of the work in any average.</p>\n<p>Pew published that survey in bands and gave no mean. Its 2025 median also looks lower than the five of 2021, but Pew moved from telephone interviews to a mix of web and phone in 2025 and marks the change of method in its charts. Its own <a rel=\"external\" href=\"https://www.pewresearch.org/short-reads/2026/04/09/americans-still-opt-for-print-books-over-digital-or-audio-versions-few-are-in-book-clubs/\">report on the survey</a> says the share of Americans who read books has barely changed since 2021.</p>\n<p>Gallup's drop in 2021 came from the heavy readers too. The share who had read more than 10 books fell from 35% in 2016 to 27%, while the share who read none, 17%, stayed close to the 16% to 18% of the surveys from 2002 to 2016. College graduates fell furthest: they averaged 21.1 books across the polls of 2002 to 2016 and <a rel=\"external\" href=\"https://news.gallup.com/poll/388541/americans-reading-fewer-books-past.aspx\">14.6 in 2021</a>.</p>\n<h2 id=\"who-reads-more\">Who reads more</h2>\n<p>Education makes a large difference in <a rel=\"external\" href=\"https://www.pewresearch.org/short-reads/2026/04/09/americans-still-opt-for-print-books-over-digital-or-audio-versions-few-are-in-book-clubs/\">Pew's 2025 data</a>: 88% of adults with a bachelor's degree had read at least part of a book in the past year, against 60% of those with a high school education or less. Women were more likely to have read one than men (78% and 71%). <a rel=\"external\" href=\"https://news.gallup.com/poll/388541/americans-reading-fewer-books-past.aspx\">Gallup found</a> the same gap in volume, with women averaging 15.7 books in 2021 and men 9.5.</p>\n<h2 id=\"how-many-read-ebooks\">How many read ebooks</h2>\n<p>Pew's questions about format give the share of people who read one, rather than a count of books. In October 2025, 64% of adults had read a print book in the past year, down from 72% in 2011. The share who had read an ebook rose from 17% to 31% over the same years, and 26% had listened to an audiobook. Young adults lean digital: <a rel=\"external\" href=\"https://www.pewresearch.org/short-reads/2026/04/09/americans-still-opt-for-print-books-over-digital-or-audio-versions-few-are-in-book-clubs/\">41% of adults under 30</a> had read an ebook, compared with 23% of those aged 65 and over.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-many-books-fit-on-an-e-reader/","url":"https://readwyn.com/blog/how-many-books-fit-on-an-e-reader/","title":"How many books fit on an e-reader?","summary":"A 16 GB Kindle or Kobo holds roughly 10,000 plain novels but only hundreds of illustrated books, going by the makers' sums and real file sizes.","content_html":"<p>Kobo says the 16 GB in its 6-inch Clara BW lets you carry “up to 12,000 eBooks or 75 Kobo Audiobooks”, and a footnote on <a rel=\"external\" href=\"https://ereader.kobo.com/en-ww/products/kobo-clara-bw\">the product page</a> shows the working: a gigabyte counted as a billion bytes, and about 1.3 MB per book. That's a fair figure for novels. A 16 GB Kindle or Kobo has room for something like ten thousand plain novels, more if they're short ones, and the count drops to the hundreds once the books are full of pictures.</p>\n<h2 id=\"the-makers-sums\">The makers' sums</h2>\n<p>Sixteen billion bytes divided by 1.3 MB comes to about 12,300, which Kobo rounds down. The same rate runs through its range: in November 2026 the 32 GB <a rel=\"external\" href=\"https://ereader.kobo.com/en-ww/products/kobo-libra-colour\">Libra Colour</a> is sold for up to 24,000 ebooks or 150 audiobooks. For audiobooks the footnote assumes 6.5 hours a title at 28.8 MB an hour. That's about 187 MB a book, a bitrate of 64 kilobits a second, and as much space as some 144 ebooks.</p>\n<p>Amazon doesn't print its arithmetic. When it announced its new Kindle line on 16 October 2024, it gave the basic Kindle and the Paperwhite <a rel=\"external\" href=\"https://press.aboutamazon.com/2024/10/amazon-launches-entirely-new-kindle-lineup-including-reimagined-kindle-scribe-and-first-ever-color-kindle\">16 GB each</a> and said only that this held thousands of books. Its early press releases were more precise. The first Kindle of November 2007 <a rel=\"external\" href=\"https://press.aboutamazon.com/2007/11/introducing-amazon-kindle\">stored more than 200 titles</a>, and the Kindle 2 of February 2009 had <a rel=\"external\" href=\"https://press.aboutamazon.com/2009/2/introducing-amazon-kindle-2\">2 GB for more than 1,500 books</a>. At the launch, TidBITS put the space for books at <a rel=\"external\" href=\"https://tidbits.com/2009/02/09/amazon-announces-kindle-2-ebook-reader/\">1.4 GB on the Kindle 2</a>, against about 180 MB on the first model. Both work out to roughly 0.9 MB a book, a little under the size Kobo assumes now.</p>\n<h2 id=\"what-real-ebooks-weigh\">What real ebooks weigh</h2>\n<p>Project Gutenberg prints the size of every download, and as of November 2026 a novel of ordinary length comes in well under Kobo's 1.3 MB. The EPUB 3 of <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/84\">Frankenstein</a> is 430 kB. <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2600\">War and Peace</a> goes over it, at 1.7 MB. At those sizes a 16 GB reader would take about 37,000 copies of Shelley's novel or 9,400 of Tolstoy's. The Kindle (KF8) files of the same two books are larger, at 535 kB and 2.4 MB.</p>\n<p>Text squeezes down well. War and Peace is 3.2 MB as plain text but 1.7 MB as an EPUB, because an EPUB is a ZIP archive and the standard lets the files inside it be <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-zip-container-zipreqs\">compressed with Deflate</a>.</p>\n<p>Pictures change the sums. Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1342\">Pride and Prejudice</a> is 595 kB without its images and 23.7 MB with them, so the illustrated file would fit on a 16 GB reader only about 675 times. <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/14838\">The Tale of Peter Rabbit</a> is 70 kB as an EPUB without its pictures and 1.4 MB as the picture book, twenty times the size. A comic, with a picture on every page, behaves like the illustrated Austen and not like Tolstoy. <a href=\"/blog/why-ebook-files-are-large/\">Why some ebook files are so large</a> looks at where those bytes sit inside the file.</p>\n<h2 id=\"why-16-gb-shows-up-as-less\">Why 16 GB shows up as less</h2>\n<p>Connect a new e-reader to a computer and you may see less space than the box promised, before you've added a single book. Makers count a gigabyte as a billion bytes, but some software counts it as 1,073,741,824 bytes. <a rel=\"external\" href=\"https://support.apple.com/en-us/102119\">Apple's page on storage capacity</a> names iOS 10 and earlier, and OS X Leopard and earlier, as systems that reported space in binary units. Sixteen billion bytes comes to 14.9 of those. Apple adds that the operating system takes its own share, and both Kobo footnotes end on the same point (some storage is used by software) without saying how much.</p>\n<p>For planning, go by the free space the device reports (for a Kobo, Readwyn's Sync screen gives it on one line with the number of books already on the device). Divide that by 1.3 MB for a rough count of the novels still to come, or by about 24 MB if they're illustrated like Gutenberg's Austen.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-many-words-in-a-novel/","url":"https://readwyn.com/blog/how-many-words-in-a-novel/","title":"How many words are in a novel?","summary":"A novel is 40,000 words or more by award rules, Harlequin asks for 50,000 to 100,000, and classics run from 48,000 words to over half a million.","content_html":"<p>Project Gutenberg's text of <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/64317\">The Great Gatsby</a> comes to about 48,600 words, which clears the line for a novel by fewer than 9,000. That line is 40,000 words, and it's the one the Nebula and Hugo awards for science fiction and fantasy use. Above it, the range is wide. Harlequin asks for 50,000 words in its shortest romance lines and up to 100,000 in its longest, and Project Gutenberg's English translations of War and Peace and Les Misérables each pass 560,000.</p>\n<h2 id=\"where-the-award-rules-draw-the-lines\">Where the award rules draw the lines</h2>\n<p>The <a rel=\"external\" href=\"https://nebulas.sfwa.org/about-the-nebulas/nebula-rules/\">Nebula Awards rules</a> of the Science Fiction and Fantasy Writers Association set four lengths for the 2026 award year. A short story has fewer than 7,500 words, a novelette at least 7,500 and fewer than 17,500, a novella at least 17,500 and fewer than 40,000, and a novel 40,000 or more. A translated work is judged by its word count in English.</p>\n<p>The Hugo Awards use the same four boundaries. Section 3.3.1 of the <a rel=\"external\" href=\"https://www.wsfs.org/wp-content/uploads/2026/06/WSFS-Constitution-as-of-2025-10-07-Clean_20260606.pdf\">WSFS Constitution</a> defines Best Novel as a story of 40,000 words or more, and section 3.2.10 adds some give: the Worldcon committee may move a story into a more fitting category if its length is within 20 per cent of that category's limits. A novella of 33,000 words can therefore end up on the novel ballot. The same document defines Best Series by total length, at least 240,000 words over three or more instalments.</p>\n<h2 id=\"what-publishers-ask-for\">What publishers ask for</h2>\n<p>Harlequin publishes exact targets. In November 2026, its <a rel=\"external\" href=\"https://harlequin.submittable.com/submit\">submission page</a> gives these for its series:</p>\n<table><thead><tr><th>Series</th><th>Words</th></tr></thead><tbody>\n<tr><td>Harlequin Presents (Mills &amp; Boon Modern), Medical Romance, Harlequin Romance (Mills &amp; Boon Love Always)</td><td>50,000</td></tr>\n<tr><td>Love Inspired, Love Inspired Suspense</td><td>55,000</td></tr>\n<tr><td>Harlequin Intrigue</td><td>55,000 to 60,000</td></tr>\n<tr><td>Harlequin Heartwarming, Special Edition, Romantic Suspense</td><td>70,000</td></tr>\n<tr><td>Afterglow by Harlequin</td><td>at least 75,000</td></tr>\n<tr><td>Love Inspired Trade</td><td>85,000 to 100,000</td></tr>\n</tbody></table>\n<p>The table leaves out Harlequin Historical (70,000 to 75,000 words), because a note of 11 February 2026 on the same page closes it to submissions and says the series will end in September 2027. The page also says what the extra length is for. A 70,000-word Romantic Suspense novel has room for subplots, while the 50,000-word lines are asked to carry the book on emotional conflict.</p>\n<p>Outside publishing, the best-known target was the 50,000 words of National Novel Writing Month: a draft of that length in the thirty days of November, about 1,667 words a day, and about 1,400 words more than Gutenberg's Great Gatsby. According to <a rel=\"external\" href=\"https://www.theguardian.com/books/2025/apr/02/creative-writing-nanowrimo-to-close-after-20-years\">The Guardian</a>, the event began informally in 1999 and became a nonprofit in 2006. At the end of March 2025 it announced that it would close, citing money problems made worse by damage to its reputation.</p>\n<h2 id=\"word-counts-of-well-known-novels\">Word counts of well-known novels</h2>\n<p>The counts below come from the plain-text files that Project Gutenberg served on 2 November 2026. Each covers everything between Gutenberg's start and end lines, so the title, the contents list and the chapter headings are included. The text is split at spaces, at em dashes and en dashes, and at the double hyphens that some of the files use in place of a dash, and the total is rounded to the nearest hundred. If a file changes after a correction, a count made later differs by a few words. Different editions and translations give different totals.</p>\n<table><thead><tr><th>Book</th><th>Words</th></tr></thead><tbody>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/43\">Strange Case of Dr Jekyll and Mr Hyde</a></td><td>25,700</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/11\">Alice's Adventures in Wonderland</a></td><td>26,800</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/46\">A Christmas Carol</a></td><td>28,600</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/35\">The Time Machine</a></td><td>32,600</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/219\">Heart of Darkness</a></td><td>38,500</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/64317\">The Great Gatsby</a></td><td>48,600</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/120\">Treasure Island</a></td><td>69,100</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/84\">Frankenstein</a></td><td>75,200</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/768\">Wuthering Heights</a></td><td>116,600</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/42671\">Pride and Prejudice</a></td><td>122,500</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1260\">Jane Eyre</a></td><td>187,200</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2701\">Moby-Dick</a></td><td>214,400</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/145\">Middlemarch</a></td><td>318,600</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2600\">War and Peace</a>, translated by Louise and Aylmer Maude</td><td>565,400</td></tr>\n<tr><td><a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/135\">Les Misérables</a>, translated by Isabel Florence Hapgood</td><td>567,500</td></tr>\n</tbody></table>\n<p>By the award rules, the first five would be novellas. Heart of Darkness misses 40,000 by 1,500 words, close enough for the Hugo margin. In this pair of translations, Victor Hugo's novel comes out slightly longer than Tolstoy's, and a different translator could reverse that.</p>\n<h2 id=\"why-two-counts-of-one-book-differ\">Why two counts of one book differ</h2>\n<p>A word count depends on where the program decides a word ends. Splitting the same Gutenberg files only at spaces gives smaller totals for books with many unspaced dashes. Wuthering Heights opens with “1801—I have just returned”, and a count by spaces reads “1801—I” as one word. For Heart of Darkness the space count is 37,900 instead of 38,500, and for Jane Eyre it's 185,400 instead of 187,200, a gap of about 1,850 words.</p>\n<p>Readwyn splits at spaces only when it counts the words of an ebook, and it shows the total in the Details tab of the book screen, with \"Count again\" beside it. It divides that number by 250 for the <a href=\"/blog/ebook-page-count-from-word-count/\">estimated page count</a>, shown as \"~880\". The post on <a href=\"/blog/count-words-in-ebook/\">counting the words in an ebook</a> gives the rules it follows.</p>\n<p>Front matter adds noise of its own. Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1342\">illustrated edition of Pride and Prejudice</a> counts about 5,000 words more than the plain one: it opens with a preface by George Saintsbury and a list of its illustrations, and 76 of its 162 pictures have a caption.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ideal-line-length-for-reading/","url":"https://readwyn.com/blog/ideal-line-length-for-reading/","title":"What is the best line length for reading?","summary":"Typographers advise 45 to 75 characters per line, with 66 as the ideal, though screen studies found people read long lines faster even when they liked them less.","content_html":"<p>For a single column of continuous text, aim for 45 to 75 characters per line, counting the spaces. Robert Bringhurst, in <em>The Elements of Typographic Style</em>, writes that the 66-character line is widely regarded as ideal. Richard Rutter quotes those figures on <a rel=\"external\" href=\"http://webtypography.net/2.1.2\">webtypography.net</a>, a site that applies Bringhurst's rules to text on the web.</p>\n<p>The measure is counted in characters rather than centimetres because the same width holds a different number of letters at each type size. Rutter works it through: a box 400 pixels wide holds about 66 characters of 12 px text, but only about 50 at 16 px. Since an average character is roughly half an em wide, a column 33 em across stays near 66 characters whatever the size. Matthew Butterick's <a rel=\"external\" href=\"https://practicaltypography.com/line-length.html\"><em>Practical Typography</em></a> allows a slightly wider band, 45 to 90 characters, and has a test that needs no counting: a comfortable line holds two to three lowercase alphabets.</p>\n<p>Other guidance lands close by. A <a rel=\"external\" href=\"https://baymard.com/blog/line-length-readability\">2022 article by the Baymard Institute</a> cites the Swiss typographer Emil Ruder, who preferred 50 to 60 characters, and settles on 50 to 75 for body text. Baymard also reports that people in its usability testing grew tired once lines reached 100 characters or more. The W3C's accessibility guidelines set a ceiling instead of a target. <a rel=\"external\" href=\"https://www.w3.org/WAI/WCAG22/Understanding/visual-presentation.html\">Success criterion 1.4.8</a> says a reader must be able to get lines of no more than 80 characters, or 40 in Chinese, Japanese and Korean, whose characters are about twice as wide. It also asks for line spacing of at least one and a half.</p>\n<p>The reasons are mechanical. At the end of a long line the eye has a long way back to the start of the next, and it can drop onto the wrong one. A very short line forces that return trip so often that the reading turns jerky (<a rel=\"external\" href=\"https://baymard.com/blog/line-length-readability\">Baymard</a> gives both reasons), and if the text is justified, a narrow measure opens wide gaps between words (the trade-off is covered in <a href=\"/blog/justified-or-ragged-right/\">justified or ragged right</a>).</p>\n<p>On an e-reader you set the line length indirectly. A larger font puts fewer characters on each line, and so do wider margins, so the settings that make the text easier to see also shorten the measure. On a <a href=\"/blog/e-reader-screen-size/\">small screen</a> at a large size, lines can end up shorter than typographers like; on a tablet held sideways or a laptop, they can run far past 75. Splitting a wide area into columns shortens each line, and for work set in several columns Bringhurst gives a lower range, 40 to 50 characters.</p>\n<p>To check a book on your own device, pick an ordinary line in the middle of a paragraph and count every letter, space and punctuation mark. If the count falls well under 45, a smaller font or narrower margins will lengthen it; if it passes 75, the opposite applies. On a wide screen the fix can be the layout rather than the type. In a window at least 1,000 px wide, the Readwyn reader on a Mac shows two pages side by side (unless you've set it to always show one page), so a line runs across one page rather than the whole window.</p>\n<p>On screens, some experiments favoured long lines. Mary Dyson's <a rel=\"external\" href=\"https://doi.org/10.1080/01449290410001715714\">2004 review in <em>Behaviour &amp; Information Technology</em></a> went through the experiments on line length, columns, window size and line spacing, and concluded that characters per line, not the physical width of the line, is the variable that counts. <a rel=\"external\" href=\"https://journals.uc.edu/index.php/vl/article/view/5671\">Dyson and Kipping found in 1998</a> that people read text at 100 characters per line faster than text at 25, with the same comprehension, though the readers judged the long lines the hardest and lines of 55 characters the easiest. Later, <a rel=\"external\" href=\"https://doi.org/10.1006/ijhc.2001.0458\">Dyson and Haselgrove proposed in 2001</a> that about 55 characters gives the best balance of speed and comprehension on screen.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/is-e-ink-better-for-your-eyes/","url":"https://readwyn.com/blog/is-e-ink-better-for-your-eyes/","title":"Is e-ink easier on your eyes than a tablet?","summary":"What the studies found when people read for hours on e-ink, on a backlit tablet and on paper, and why brightness and blinking matter more than the screen type.","content_html":"<p>In a 2013 study, yes: people who read for over an hour on a backlit Kindle Fire HD reported more eye fatigue and blinked less than when they read on an e-ink Kindle Paperwhite or a paper book, and the e-ink reader and paper came out alike. But the evidence is a handful of small studies, and the researchers themselves suggest the difference came from how bright the screen was rather than from e-ink as such.</p>\n<h2 id=\"light-that-comes-from-the-page-and-light-that-comes-through-it\">Light that comes from the page, and light that comes through it</h2>\n<p>A tablet's LCD makes its own light: a backlight shines through a layer of liquid crystals, which let more or less of it reach you. An E Ink screen has no backlight. Room light falls on black-and-white pigment and bounces back, as it does from a printed page. E Ink's <a rel=\"external\" href=\"https://www.eink.com/tech/detail/Benefits\">page on the benefits of its displays</a> says the brightness you see depends on the light around you, and adds that users have told it they don't get the eye fatigue of an LCD over long sessions. E Ink gives no measurement for that. In the dark an e-reader needs its <a href=\"/blog/front-light-and-warm-light/\">front light</a>: LEDs in the sides of the screen whose light, <a rel=\"external\" href=\"https://www.eink.com/tech/detail/Front_Light\">E Ink says</a>, is angled down onto the ink instead of shining through the display towards you.</p>\n<h2 id=\"what-the-2013-study-measured\">What the 2013 study measured</h2>\n<p>Simone Benedetto and four colleagues published <a rel=\"external\" href=\"https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0083676\">E-Readers and Visual Fatigue</a> in PLOS ONE in December 2013. Twelve adults, none of whom had used an e-reader before, each read Maupassant's <em>Bel-Ami</em> in French on all three: the Fire HD, the Paperwhite and a paper copy set in the same type. The sessions were about ten days apart, in a random order, and averaged 73 minutes, under artificial light that measured 54 lux at the reader's forehead.</p>\n<p>On a six-item visual fatigue questionnaire, scores rose after the tablet and didn't change significantly after the Paperwhite or the book. Blinks per second were lower on the tablet than on either of the others, and pupils were smaller. A test of flicker sensitivity, which tracks alertness, fell after reading on every device, with no difference between them. Asked which they liked, readers preferred paper to both screens.</p>\n<p>The team also measured how bright each page looked. From 60 cm at eye level, the white of the tablet page gave about 28 cd/m², the paper 16 and the Paperwhite 11. The e-ink page was dimmer than the book. The authors suggest the tablet's higher luminance explains both the smaller pupils and the extra fatigue, and they note that a test of particular devices gives results that hold for those devices.</p>\n<h2 id=\"brightness-rather-than-technology\">Brightness rather than technology</h2>\n<p>The same group tested that idea in 2014. In <a rel=\"external\" href=\"https://doi.org/10.1016/j.chb.2014.09.023\">a study in Computers in Human Behavior</a>, 48 people read for an hour on a 22 inch Dell LCD set to either 20 or 140 cd/m², in a room lit at either 5 or 85 lux. Readers blinked less at the brighter setting. Self-rated eye fatigue rose after an hour in every group, though, with no effect of screen brightness or room light at all.</p>\n<p>Two years earlier, a Swiss team had found little to choose between the two. Eva Siegenthaler and colleagues had people read for several hours on either e-ink or an LCD and found them <a rel=\"external\" href=\"https://europepmc.org/article/MED/22762257\">very similar on both subjective and objective measures</a>. Those measures ran from self-rated fatigue and a letter search to reading speed, eye movements and the pupil's response to light. Their conclusion: \"It is not the technology itself, but rather the image quality\" that matters.</p>\n<h2 id=\"why-blinking-keeps-coming-up\">Why blinking keeps coming up</h2>\n<p>The American Academy of Ophthalmology's page on <a rel=\"external\" href=\"https://www.aao.org/eye-health/tips-prevention/computer-usage\">computers and eye strain</a>, dated June 2024, gives a normal rate of about 15 blinks a minute, falling to about 5 to 7 at a screen. It describes the result (dry, stinging or tired eyes, blurred vision) as the same strain that other close work such as reading or sewing brings, and says screens don't permanently damage the eyes. Its advice includes blinking on purpose, looking at something far away from time to time, and matching the screen's brightness to the room.</p>\n<p>That's why Benedetto's team counted blinks: their paper links fewer blinks to a poorer tear film and dry eyes. It also notes that people blink less when they read on any medium, and less again on a backlit display, so an e-reader is still close work.</p>\n<h2 id=\"blue-light\">Blue light</h2>\n<p>Blue light also comes up when people compare a tablet with an e-reader. What it does to sleep is a <a href=\"/blog/reading-on-an-e-reader-before-bed/\">separate question</a>; the case against it for eye strain is weak. The Academy's 2021 page <a rel=\"external\" href=\"https://www.aao.org/eye-health/tips-prevention/should-you-be-worried-about-blue-light\">on blue light</a> says there's no scientific evidence that light from digital devices damages the eye, and it doesn't recommend blue-light glasses for computer use. A <a rel=\"external\" href=\"https://www.cochrane.org/CD013244/EYES_do-blue-light-filtering-spectacle-lenses-reduce-digital-eye-strain\">Cochrane review published in August 2023</a>, covering 17 trials and 619 adults, found that blue-light filtering lenses may not reduce short-term eye strain from computer work, and rated that evidence low certainty.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/is-removing-drm-legal/","url":"https://readwyn.com/blog/is-removing-drm-legal/","title":"Is it legal to remove DRM from ebooks you bought?","summary":"What US, UK and EU law say about breaking the DRM on an ebook you paid for, the narrow exemptions for readers with print disabilities, and what store terms add.","content_html":"<p>In the United States, the United Kingdom and the European Union, breaking the DRM on an ebook that's still in copyright is itself unlawful, even when you paid for the book and only want to read it on another device. The exceptions are narrow, and the one that reaches ordinary buyers is for people with print disabilities. In the US they may break DRM that blocks a screen reader. In the UK and the EU the rights holder has to give them the means to make an accessible copy instead, and in the UK that follows a complaint to the Secretary of State. Outside the US exemption, a reader in the US or the UK who breaks DRM for their own use risks a civil claim; the criminal law is aimed at people who sell tools or act for profit. What follows describes the texts of the laws and isn't legal advice.</p>\n<p>Store terms add a contract on top of the law. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=201014950\">Kindle Store Terms of Use</a>, last updated on 21 October 2026, say Kindle books are licensed, not sold, and forbid any attempt to bypass or circumvent the service's digital rights management.</p>\n<h2 id=\"united-states\">United States</h2>\n<p><a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/1201\">Section 1201</a> of the Copyright Act, added by the DMCA on 28 October 1998, says no person shall circumvent a technological measure that effectively controls access to a protected work, and it counts decrypting as circumventing. The section's own exemptions are for nonprofit libraries, archives and schools, law enforcement, reverse engineering, encryption research and security testing, among a few others. None of them is for someone who just bought the copy.</p>\n<p>Anyone injured by a breach can sue under <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/1203\">section 1203</a> for statutory damages of $200 to $2,500 per act of circumvention, which a court may cut for someone who had no reason to know it was a violation. Crime starts only under <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/1204\">section 1204</a>: a wilful breach for commercial advantage or private financial gain carries up to $500,000 and five years for a first offence.</p>\n<p>Section 1201(c)(1) keeps fair use as a defence to infringement, but the ban on circumvention is a separate rule, and the way around it is an exemption. The Librarian of Congress grants these every three years, on the Register of Copyrights' recommendation. The <a rel=\"external\" href=\"https://www.copyright.gov/1201/2024/\">ninth rulemaking</a> ended with a final rule on 28 October 2024, and <a rel=\"external\" href=\"https://www.law.cornell.edu/cfr/text/37/201.40\">two of its exemptions</a> concern ebooks. One allows breaking DRM that blocks read-aloud or screen readers, when an eligible person under section 121 lawfully obtains the book and the rights owner is paid the market price of an inaccessible copy. The other lets researchers at nonprofit universities mine a corpus of ebooks for text and data. An exemption covers the act and nothing else: subsection (a)(1)(E) says it's no defence under any other provision, including the ban on making or selling circumvention tools. As of November 2026, the Copyright Office has opened the <a rel=\"external\" href=\"https://www.copyright.gov/1201/\">tenth proceeding</a>, and exemptions it renews would run from October 2027 to October 2030.</p>\n<h2 id=\"united-kingdom\">United Kingdom</h2>\n<p><a rel=\"external\" href=\"https://www.legislation.gov.uk/ukpga/1988/48/section/296ZA\">Section 296ZA</a> of the Copyright, Designs and Patents Act 1988 covers anyone who gets round DRM on a book, a film or a song while knowing, or having good reason to know, what they're doing (computer programs have a rule of their own). The section gives the publisher, the copyright owner and an exclusive licensee the same rights against that person as against an infringer, so the claim is civil. The offences in <a rel=\"external\" href=\"https://www.legislation.gov.uk/ukpga/1988/48/section/296ZB\">section 296ZB</a> target trade in circumvention devices and services, with up to two years in prison on indictment.</p>\n<p>If DRM stops you doing something the Act permits, <a rel=\"external\" href=\"https://www.legislation.gov.uk/ukpga/1988/48/section/296ZE\">section 296ZE</a> lets you complain to the Secretary of State, who can direct the rights owner to give you the means. Subsection 9 shuts that door for works made available on agreed contractual terms for access at a place and time each person chooses, a description that fits an ebook bought from an online store. Since 11 October 2018, subsection 9A reopens it in one case: when DRM prevents an accessible copy being made for a blind, visually impaired or otherwise print-disabled reader.</p>\n<h2 id=\"european-union\">European Union</h2>\n<p><a rel=\"external\" href=\"https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32001L0029\">Article 6 of Directive 2001/29/EC</a> requires member states to give adequate legal protection against knowing circumvention and against trade in tools. Each country writes its own rule and sets its own penalties. Article 6(4) asks states to make rights holders give the beneficiaries of certain exceptions the means to use them, but its fourth subparagraph excludes works supplied on demand under contract, in the same words as the UK's subsection 9.</p>\n<p>The <a rel=\"external\" href=\"https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32017L1564\">Marrakesh Directive (EU) 2017/1564</a> lets a print-disabled person, or someone acting for them, make an accessible copy of a work they have lawful access to. Its Article 3(4) applies parts of Article 6(4) to that exception, but not the fourth subparagraph, so the carve-out for store downloads doesn't apply. Article 3(5) adds that a contract can't override the exception.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/isbn-for-ebooks/","url":"https://readwyn.com/blog/isbn-for-ebooks/","title":"Does an ebook need its own ISBN?","summary":"Kindle and Apple Books sell ebooks without an ISBN, but an ebook that has one needs its own number, apart from the paperback and from each other format.","content_html":"<p>Not to sell it through Amazon or Apple: both take ebooks without one. If you do give an ebook an ISBN, though, the number is the ebook's alone, never the paperback's. Kindle Direct Publishing's help says <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G7DMSKCM9DVS65TC\">eBooks don't require an ISBN</a>, and Amazon gives every book an ASIN, its own number, whether or not it has an ISBN. Apple's Pages guide says you <a rel=\"external\" href=\"https://support.apple.com/guide/pages/tan229f2ef0c/mac\">don't need an ISBN-13 to publish to Apple Books</a>, though if you have one, enter it on the first submission: the same guide says that's the only time you can.</p>\n<p>The rule against sharing comes from the International ISBN Agency. An ISBN names one title in one edition and one format, so a paperback, an EPUB and a PDF of the same text <a rel=\"external\" href=\"https://www.isbn-international.org/content/what-isbn/10\">should each be identified separately</a>.</p>\n<p>The agency's <a rel=\"external\" href=\"https://www.isbn-international.org/content/guidelines-assignment-e-books/26\">guidelines for e-books</a> use the Kindle as their example of a version that can go without. Only Amazon sells the Kindle format, so with one source of supply the retailer doesn't need an ISBN. A publisher who assigns one anyway must keep it unique to the Kindle version. Even one format can need several numbers. Two EPUBs of one book count as separate products if their DRM gives different usage rights, for instance one that allows printing and one that doesn't. A version with unobtrusive “social” DRM, which enforces no usage constraints, needn't get a number apart from the version with no DRM at all. Editions with colour and with greyscale images are different products, but a colour edition read on a greyscale screen is the same product. A master file sent to a conversion service gets no ISBN unless the public can also get it in exactly that form, same format and no DRM; each version the service makes from it gets its own.</p>\n<p>In the UK there's no law that a book must have one, as the <a rel=\"external\" href=\"https://www.nielsenisbnstore.com/Home/Isbn\">Nielsen UK ISBN Store</a> points out, but booksellers and libraries order and list books by it. Nor does it protect the text: the International ISBN Agency says the number is only an identifier and <a rel=\"external\" href=\"https://www.isbn-international.org/content/what-isbn/10\">gives no legal or copyright protection</a>.</p>\n<p>Since <a rel=\"external\" href=\"https://www.isbn-international.org/content/what-isbn/10\">1 January 2007</a> every ISBN has 13 digits and starts with 978 or 979. Inside an EPUB it goes in <code>dc:identifier</code>, next to the title, the author and the rest of the <a href=\"/blog/epub-metadata-dublin-core/\">Dublin Core metadata</a>. <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/\">EPUB 3.3</a> requires that element in every book but names a UUID and a DOI as examples alongside the ISBN, so an EPUB with no ISBN at all is still a valid file.</p>\n<p>What a number costs depends on where you live. As of November 2026, <a rel=\"external\" href=\"https://www.nielsenisbnstore.com/Home/Isbn\">Nielsen</a>, which sells to publishers in the UK, Ireland and the British Overseas Territories, charges £93 for one ISBN, £174 for ten and £387 for a hundred, VAT included. In Canada, <a rel=\"external\" href=\"https://www.canada.ca/en/library-archives/services/publishers/isbn/about.html\">Library and Archives Canada</a> gives free ISBNs for ebooks as well as print to eligible publishers: a self-publishing author must be physically in Canada and must declare that a human wrote the book. KDP's own free ISBNs are for <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G201834170\">paperbacks and hardcovers only</a>, so an ebook on Amazon that should carry one needs a number from the national agency.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/justified-or-ragged-right/","url":"https://readwyn.com/blog/justified-or-ragged-right/","title":"Should ebook text be justified or ragged right?","summary":"Justified text reads well on a wide, hyphenated line, but ragged right is better on a narrow page at a large size, and some readers with cognitive disabilities find justified text hard.","content_html":"<p>Either works when each line holds enough words and the software hyphenates, because the extra space is then shared out in small amounts. On a narrow page at a large font size, ragged right reads better: there are fewer gaps to share the space between, so each gap grows. Amazon's layout engine allows for this: its <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GNY87A6WM6EK6YEE\">help page for publishers</a> says a Kindle switches to left-aligned text by itself when the font is large or the line is short.</p>\n<h2 id=\"where-the-setting-lives\">Where the setting lives</h2>\n<p>On a Kobo, tap the middle of the page to open the Reading Menu, tap the Text icon, and pick an option beside \"Justification\" (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017639913-Adjust-font-size-and-change-the-font-style-on-your-Kobo-eReader\">Kobo help</a>, November 2026). In the Kobo Books app for phones the setting is called \"Alignment\" (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017865293-Adjust-page-margins-line-spacing-and-more-on-the-Kobo-Books-app-iOS-Android\">Kobo help</a>). A Kindle book made for Amazon's <a href=\"/blog/kindle-enhanced-typesetting/\">newer layout engine</a> opens fully justified, and the same <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GNY87A6WM6EK6YEE\">publishing help</a> says readers can change that to left aligned (ragged right). In April 2025 Amazon <a rel=\"external\" href=\"https://www.aboutamazon.com/news/books-and-authors/kindle-accessibility-features-for-all-readers\">wrote</a> that for some readers left-aligned text, with wider spacing and margins, can reduce eye strain.</p>\n<h2 id=\"why-printed-books-justify\">Why printed books justify</h2>\n<p>Many readers take a straight right edge as the mark of a proper book. Matthew Butterick, a typographer, calls justification a matter of personal preference, not a sign of professional work, in <a rel=\"external\" href=\"https://practicaltypography.com/justified-text.html\">Practical Typography</a>. He notes that books tend to justify while newspapers and magazines mix both, yet he left-aligns his own web pages and word-processor documents, because the justification in browsers and word processors is crude next to a page-layout program.</p>\n<p>Early printers also had more ways to fill a line. In Gutenberg's 42-line Bible, as Johannes Ammon <a rel=\"external\" href=\"https://finaltype.de/en/topics/gutenbergs-justification\">describes it</a>, the compositors had about 290 different glyphs to choose from, including lowercase letters of more than one width and a heavy use of Latin abbreviations, and they let hyphens hang slightly into the margin. A printer fits each line once, for one typeface at one size. An e-reader has to do it again every time you change the font, the size or the margins.</p>\n<h2 id=\"what-the-reading-studies-found\">What the reading studies found</h2>\n<p>Studies from 1986 and 1991 disagree about whether justified text is slower to read. Justified text was slower in <a rel=\"external\" href=\"https://doi.org/10.1177/001872088602800204\">two studies</a> that Stanley Trollip and Gregory Sales ran in 1986, when computers often justified a line by inserting extra spaces between words. Comprehension didn't differ.</p>\n<p>Five years later, Rebecca Jubis and Paul Lee <a rel=\"external\" href=\"https://doi.org/10.2466/pms.1991.72.3c.1251\">tested 96 university students</a> on a CRT screen with four layouts. Justification made no difference to reading time. Finding typographical errors was fastest with \"even-gap\" justification, where small spaces were spread proportionally between the words, and slowest with ragged text whose word gaps varied. Their explanation was that a reader scans more slowly when unsure where the next word starts.</p>\n<h2 id=\"rivers-and-readers-with-cognitive-disabilities\">Rivers and readers with cognitive disabilities</h2>\n<p>Rivers, the channels of white that form when gaps line up down a page, are often treated as a cosmetic flaw. The web's accessibility standard takes them more seriously. The W3C's <a rel=\"external\" href=\"https://www.w3.org/WAI/WCAG22/Understanding/visual-presentation.html\">explanation of WCAG success criterion 1.4.8</a> says some people with cognitive disabilities find justified text hard to read, that rivers can make reading difficult or even impossible, and that squeezed lines make word boundaries hard to find. Among other things, the criterion asks for a way to get text that isn't justified. It also caps the length of a line, which the post on <a href=\"/blog/ideal-line-length-for-reading/\">line length</a> takes up.</p>\n<h2 id=\"hyphenation-and-the-language-of-the-book\">Hyphenation and the language of the book</h2>\n<p>Butterick's rule is that justified text needs hyphenation, or the word gaps become huge. The software that does it misses breaks on purpose. Frank Liang's 1983 Stanford thesis, <a rel=\"external\" href=\"https://www.tug.org/docs/liang/\">Word Hy-phen-a-tion by Com-put-er</a>, produced the pattern method that is standard in TeX and has since been adapted to many languages. His English set of about 4,500 patterns found 89% of the permitted breaks in a pocket dictionary's word list, and put virtually none in the wrong place. The missing 11% was a choice: the last breaks proved expensive to capture, and Liang judged a missed break safer than a wrong one. A few can't be found from the spelling at all, such as \"re-cord\" the verb beside \"rec-ord\" the noun, and the dictionaries don't always agree either: the thesis shows Webster's and the American Heritage breaking \"independent\" differently.</p>\n<p>Patterns are per language, so the software has to know the language first. The <a rel=\"external\" href=\"https://www.w3.org/TR/css-text-3/#hyphenation\">CSS Text standard</a>, which EPUB styling builds on, lets a reading system hyphenate automatically only text whose language is declared. Readwyn works the same way: it hyphenates a book whose language is English with the Knuth and Liang patterns of TeX, and breaks a book in any other language, or one with no language set, only between words.</p>\n<p>Not every script stretches its word spaces. The W3C's note on <a rel=\"external\" href=\"https://www.w3.org/International/articles/typography/justification\">approaches to full justification</a> explains that Arabic can lengthen the joins between letters (kashida), that Chinese and Japanese, which have no word spaces, adjust the space around punctuation and between characters, and that Thai breaks lines at word boundaries that aren't marked by any space.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kindle-direct-publishing-file-formats/","url":"https://readwyn.com/blog/kindle-direct-publishing-file-formats/","title":"Which file should you upload to Kindle Direct Publishing?","summary":"KDP recommends DOCX, EPUB or KPF for ebooks and no longer takes MOBI. Which one suits a novel, a picture book or a PDF, and how to check the file first.","content_html":"<p>A Word file (DOC or DOCX), an EPUB, or a KPF made in Kindle Create: those are the formats that Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200634390\">list of recommended formats</a> still takes. The table keeps a fourth row for MOBI, but all it holds now is a notice that MOBI is no longer accepted for fixed-layout books. Kindle Direct Publishing stopped taking it for reflowable books on 1 August 2021 and for fixed-layout books on 18 March 2025, as KDP's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GULSQMHU5MNH4EZM\">MOBI FAQ</a> records.</p>\n<p>For a novel, a memoir or any other book that is mostly text, upload a clean EPUB if you have one, the DOCX if you wrote in Word and kept the formatting plain, or a KPF if you'd rather let Amazon's own app do the layout. A comic or a picture book has a fixed page layout, and for those the FAQ takes an EPUB or a KPF. <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G79CTKR8BX79E96L\">Kindle Create</a> will build a comic eBook from the PDF of a comic.</p>\n<h2 id=\"amazon-converts-every-upload\">Amazon converts every upload</h2>\n<p>The file you upload is not the file a buyer downloads. According to the same FAQ, KDP processes an EPUB the way it processed a MOBI, and each reader gets a file built for what their device can do. The end of MOBI doesn't change how royalties are calculated, the same page adds.</p>\n<p>Books already on sale from a MOBI stay on sale, since the rule covers only new books and revisions. But if all you have of your own book is an old MOBI and you want to fix a typo, you need a new file. Readwyn converts a MOBI to EPUB 3 when you add it to a library, keeps the original beside it, and <code>epubcheck</code> reports no errors on the result. Open that EPUB in Kindle Previewer before you upload it, as you would any other.</p>\n<h2 id=\"word-files\">Word files</h2>\n<p>The formats page says most Word files convert well and warns that some with complex formatting don't. KDP's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200645680\">eBook Manuscript Formatting Guide</a> names the traps. Tab indents don't carry over, so it asks for a first-line indent of 0.2 inches (5 mm) set in the Normal style. Chapter titles take Heading 1. Leave out page numbers, headers, footers and the cover image, which KDP adds from the cover you supply at title setup. Footnotes turn into endnotes on upload, and a Word file that already went to print as a paperback has to be reformatted for the ebook. <a href=\"/blog/make-an-epub-from-word/\">Making an EPUB from Word</a> covers the other route.</p>\n<h2 id=\"epub\">EPUB</h2>\n<p>KDP takes both <a href=\"/blog/epub-2-or-epub-3/\">EPUB 2 and EPUB 3</a>: the FAQ says it accepts either if the file follows the Kindle Publishing Guidelines, and it names five programs whose exports work (Vellum, Scrivener, Sigil, Reedsy and Jutoh). A few of the guidelines go further than the EPUB specification. A <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GPNJPYK298J8TRRV\">reflowable book</a> must use a single column, must not use the CSS <code>position</code> property for alignment, and needs <code>dc:title</code> and <code>dc:language</code> in its OPF file. The <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GPST5EKFJ6MX5UPK\">cover guidelines</a> want the cover image declared in the OPF, with no extra HTML cover page, which can show the cover twice or make the conversion fail.</p>\n<h2 id=\"kpf-and-kindle-create\">KPF and Kindle Create</h2>\n<p>Kindle Create, Amazon's app for Windows and Mac, makes KPF files (for reflowable books the MOBI FAQ names Kindle Previewer as well), and the formats page calls KPF the go-to way to avoid formatting problems. Kindle Create <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G79CTKR8BX79E96L\">imports</a> DOC and DOCX files from Word, Pages or Google Docs, and PDFs for comics and Print Replica books. Alongside the KPF it saves a KCB project, and the <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GMUKXSTBCN6F2TGM\">Print Replica help page</a> tells you to back up the folder that holds it: the KPF is only for publishing, and the KCB is what you open to make an update. Kindle Create can also export a reflowable book as an EPUB, but <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GUGQ4WDZ92F733GC\">Getting Started with Kindle Create</a> lists publishing outside Kindle as unsupported and says that EPUB leaves out some Kindle features, such as image bleeds and chapter start images.</p>\n<h2 id=\"pdf-and-the-other-formats\">PDF and the other formats</h2>\n<p>KDP also accepts HTML (as ZIP, HTM or HTML), RTF, plain text and PDF, though PDF only for books in English, French, German, Italian, Portuguese, Spanish, Catalan, Galician and Basque. Kindle Create turns the PDF of a book with many images and complex formatting into a Print Replica book, and Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G42HENP2VHSN8VW8\">format comparison</a> says Print Replica doesn't work on Kindle e-readers, doesn't let readers resize text and doesn't support screen readers. Amazon advises it only for a book that must keep its exact print layout.</p>\n<p>One of KDP's own pages hasn't caught up. On 1 November 2026 its <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200673300\">supported languages page</a> still listed .mobi among the upload formats for languages that can't use PDF.</p>\n<h2 id=\"checking-the-file-before-you-publish\">Checking the file before you publish</h2>\n<p><a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G202131170\">Kindle Previewer</a> is free and opens .epub, .htm, .html, .xhtml, .opf, .kpf, .doc and .docx files, with Word support limited to English. On a Mac it needs macOS 12 or later. Under View › Book Information, the “Enhanced Typesetting” line reads “Supported” when a reflowable book qualifies for <a href=\"/blog/kindle-enhanced-typesetting/\">Amazon's newer layout engine</a>. The Online Previewer inside KDP does the same basic job with fewer options.</p>\n<p>To read the converted book on your own Kindle, the FAQ gives a roundabout route: download the HTML of the converted book from the previewer as a ZIP file, rename it from .zip to .epub and <a href=\"/blog/send-epub-to-kindle/\">email it to your Kindle</a>. Amazon warns that a personal file sent this way won't look exactly like the copy a buyer receives.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kindle-download-and-transfer-removed/","url":"https://readwyn.com/blog/kindle-download-and-transfer-removed/","title":"Why Kindle's Download & Transfer via USB is gone","summary":"Amazon removed Download & Transfer via USB on 26 February 2025 without giving a reason. What the option did, who lost most, and how Kindle books reach a device now.","content_html":"<p>Amazon switched off \"Download &amp; Transfer via USB\" on 26 February 2025 and gave no reason: its statement to The Verge confirmed the change and named the routes that still work, nothing more. Until then, the Content Library on Amazon's website let you save a copy of a Kindle book you'd bought to your computer and copy it to a Kindle by cable. Since that date the option isn't there. Books that were already on a Kindle stayed readable, and wireless delivery carried on as before.</p>\n<h2 id=\"what-the-option-did\">What the option did</h2>\n<p>The route was a leftover from the years when not every Kindle had Wi-Fi. As <a rel=\"external\" href=\"https://www.theverge.com/news/612898/amazon-removing-kindle-book-download-transfer-usb\">The Verge described it</a> just before the cut-off, you opened your Content Library while signed in to Amazon, chose \"More actions\" next to a book, picked \"Download &amp; transfer via USB\", and then chose one of the Kindles registered to your account. A copy of the book then downloaded to your computer. You plugged the Kindle into the computer and dragged the file across.</p>\n<p>It handled one book at a time, which made a backup of a large library slow work. Still, it got books onto a Kindle with no wireless connection, and it left you with an offline copy of each purchase on your own disk. The Verge also noted the format: the downloaded copies were AZW3 files, an older Kindle format, while current Kindles store books as KFX, a format Amazon began rolling out in 2015.</p>\n<h2 id=\"how-amazon-announced-it\">How Amazon announced it</h2>\n<p>In mid-February 2025 a warning appeared in the download pop-up itself, saying the option \"will no longer be available\" from 26 February and pointing to \"Deliver or Remove from Device\" for Kindles with Wi-Fi. <a rel=\"external\" href=\"https://goodereader.com/blog/kindle/amazon-is-not-to-be-trusted-anymore-with-their-kindle-e-reader\">Good e-Reader reported it</a> on 14 February and urged readers to back up their books while they still could; twelve days remained. The same article pointed out that Amazon was closing Kindle Vella, its serialised fiction platform, in the same month.</p>\n<p>Amazon then confirmed the change to The Verge. Its spokesperson, Jackie Burke, said customers could keep reading the books they'd already downloaded to their Kindles, and could get new content through the Kindle app, Kindle for Web, and Kindles with Wi-Fi.</p>\n<h2 id=\"who-lost-the-most\">Who lost the most</h2>\n<p>The original Kindle of 2007, the Kindle 2 of 2009 and the Kindle DX of the same year (Amazon counts it in its second generation) connected only over 2G and 3G mobile networks and had no Wi-Fi. When US carriers began switching those networks off in December 2021, Amazon told owners that these models <a rel=\"external\" href=\"https://www.pcworld.com/article/3629405/which-amazon-kindles-will-lose-internet-in-december.html\">could no longer download content wirelessly</a>, and PCWorld, citing Amazon's support FAQ, said the three had no other way to connect to a network. From then on, a cable from a computer was the only way to put a newly bought Kindle book on one of them, and Download &amp; Transfer via USB was the step that produced the file. After 26 February 2025 those Kindles could take no new store books at all.</p>\n<p>Amazon went further in 2026. On 20 May it ended support for every Kindle e-reader and tablet released in 2012 or earlier, and <a rel=\"external\" href=\"https://techcrunch.com/2026/04/08/amazon-to-end-support-for-older-kindle-devices-2012/\">TechCrunch reported</a> that those devices can no longer buy, borrow or download books, only read what's already on them. The post on <a href=\"/blog/old-kindle-end-of-support/\">old Kindles and the store</a> lists the models.</p>\n<p>Readers with a Wi-Fi Kindle lost the backup. The Verge, citing Good e-Reader, pointed out that Amazon has now and then removed books from its store and deleted them from Kindles, as it did with <a href=\"/blog/amazon-deleted-1984-from-kindles/\">Orwell's Nineteen Eighty-Four in 2009</a>, and that edited versions of some books, Roald Dahl's among them, have replaced the old copies on ebook platforms. A copy on your own disk was the one thing such a removal couldn't reach, and for most Kindle books the store no longer hands one out. The post on <a href=\"/blog/do-you-own-your-kindle-books/\">whether you own your Kindle books</a> goes through the licence terms behind that.</p>\n<h2 id=\"what-still-works\">What still works</h2>\n<p>A Kindle with Wi-Fi gets its books as it always did. In the Content Library, \"Deliver or Remove from Device\" sends a purchase to a registered Kindle over the air. This was the route Amazon's warning pointed to.</p>\n<p>The apps are the other route Amazon named. As of November 2026, the <a rel=\"external\" href=\"https://apps.apple.com/us/app/amazon-kindle/id302584613\">Kindle app on the App Store</a> runs on iPhone, iPad and Mac (the Mac version needs macOS 14 or later), and its listing says the books you buy on Amazon appear in the app automatically. Kindle for Web reads them in a browser.</p>\n<p>The cable itself still works. The Verge noted that you can still copy ebook files and documents of your own to a Kindle over USB, with Amazon's apps or with other software. The removal touched only the website's download of books bought from the Kindle Store. Your own files can also go wirelessly: in November 2026 Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G5WYD9SAF7PGXRNA\">Send to Kindle help page</a> lists EPUB, PDF, HTML, RTF, plain text and the common image types, and its web uploader takes files of 200 MB or smaller. Amazon converts an EPUB on the way, and <a href=\"/blog/send-epub-to-kindle/\">sending an EPUB to a Kindle</a> explains what arrives and why a file with DRM fails.</p>\n<p>One kind of purchase can still be downloaded: since <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GDDXGH9VR22ACM8U\">20 January 2026</a>, self-published books whose authors have confirmed them as DRM-free, with the conditions set out in <a href=\"/blog/where-to-buy-drm-free-ebooks/\">where to buy DRM-free ebooks</a>.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kindle-enhanced-typesetting/","url":"https://readwyn.com/blog/kindle-enhanced-typesetting/","title":"What is “Enhanced Typesetting” on a Kindle?","summary":"“Enhanced Typesetting: Enabled” means a Kindle book uses Amazon's 2015 layout engine, which hyphenates, kerns and reflows pages on Kindles from the sixth generation.","content_html":"<p>On a Kindle book's Amazon page you may see the line “Enhanced Typesetting: Enabled” among the product details. It means Amazon has built that book for the layout engine it introduced in 2015, so your Kindle or Kindle app lays out each page itself: it hyphenates long words, evens out the gaps between words, applies kerning and ligatures, and sizes drop caps to your font settings. <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G202087570\">Amazon's help for publishers</a> says a book with the feature also shows “Page Flip: Enabled”, and that a device which can't use it receives an older KF8 version of the same title.</p>\n<h2 id=\"the-rollout-in-2015\">The rollout in 2015</h2>\n<p>The iPhone and iPad app got the engine first: on 27 May 2015 Amazon updated it with a new default font, <a href=\"/blog/bookerly-kindle-font/\">Bookerly</a>, and a layout engine that Amazon said would adapt the page to \"any combination of screen size and font setting\", as <a rel=\"external\" href=\"https://thenextweb.com/news/buy-the-book\">The Next Web reported</a>. Hyphenation was on iOS only at that point, and Marco Arment wrote that Kindle e-readers would have to wait until later in the summer (<a rel=\"external\" href=\"https://mjtsai.com/blog/2015/05/28/kindle-typography-improvements/\">collected by Michael Tsai</a>).</p>\n<p>Three weeks later, on 17 June, Amazon announced a Kindle Paperwhite with a 300 ppi screen at $119. Its <a rel=\"external\" href=\"https://press.aboutamazon.com/2015/6/amazon-introduces-new-kindle-paperwhite-the-most-popular-kindle-now-even-better-still-only-119\">press release</a> spelled out what the engine does on an e-reader: hyphenation to break words in the right place, kerning to adjust the space between neighbouring letters, and ligatures, with the f and i of \"first\" drawn as one shape as its example. Amazon said the features would come in a free over-the-air software update in the coming weeks.</p>\n<p>That update arrived on 10 August 2015 as software version 5.6.5 for the Kindle Voyage, the $79 basic Kindle with a touch screen, and the second and third Kindle Paperwhite (<a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2015/08/10/major-kindle-5-6-5-update-finally-arrives-with-bookerly-font-and-hyphens/\">The eBook Reader</a>). Its list of changes promised hyphenation, justification, ligatures, kerning and pop-up footnotes in hundreds of thousands of books, for titles in seven languages: English, French, Italian, German, Spanish, Portuguese and Dutch. The update also brought Bookerly, which had replaced Caecilia on the 2015 Paperwhite, to these older models (<a rel=\"external\" href=\"https://www.engadget.com/2015-08-11-kindle-update-bookerly.html\">Engadget</a>).</p>\n<h2 id=\"a-new-file-format-behind-the-label\">A new file format behind the label</h2>\n<p>Owners soon found that a label on the store page didn't change the books already on their Kindles. A day after the update, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2015/08/11/kindle-ebooks-with-enhanced-typesetting-use-new-kfx-kindle-format\">The eBook Reader</a> explained that the new layout needs a new file format with the extension KFX, while most Kindle books were then AZW or AZW3 files. A book bought earlier kept its old format until you deleted it from the device and downloaded it again, most titles didn't support the engine yet, and KFX files went only to devices that could read them.</p>\n<p>The split between the two formats is still there. In November 2026 Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G202087570\">publisher help</a> lists the readers that get the engine: Kindle e-readers from the sixth generation (the <a href=\"/blog/which-kindle-do-i-have/\">Kindle model guide</a> helps you find yours), Fire tablets from the third, Kindle for iOS from version 4.9, and Kindle for Android from version 4.18.0; its <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GNY87A6WM6EK6YEE\">guidelines page</a> adds Kindle for PC and Mac from version 1.1. Older devices get KF8, the format Amazon <a rel=\"external\" href=\"https://techcrunch.com/2011/10/20/amazon-throws-a-minor-curveball-with-html5-kindle-8-format\">announced in October 2011</a> after it had <a rel=\"external\" href=\"https://www.droid-life.com/2011/09/28/amazon-announces-the-kindle-fire-for-199/\">shown the first Kindle Fire</a> on 28 September, and whose files end in .azw3 (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Kindle_File_Format\">Wikipedia</a>). Readwyn opens DRM-free AZW3 files and converts each one to EPUB 3 when it imports it, with the original file kept beside the copy (<a href=\"/blog/open-azw3-files-on-mac/\">more on AZW3 files</a>).</p>\n<h2 id=\"what-the-label-covers-now\">What the label covers now</h2>\n<p>The engine does more than it did in 2015. Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GNY87A6WM6EK6YEE\">help page on the feature</a> now also credits it with steadier line heights and margins, and with pictures that stay nearer their intended size and place. Text contrast adjusts so that it stays readable on the night, sepia and green pages, tables can be panned and zoomed and their text selected, and equations can be written in MathML. At a large font size or on a short line the layout changes: justified text falls back to left alignment, and two columns become one. The reader can also switch from full justification to a <a href=\"/blog/justified-or-ragged-right/\">ragged right edge</a>. Page Flip, the skimming view that Amazon <a rel=\"external\" href=\"https://press.aboutamazon.com/2016/6/amazon-announces-page-flip-a-new-way-to-hop-skim-and-jump-through-kindle-books\">announced in June 2016</a>, depends on the engine too, as do progressive download (you start reading before the whole book has arrived) and high-definition images. The same page lists more than forty languages where there were seven in 2015, among them Japanese, Hindi, Tamil, Malayalam, Gujarati, Marathi, Icelandic and Welsh.</p>\n<p><a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G202087570\">Amazon turns the feature on itself</a> “when possible”, and only for reflowable books uploaded as EPUB, DOC, DOCX or HTML; a publisher can check a file beforehand in Kindle Previewer, which shows an “Enhanced Typesetting” label in its navigation pane when the book qualifies. Kindle Direct Publishing stopped taking <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GULSQMHU5MNH4EZM\">MOBI uploads</a> for new and updated reflowable books on 1 August 2021 and for fixed-layout books on 18 March 2025 (the <a href=\"/blog/kindle-direct-publishing-file-formats/\">KDP formats guide</a> has the full list). A fixed-layout book never carries the label, because the engine works only on text that reflows.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kindle-popular-highlights/","url":"https://readwyn.com/blog/kindle-popular-highlights/","title":"How Kindle's popular highlights work","summary":"Why a Kindle underlines passages you never marked, who decides which ones, what the setting turns off, and what the underlines show about how far people read.","content_html":"<p>Read a bestseller on a Kindle and now and then a sentence turns up underlined that you never touched. That underline is a popular highlight: a passage that a number of other Kindle readers marked in their own copies. Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">user guide for the Kindle Colorsoft</a> says you see it when \"several other Kindle users have highlighted a particular passage\" of the book you're reading, and that you can turn the underlines off under the Aa menu, then More, then Popular Highlights.</p>\n<p>The guides never put a figure on \"several\". When <a rel=\"external\" href=\"https://www.nbcnews.com/business/consumer/amazon-peeking-over-kindle-users-shoulders-flna6C10406354\">NBC News wrote about the new feature in May 2010</a>, a month after it had begun reaching Kindles with the latest software, only passages highlighted by three or more users were included. The settings chapter of the <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_voyage/v1/en-US/html/kug.html\">Kindle Voyage guide</a> describes the switch as showing the passages that other Kindle users highlight most often: the underline is a tally of readers. The same data turns up in two more places: the Notes tab, which you reach from Go To on the reading toolbar, lists popular highlights next to your own, and About This Book, the card a Kindle can show when you first open a book, may include them.</p>\n<h2 id=\"the-hawking-index\">The Hawking Index</h2>\n<p>In July 2014 the mathematician Jordan Ellenberg used the underlines in <em>The Wall Street Journal</em> to guess how far buyers read. As <a rel=\"external\" href=\"https://gizmodo.com/the-books-everyone-starts-and-no-one-finishes-accordin-1600866004\">Gizmodo described his method</a>, he took the five most highlighted passages that Amazon listed for a book, averaged their page numbers and divided by the length of the book. If most readers reach the end, the highlights come from all over; if they give up in chapter three, the highlights bunch up near the front.</p>\n<p>He named the score after <em>A Brief History of Time</em>, which came out at 6.6 per cent. Daniel Kahneman's <em>Thinking, Fast and Slow</em> scored 6.8 per cent, and Thomas Piketty's <em>Capital in the Twenty-First Century</em> managed 2.4. Donna Tartt's <em>The Goldfinch</em>, at the other end, scored 98.5 per cent (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Hawking_Index\">Wikipedia's table of his scores</a>). Gizmodo called the method \"not entirely rigorous\", and five passages are a small sample. The scores do show what a popular highlight records: where a lot of readers were when they marked something, which for the books at the bottom of his list was the first few per cent of the text.</p>\n<h2 id=\"what-amazon-keeps\">What Amazon keeps</h2>\n<p>The guides don't say whether hiding the underlines also takes your own highlights out of the count. On the Voyage the switch sits under All Settings, Reading Options, Highlights &amp; About This Book, and its description talks only about what the Kindle displays, not about what it sends. In 2010 NBC News reported that the only way to stay out of the count was to turn off the automatic backup of your notes and highlights. Your highlights reach Amazon anyway while Whispersync for Books is on, as the Colorsoft guide explains; <a href=\"/blog/export-kindle-highlights/\">where your Kindle highlights are</a> covers that copy and the My Clippings file.</p>\n<p>Popular highlights carry no names. The same report says Amazon does not reveal the preferences of individual users. Seeing particular people's highlights once took a separate setting, Public Notes, which the guide for the Voyage, a 2014 model, says displayed the highlights and notes of users you followed in the Kindle community. The current Colorsoft guide doesn't list it.</p>\n<p>Amazon itself sees more. Back in 2012 the EFF's <a rel=\"external\" href=\"https://www.eff.org/pages/reader-privacy-chart-2012\">e-reader privacy chart</a> said the Kindle stores the last page read and may store annotations, highlights and other markings. Writing in <a rel=\"external\" href=\"https://slate.com/technology/2015/02/fifty-shades-of-grey-and-the-paradox-of-e-reader-privacy.html\">Slate in February 2015</a>, Neil Richards went further and listed what the design of the Kindle lets Amazon know about a reader: whether they finished the book, the page they're on, how long they spent on each page, and which passages they highlighted. What Kobo and Libby keep is in <a href=\"/blog/does-your-e-reader-track-what-you-read/\">does your e-reader track what you read</a>.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/library-ebooks-on-kobo-and-kindle/","url":"https://readwyn.com/blog/library-ebooks-on-kobo-and-kindle/","title":"How to read library ebooks on a Kobo or a Kindle","summary":"A Kobo borrows library ebooks over Wi-Fi in 13 countries, a Kindle only in the US. The steps, the models that work, and the Adobe route for other e-readers.","content_html":"<p>A Kobo borrows the book itself: most models since 2016 have the library catalogue built in and fetch loans over Wi-Fi, with no computer. A Kindle gets its loans through the Libby app, which hands them to Amazon to deliver as Kindle books; that route exists only in the United States, and Libby's help files the whole feature under <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/categories/kindle-books.htm\">Kindle Books (U.S. only)</a>. Both run through OverDrive, the company behind Libby, which <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019374554-Borrow-eBooks-from-the-public-library-using-your-library-s-website\">Kobo's help</a> says most public libraries use to manage their ebooks.</p>\n<h2 id=\"on-a-kobo\">On a Kobo</h2>\n<p>Kobo <a rel=\"external\" href=\"https://web.archive.org/web/20161201165517/http://news.kobo.com/press/releases/announcing-kobo-aura-one:-a-breakthrough-waterproof-ereader-designed-with-the-help-of-real-booklovers\">announced built-in OverDrive lending</a> with the Aura ONE on 17 August 2016. At launch it covered libraries in North America, the UK, Australia and New Zealand. As of October 2026, the feature works in <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017677053-Borrow-eBooks-from-the-public-library-using-your-Kobo-eReader\">13 countries</a>: Canada, the US, the UK, Australia, New Zealand, Denmark, Finland, Norway, Sweden, Taiwan, Hong Kong, Singapore and Thailand.</p>\n<p>Kobo's page names 15 models with the feature: the Libra Colour, Clara Colour and Clara BW, the Elipsa 2E, Clara 2E, Sage, Libra 2, Elipsa, Nia, Libra H2O and Forma, and the Aura ONE, Aura Edition 2, Aura H2O Edition 2 and Clara HD. The original Aura H2O isn't on it, only its Edition 2. If you're unsure which you own, <a href=\"/blog/which-kobo-do-i-have/\">the model number tells you</a>.</p>\n<p>The library card belongs to your Kobo account rather than to the device. To add it, go to My Account, Settings, Libby, Get Started, <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/4477058367895-About-the-Libby-app\">Add Library</a>, then search for your library and sign in. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017677053-Borrow-eBooks-from-the-public-library-using-your-Kobo-eReader\">borrowing page</a> gives the path as My Account, Library Cards, Add Library Card instead; if your Kobo shows no Libby entry, follow that one. Either way the Kobo shows a QR code to scan with your phone, or a web address and a code to type in a browser. Kobo recommends the QR code over “Sign in on this device”, which can run into sign-in problems. If you used to sign in to Libby with an email address and password, that may no longer work, and you'll need your card number and PIN instead. After the software update to version 6.0 or later, you have to add the card again.</p>\n<p>To borrow on the reader, go to Home, tap Discover, then OverDrive (or Libby), and search. On the book's page, tap the More icon and choose “Borrow with Overdrive/Libby App”, then sync. If no copy is free, the same menu offers a hold, and Kobo emails you when the book is ready; the <a href=\"/blog/why-library-ebooks-have-waitlists/\">licences that libraries buy</a> explain the queue. Loans return themselves when they expire. Renewing means borrowing again: a <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017676973-Can-I-renew-eBooks-I-ve-borrowed-with-Overdrive\">“request again” option</a> appears three days before the loan ends, and you may have to wait if someone else holds the book.</p>\n<p>You can also borrow in Libby on a phone and let the book reach the Kobo at its next sync, as long as the Kobo is signed in with the same card. Not every Libby ebook is <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6251.htm\">compatible with Kobo</a>, so some loans never appear, and Libby audiobooks and magazines don't work on Kobo readers at all.</p>\n<h2 id=\"on-a-kindle\">On a Kindle</h2>\n<p>Kindle library lending began in September 2011, when OverDrive added Kindle-format titles for American public and school libraries; <a rel=\"external\" href=\"https://tidbits.com/2011/09/23/overdrive-adds-kindle-format-titles-to-library-ebook-lending/\">TidBITS reported</a> that OverDrive served more than 11,000 of those libraries at the time. Fifteen years on, it's still an American feature, and Libby's help describes no way to get a library book onto a Kindle e-reader anywhere else.</p>\n<p>To <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6017.htm\">send a loan to a Kindle</a>, open Shelf in Libby, then Loans, tap Read With... and choose Kindle. Libby sends you to Amazon's website, where you sign in if asked and tap Get Library Book; the book then goes to the devices on your Amazon account. On a Fire tablet, Libby recommends installing the Libby app and reading there instead.</p>\n<p>The Kindle option isn't always there. A book's page in Libby shows a “Read With” section, and if Kindle isn't listed, that title isn't available for Kindle. There's a <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6027.htm\">less obvious cause</a> too: once you've downloaded a loan in another format, such as EPUB, you can't send it to Kindle unless you borrow it again. A reading preference set to Libby also blocks it.</p>\n<p>A Kindle loan goes back to the library automatically when its lending period ends. To <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6015.htm\">return one early</a>, find it on Amazon's “Your Books” page and choose “Return borrowed book”, or open the menu on its cover on the Kindle and tap “Remove from Library” (“Return to Public Library” on older models). If you start from Libby instead (Manage Loan, Return Early, Return with Kindle), it opens that Amazon page and afterwards asks you to tap “Yes, I've Returned It”. If the loan still shows in Libby after that, Libby's help says to sync your shelf by hand.</p>\n<h2 id=\"older-kobos-and-other-e-readers\">Older Kobos and other e-readers</h2>\n<p>E-readers without a built-in catalogue take a longer route through Adobe Digital Editions (ADE), free software for Windows and Mac. The <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6059.htm\">transfer</a> starts at libbyapp.com on a computer: Shelf, Loans, Read With..., Other Options..., then EPUB or PDF. What downloads is a small file ending in .acsm, not the book: OverDrive calls it <a rel=\"external\" href=\"https://help.overdrive.com/en-us/0577.html\">download instructions</a> for ADE. ADE opens it, fetches the book, and copies it to the e-reader when you drag it onto the device under Devices. You authorise ADE on the computer once, and the e-reader is <a rel=\"external\" href=\"https://help.overdrive.com/0005.html\">authorised</a> with the same account during the first transfer. OverDrive lists two exceptions: the 2015 NOOK GlowLight Plus 6\" doesn't use ADE at all, and a Tolino downloads library books itself. On a Tolino you <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6282.htm\">borrow in Libby</a> and then download on the reader, signed in with the same library card.</p>\n<p>The ADE sign-in is changing. As of October 2026, <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017814074-Add-eBooks-with-Adobe-Digital-Editions\">Kobo's ADE page</a> says Adobe is replacing the Adobe ID with a ByteBooks ID for DRM-protected books. A computer or reader that is already authorised keeps working until ADE asks you to sign in again. Kobo advises creating the ByteBooks ID with the email address of your old Adobe ID, so that you keep access to the books you already have, and says the error E_ADEPT_RESET_PW_REQUIRED means you need to make that switch.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/literata-font/","url":"https://readwyn.com/blog/literata-font/","title":"Literata, the font Google commissioned for ebooks","summary":"Literata is the serif TypeTogether drew for Google Play Books in 2015, designed for phone and tablet screens, and free as a variable font with optical sizes since Literata 3.","content_html":"<p>Literata is a serif typeface that the foundry TypeTogether designed for Google Play Books. It shipped in version 3.4.5 of the Play Books app on 6 May 2015, replacing Droid Serif as the default reading font, and Google announced it twelve days later as “perfect for long reads on all devices” (<a rel=\"external\" href=\"https://9to5google.com/2015/05/18/google-play-books-default-font-replace-droid-serif/\">9to5Google, 18 May 2015</a>). In December 2018 the family went open source as a variable font, and in July 2020 it gained an optical size axis. Anyone can download it free from <a rel=\"external\" href=\"https://fonts.google.com/specimen/Literata\">Google Fonts</a>.</p>\n<h2 id=\"the-brief\">The brief</h2>\n<p>Google's creative brief is dated 1 March 2014, and TypeTogether published it with its sketches in a <a rel=\"external\" href=\"https://github.com/googlefonts/literata/blob/main/Documentation/v2/specimen_material/Process%20and%20development%20of%20Literata.pdf\">process document</a> that now sits in the font's repository. It asked for a face for continuous reading on high resolution phones and tablets, a look that would set Play Books apart from other ebook apps, and a family that would sit well beside Google's Roboto. The target reader was described in some detail: someone who reads on a phone in short bursts during the day and moves to a tablet for longer sessions in the evening. The same family had to hold up on everything from electronic ink to LED screens with different rendering engines.</p>\n<p>TypeTogether's <a rel=\"external\" href=\"https://www.typetogether.com/literata-font\">history of the project</a> puts the commission in context: Apple had already added better fonts to iBooks (TypeTogether's own Athelas among them), and Amazon had commissioned a new face for its Kindles. The Kindle's font, Bookerly, has <a href=\"/blog/bookerly-kindle-font/\">a post of its own</a>.</p>\n<h2 id=\"how-it-was-drawn-for-screens\">How it was drawn for screens</h2>\n<p>The designers turned the brief into five rules in April 2014: exaggerated details such as serifs and terminals that stay recognisable at small sizes, open apertures that don't fill in when the pixels spread (the document calls this \"pixel gain\"), low contrast without thin strokes, wider letters with loose spacing, and a large x-height with short ascenders and descenders. Their wish list also floated Georgia as a benchmark and complained that Caecilia and Droid Serif looked too even and mechanical.</p>\n<p>The first round of sketches had six designs of only eight letters each, laid out on a scale from old style to modern. TypeTogether dropped the two most extreme before showing Google the rest. Google's feedback pushed the design towards a hybrid: a Scotch Roman with old style features, slanted stress, and letters of varied width, which the team expected to give the page a less mechanical texture.</p>\n<p>For the italic, TypeTogether offered three options in June 2014, and the one that went ahead was upright: italic letter shapes with no slant. The aim was to sidestep the limits of the square pixel grid, and the wish list had already asked for horizontal and vertical strokes, which the grid draws most cleanly. Looking back, TypeTogether says the darkish regular weight and the generous letter spacing both tested well on backlit screens, and that the upright italic fixed many pixelation problems.</p>\n<p>Despite that rule of a large x-height, 9to5Google found Literata's x-height lower than that of Droid Serif, the font it replaced, and its ascenders taller. The first release had two weights, regular and bold, each with an italic, for Latin, Greek and Cyrillic.</p>\n<h2 id=\"literata-3-and-its-optical-sizes\">Literata 3 and its optical sizes</h2>\n<p>Variable fonts, an extension of OpenType that puts a whole range of styles into one file, were announced by Adobe, Apple, Google and Microsoft near the end of 2016. Dave Crossland of Google Fonts had imagined Literata growing into a much larger family, and TypeTogether <a rel=\"external\" href=\"https://www.typetogether.com/literata-font\">redesigned it for the new format</a>, a project it dates to January 2020. Starting from the regular, it sketched a headline style with more contrast, slightly narrower letters and tighter spacing, then a caption style that went the other way, with sturdier, flatter serifs. It drew the four extremes of the design space first, to fix the ends of both axes, and then added three of the four corner masters, because interpolating from the extremes alone gives kinks, wrong modulation and uneven stroke thickness at the corners of the design space.</p>\n<p>The <a rel=\"external\" href=\"https://github.com/googlefonts/literata/blob/main/Documentation/WhatsNew-in-v3.md\">notes for version 3</a> say the family grew from 8 styles to 50, spread over four optical size series. The repository's static files carry the series in their names: plain <code>Literata</code>, then <code>Literata7pt</code>, <code>Literata36pt</code> and <code>Literata72pt</code>. Google Fonts serves two variable files instead, one roman and one italic, and its <a rel=\"external\" href=\"https://github.com/google/fonts/blob/main/ofl/literata/METADATA.pb\">metadata for the family</a> gives the weight axis as 200 to 900 and the optical size axis as 7 to 72. On a web page a browser picks the optical size by itself: <a rel=\"external\" href=\"https://developer.mozilla.org/en-US/docs/Web/CSS/font-optical-sizing\">MDN</a> says optical sizing is on by default for any font with an <code>opsz</code> axis.</p>\n<p>Version 3 made Bold and SemiBold heavier, so a design moved from version 2 comes out darker. To get the colour of version 2 back, the notes recommend weight axis positions of 450 for Medium, 505 for SemiBold and 580 for Bold.</p>\n<h2 id=\"versions-and-dates\">Versions and dates</h2>\n<p>The <a rel=\"external\" href=\"https://github.com/googlefonts/literata/releases\">releases on GitHub</a> give this timeline:</p>\n<ul>\n<li>7 December 2018, version 2.100: the first public release, with variable fonts and TrueType files.</li>\n<li>12 June 2019, versions 2.200 and 2.201.</li>\n<li>29 July 2020, version 3.002: the optical size axis arrives.</li>\n<li>5 to 19 May 2023, versions 3.101 to 3.103: among the fixes, kerning for fractions and interpolation. As of November 2026, 3.103 is still the latest release.</li>\n</ul>\n<p>In 2021 the family won gold at the Indigo Awards and was a Modern Cyrillic winner, the repository's <a rel=\"external\" href=\"https://github.com/googlefonts/literata\">README</a> says. The <a rel=\"external\" href=\"https://github.com/google/fonts/blob/main/ofl/literata/DESCRIPTION.en_us.html\">Google Fonts description</a> adds that the family exists in two versions, one for print and one for ebooks, and that Google Fonts serves the print version. The files are under the <a rel=\"external\" href=\"https://openfontlicense.org/open-font-license-official-text/\">SIL Open Font License 1.1</a>, which is why they can go into an app, an ebook or the <code>fonts</code> folder of <a href=\"/blog/add-fonts-to-kobo-and-kindle/\">a Kobo or a Kindle</a> without asking anyone's permission.</p>\n<h2 id=\"where-you-ll-meet-it\">Where you'll meet it</h2>\n<p>Readwyn sets book text in Literata, and it's one of six font tiles in the reader's Text panel, with Georgia, Palatino, Charter, Avenir and Helvetica. The text you're reading on this page is Literata too.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/make-an-epub-from-word/","url":"https://readwyn.com/blog/make-an-epub-from-word/","title":"How to make an EPUB from a Word document","summary":"Word can't save an EPUB, so you style the manuscript and convert the .docx with pandoc, Pages, calibre or WordToEPUB. The steps, and what gets lost.","content_html":"<p>Save the manuscript as a .docx with proper heading styles, then convert it with a tool such as pandoc, Apple Pages, calibre or the DAISY Consortium's WordToEPUB. Word itself has no EPUB option in its Save As list (checked in November 2026): Microsoft's <a rel=\"external\" href=\"https://learn.microsoft.com/en-us/office/compatibility/office-file-format-reference\">file format reference for Word</a> (Office 2016 to Office LTSC 2024) runs from .docx and RTF through OpenDocument and PDF to web pages, and EPUB isn't on it.</p>\n<p>With <a rel=\"external\" href=\"https://pandoc.org/MANUAL.html\">pandoc</a>, the command-line converter, the job takes five steps:</p>\n<ol>\n<li>\n<p>Give every chapter title Word's Heading 1 style and every section title Heading 2. Don't fake a heading with bold 16 pt text.</p>\n</li>\n<li>\n<p>Accept or reject every tracked change. You can leave the comments in: pandoc drops them.</p>\n</li>\n<li>\n<p>Save the file as a Word Document (.docx), not the older .doc.</p>\n</li>\n<li>\n<p>In Terminal, run the conversion (change the names to yours):</p>\n<pre><code>pandoc manuscript.docx -o book.epub -M title=&quot;My Book&quot; -M author=&quot;Your Name&quot; -M lang=en-GB --epub-cover-image=cover.jpg\n</code></pre>\n</li>\n<li>\n<p>Run the result through <a href=\"/blog/check-epub-with-epubcheck/\">EPUBCheck</a>, then open it in a reading app and page through the chapters.</p>\n</li>\n</ol>\n<h2 id=\"why-the-styles-do-the-work\">Why the styles do the work</h2>\n<p>Converters find chapters by style, not by how a line looks. According to <a rel=\"external\" href=\"https://manual.calibre-ebook.com/conversion.html#convert-microsoft-word-documents\">calibre's manual</a>, it builds the book's table of contents from paragraphs marked Heading 1, Heading 2 and so on. Apple's notes on <a rel=\"external\" href=\"https://support.apple.com/108362\">book creation in Pages</a> describe chapters the same way: a new chapter starts at each paragraph in a heading style that the document's table of contents lists without an indent. (You remove the indent in the table of contents view, with the style selected.)</p>\n<p>Pandoc reads Word styles too. Its manual shows a paragraph in a style based on Word's Block Text coming out as a quotation. By default it also splits the EPUB into a separate file at each top-level heading. <code>--split-level 2</code> splits at the next level down, and the manual suggests it for a long document with few chapters, because some reading apps slow down on very large chapter files. Readers see the same chapters either way.</p>\n<p>Steps 2 and 3 come from how pandoc reads the file. Its <code>--track-changes</code> option is set to accept by default, so an insertion you meant to reject still ends up in the book, and a passage marked for deletion disappears. Pandoc reads .docx but not .doc. For an old .doc, calibre's manual gives two ways round: save it as .docx from a newer Word or from LibreOffice, or save it from Word as a web page with “Save as Web Page, Filtered” and convert that HTML instead.</p>\n<h2 id=\"metadata-and-the-cover\">Metadata and the cover</h2>\n<p>Pandoc fills the EPUB's title, author and language from the document's metadata, and the <code>-M</code> options in the command set them outright. The language comes from <code>lang</code>; without it, pandoc falls back to the locale of the computer. Pandoc also makes a random identifier for the book, and you can give an ISBN instead in a metadata file (the manual lists the <a rel=\"external\" href=\"https://pandoc.org/MANUAL.html#epub-metadata\">fields it recognises</a>). For the cover, the manual recommends an image under 1,000 pixels in width and height. Stores ask for other sizes, which the post on <a href=\"/blog/ebook-cover-size/\">ebook cover sizes</a> lists.</p>\n<h2 id=\"pages-wordtoepub-calibre-and-google-docs\">Pages, WordToEPUB, calibre and Google Docs</h2>\n<p>On a Mac, Pages opens .docx files; Apple's guide suggests <a rel=\"external\" href=\"https://support.apple.com/guide/pages/tancdeedb11c/mac\">dragging the file onto the Pages icon</a>, because a double-click opens Word if you have it. Then choose <a rel=\"external\" href=\"https://support.apple.com/guide/pages/tance1161f26/mac\">File › Export To › EPUB</a>, fill in the title and author, pick a cover and Reflowable, and under Advanced Options turn on “Use table of contents” and “Embed fonts”.</p>\n<p>On Windows, <a rel=\"external\" href=\"https://daisy.org/activities/software/wordtoepub/\">WordToEPUB</a> is free from the DAISY Consortium, runs on Windows 7 to 11 and adds a button to Word's ribbon.</p>\n<p><a rel=\"external\" href=\"https://manual.calibre-ebook.com/conversion.html#convert-microsoft-word-documents\">Calibre</a> converts .docx files from Word 2007 and newer without a detour through another format. Google Docs has exported EPUB since <a rel=\"external\" href=\"https://workspaceupdates.googleblog.com/2016/03/export-google-docs-files-as-epub.html\">March 2016</a>, when Google put “EPUB Publication (.epub)” under File › Download as.</p>\n<h2 id=\"what-doesn-t-carry-over\">What doesn't carry over</h2>\n<p>Layout tied to the printed page goes first. Apple's notes for Pages list linked text boxes and text wrapped around objects as things that won't appear in a reflowable EPUB, and they advise setting images and shapes inline with the text. Footnotes become endnotes in that export, and a font that isn't embedded gives way to the reading app's preferred font. If the layout has to survive, that's a <a href=\"/blog/fixed-layout-vs-reflowable/\">fixed-layout EPUB</a>, and readers lose the ability to change the type size.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/marginalia/","url":"https://readwyn.com/blog/marginalia/","title":"What is marginalia, and do ebook notes count?","summary":"Marginalia are the notes readers write in the margins of books. Where they come from, and where a note you type in an ebook is kept instead.","content_html":"<p>Partly. Marginalia are the notes, underlinings, and doodles that readers leave in the margins of a book. A note in an ebook does the same job for the person who writes it, but it isn't part of the copy. The reading app usually keeps it apart from the file, tied to a device or an account, so nobody who opens another copy will see it. Delete the book, or reset the e-reader, and the note can go with it.</p>\n<h2 id=\"notes-that-outlived-their-readers\">Notes that outlived their readers</h2>\n<p>People wrote in margins long before the word marginalia existed: its first recorded use in English is in <em>Blackwood's Magazine</em> in 1819, according to the Oxford English Dictionary as <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Marginalia\">cited on Wikipedia</a>. Samuel Taylor Coleridge annotated so many books that his marginalia fill <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Samuel_Taylor_Coleridge\">six volumes of his Collected Works</a>, published from 1980.</p>\n<p>The most famous margin note belongs to mathematics. Pierre de Fermat was reading Claude Bachet's Latin edition of Diophantus's <em>Arithmetica</em>, printed in 1621, when he wrote beside one problem that no cube can be split into two cubes, nor any higher power into two powers of the same kind. He added that he had found a proof, but \"hanc marginis exiguitas non caperet\": the margin was too narrow for it (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Fermat%27s_Last_Theorem\">the Latin text</a>). MacTutor dates the note to <a rel=\"external\" href=\"https://mathshistory.st-andrews.ac.uk/HistTopics/Fermat&#x27;s_last_theorem/\">around 1630</a>, while Wikipedia's article on the theorem gives around 1637. The note only became known after his son Samuel published an edition of Bachet's translation with his father's notes in 1670 (<a rel=\"external\" href=\"https://mathshistory.st-andrews.ac.uk/Biographies/Fermat/\">MacTutor's life of Fermat</a>), and the theorem stayed unproved until Andrew Wiles's work of 1993 and 1994.</p>\n<p>Ordinary readers' notes are the subject of <a rel=\"external\" href=\"https://booktraces-public.lib.virginia.edu/about/\">Book Traces</a>, a project directed by Andrew Stauffer at the University of Virginia. It hunts for marked copies of 19th and early 20th century books in circulating library stacks, the books most at risk: older ones have mostly gone to special collections, newer ones are still in copyright, and the public domain ones in between get withdrawn once a scan sits in HathiTrust or Google Books. Until at least the 1950s, the project says, academic libraries filled their shelves with donated secondhand books, so those stacks still hold the annotations, letters, pressed flowers, and even locks of hair of the first owners. In November 2026 its <a rel=\"external\" href=\"https://booktraces-public.lib.virginia.edu/\">home page</a> counted 3,285 submissions.</p>\n<h2 id=\"where-an-ebook-note-lives\">Where an ebook note lives</h2>\n<p>On a Kobo, a note is a highlight with text attached (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017639993-Add-notes-to-your-eBook\">Add notes to your eBook</a>), and where it's kept depends on where the book came from. Notes in a book bought from Kobo follow your account to your other Kobo devices and apps. Notes in a book from anywhere else stay on that one eReader, and Kobo says they're removed if you delete the book from it. Handwriting with a <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/1500001927562-Write-notes-in-your-book-with-Kobo-Stylus\">Kobo Stylus</a> (in November 2026, on the Sage, Elipsa, Elipsa 2E, and Libra Colour) never leaves the eReader, a reset or a sign out loses it, and in an EPUB it can't be exported. In a PDF without DRM, though, the stylus writes into the file itself, so a copy taken off the device carries the notes: the nearest an ebook gets to a pencilled margin.</p>\n<p>The typed notes do have a way off the device: a sync into Readwyn brings them into its Notebook with their highlights, grouped by chapter, and they stay in the library after the book leaves the Kobo. Stylus drawings aren't read (<a href=\"/blog/kobo-notes-in-your-library/\">how Kobo notes reach the library</a>).</p>\n<h2 id=\"a-shared-format-is-still-a-draft\">A shared format is still a draft</h2>\n<p>The W3C's Publishing Maintenance Working Group is drafting <a rel=\"external\" href=\"https://www.w3.org/TR/epub-anno-10/\">EPUB Annotations 1.0</a>, an exchange format for notes, highlights, and bookmarks built as a profile of the Web Annotation Data Model. The notes go either in the <code>META-INF</code> folder of the EPUB, where they'd travel with the copy as a pencilled note does, or in a separate ZIP file with the proposed extension <code>.annotations</code>. A note's text is plain: the editors considered Markdown and left it out, because reading apps today store notes without formatting. Until a format like it is finished and adopted, an export carries a note's text and, at best, a page number, not an address that another app can follow, which is <a href=\"/blog/move-highlights-between-reading-apps/\">why highlights don't move between reading apps</a>.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/marrakesh-treaty/","url":"https://readwyn.com/blog/marrakesh-treaty/","title":"What is the Marrakesh Treaty?","summary":"The Marrakesh Treaty of 2013 lets libraries for print-disabled readers make accessible copies of books without permission and send them abroad. Who it covers and who joined.","content_html":"<p>The Marrakesh Treaty is a copyright treaty adopted on 27 June 2013 by the member states of the World Intellectual Property Organization (WIPO), and in force since 30 September 2016. A country that joins must let libraries for the blind and similar non-profit bodies turn a published work into braille, audio, large print or another accessible format without the copyright holder's permission, and must let those copies cross borders. WIPO's <a rel=\"external\" href=\"https://www.wipo.int/pressroom/en/articles/2013/article_0017.html\">press release on the day of adoption</a> put the problem in figures from the World Blind Union: of the million or so books published each year, fewer than 5 per cent reached a format that visually impaired readers could use.</p>\n<p>The release also cites a WIPO survey of 2006 that found fewer than 60 countries with any copyright exception for these readers, and those exceptions rarely reached imports or exports, because copyright stops at the frontier. To swap accessible editions abroad, a library had to negotiate a licence with the rightholders, or else produce the book again itself.</p>\n<p>In the <a rel=\"external\" href=\"https://www.wipo.int/wipolex/en/text/301016\">treaty text</a>, Article 3 reaches well beyond blindness. It covers anyone who is blind; anyone with a visual impairment or \"a perceptual or reading disability\" that stops them reading print about as well as other people; and anyone who, because of a physical disability, can't hold or handle a book, or can't focus or move their eyes well enough to read. An agreed statement adds that nobody has to go through every possible test and treatment to qualify. The works are text, music notation and related illustrations, and another agreed statement brings in audiobooks.</p>\n<p>The copying is done by an \"authorized entity\", which Article 2 defines as a non-profit or government body that its government recognises as serving these readers. Under the model exception of Article 4, it needs lawful access to the work, may change only what accessibility requires, and may supply the copies only to beneficiaries. A reader, or a carer acting for them, may also make an accessible copy for personal use of a book they lawfully have. A country may narrow this to books not already sold in an accessible format on reasonable terms in its market, if it tells WIPO (Article 4(4)), and it decides whether rightholders get paid. Article 7 deals with DRM. Where a country's law protects technical locks against circumvention, that protection mustn't stop beneficiaries from using the exception (the <a href=\"/blog/is-removing-drm-legal/\">post on whether removing DRM is legal</a> covers the EU's version).</p>\n<p><a rel=\"external\" href=\"https://www.wipo.int/wipolex/en/treaties/ShowResults?search_what=C&amp;treaty_id=843\">WIPO's table of members</a> shows India as the first to ratify, on 24 June 2014. The treaty needed 20 parties, and <a rel=\"external\" href=\"https://www.wipo.int/treaties/en/ip/marrakesh/summary_marrakesh.html\">WIPO's summary</a> dates the 20th to 30 June 2016, the day Canada acceded. Since 1 January 2019 the European Union has been bound as a single party, and the UK in its own right since 1 January 2021.</p>\n<p>The United States signed in October 2013 but first had to change its law. According to the <a rel=\"external\" href=\"https://www.copyright.gov/legislation/2018_marrakesh_faqs.pdf\">US Copyright Office</a>, the Marrakesh Treaty Implementation Act of 9 October 2018 widened section 121 of the Copyright Act from \"nondramatic literary works\" to all literary works plus music in text or notation, and the US deposited its ratification as the 50th member on 8 February 2019. It took effect there on 8 May.</p>\n<p>By the treaty's tenth anniversary, WIPO <a rel=\"external\" href=\"https://www.wipo.int/en/web/abc/w/news/2026/ten-years-of-the-marrakesh-treaty-turning-access-into-action\">counted 106 contracting parties</a> covering 132 of its member states. Türkiye, the latest, <a rel=\"external\" href=\"https://www.wipo.int/wipolex/en/treaties/ShowResults?search_what=C&amp;treaty_id=843\">ratified on 28 September 2026</a> and is bound from 28 December.</p>\n<p>The exchange has a shared catalogue, the <a rel=\"external\" href=\"https://www.accessiblebooksconsortium.org/en/globalbooks\">ABC Global Book Service</a>, hosted by WIPO's Accessible Books Consortium and once called TIGAR. Libraries pay nothing. Each signs one agreement with WIPO, adds its own catalogue, and can request files from the others; in November 2026 the page listed 156 participating entities and more than 8,000 braille music scores. Where the treaty is in national law, a library can also give its readers a login to search and download books themselves through the ABC Beneficiary Application or Dolphin's EasyReader app.</p>\n<p>The first transfer, in 2016, was an audiobook from the Canadian National Institute for the Blind to Vision Australia, which <a rel=\"external\" href=\"https://www.wipo.int/en/web/abc/w/news/2026/ten-years-of-the-marrakesh-treaty-turning-access-into-action\">by WIPO's estimate</a> saved US$2,000 it would otherwise have spent making its own copy. In 2025 the service delivered 240,500 files to readers. Bookshare, the US online library, <a rel=\"external\" href=\"https://www.bookshare.org/uncategorized/marrakesh-treaty-what-you-need-to-know\">wrote in 2019</a> that the treaty let it offer about 100,000 books, until then open only to American members, to readers in the other member countries. Its example was Latin American countries importing Spanish-language books from Spain.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/microsoft-reader-lit-files/","url":"https://readwyn.com/blog/microsoft-reader-lit-files/","title":"What was Microsoft Reader, and can you still open a .lit file?","summary":"Microsoft Reader was Microsoft's free ebook app from 2000 to 2012. Its .lit files can still be converted, unless the shop sold them copy-protected.","content_html":"<p>You can still turn a .lit book into an EPUB, though Microsoft Reader, the free Windows and Pocket PC program the format was made for, has been gone since 2012. calibre's <a rel=\"external\" href=\"https://manual.calibre-ebook.com/faq.html\">FAQ</a>, as it stands in November 2026, lists LIT among the input formats of its converter and EPUB among the outputs. That holds for an unprotected file. A copy-protected (Owner Exclusive) book opens only in a copy of Reader activated with its buyer's account, and Microsoft said it would give no way to move such books to another app.</p>\n<p>Reader was shown at CES in January 2000, and Microsoft's <a rel=\"external\" href=\"https://news.microsoft.com/2000/01/06/microsoft-announces-reader-with-cleartype-for-new-pocket-pcs/\">press release</a> called it the first product to include ClearType, a way of drawing text on LCD screens that came out of Microsoft Research. Later, Microsoft's own <a rel=\"external\" href=\"https://web.archive.org/web/20120106070110/http://www.microsoft.com/reader/info/support/faq/general.aspx\">FAQ</a> claimed ClearType improved resolution on LCD screens by up to 300 percent. After a few months preinstalled on Pocket PCs, Reader reached Windows as a free download that August, together with an add-in for Word 2000, “Read in Microsoft Reader”, that saved a document as a Reader file (<a rel=\"external\" href=\"https://news.microsoft.com/2000/08/08/microsoft-brings-ebooks-to-windows-users-with-the-free-availability-of-microsoft-reader/\">Microsoft, August 2000</a>). The same release names Barnes &amp; Noble.com as the first online bookshop to sell commercial titles for it. When Amazon opened its own ebook store that November with more than a thousand titles, its <a rel=\"external\" href=\"https://press.aboutamazon.com/2000/11/amazon-com-launches-e-books-store-offering-unique-and-exclusive-content-to-customers\">announcement</a> named only one format: Microsoft Reader.</p>\n<p>The file itself is a container from Microsoft's HTML Help family, a close relative of the .chm format of old Windows help files. The National Archives' PRONOM registry <a rel=\"external\" href=\"https://pronom.nationalarchives.gov.uk/fmt/867\">identifies a .lit file</a> by its first eight bytes, which spell <code>ITOLITLS</code>. According to the <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/LIT\">MobileRead wiki</a>, the text inside is OEBPS 1.0 markup (the Open eBook format that later grew into EPUB, see <a href=\"/blog/history-of-epub/\">the history of EPUB</a>) compressed with LZX, and the pictures are GIF, JPEG or PNG. The books were tiny by today's standards. Microsoft's general FAQ put a 300-page novel at about 250 KB.</p>\n<p>That FAQ also describes three levels of protection. Every .lit file was at least Sealed, which meant it was encrypted during conversion so that nobody could alter the text, while any copy of Reader could still open it. An Inscribed book was further encrypted by Microsoft's Digital Asset Server and showed details of the buyer, such as a name, on its cover page. Owner Exclusive was the copy-protected level: the server added an encrypted licence, and only a copy of Reader that the buyer had activated could open the book. Activation linked a Microsoft Passport account to the device, and the <a rel=\"external\" href=\"https://web.archive.org/web/20111228120708/http://www.microsoft.com/reader/info/support/faq/activation.aspx\">activation FAQ</a> allowed up to six activations per account, although a “What's new” note in the general FAQ still spoke of reading on four devices. Reformatting a hard drive or fully resetting a Pocket PC meant activating again, and that could count against the six.</p>\n<p>In October 2011 the Reader home page still advertised a <a rel=\"external\" href=\"https://web.archive.org/web/20111013110123/http://www.microsoft.com/reader/default.aspx\">marketplace of over 60,000 titles</a>, a few lines above a notice headed “Closing the Book”. Retailers would stop selling .lit books on 8 November 2011, and Reader would be discontinued on 30 August 2012, its download included. <a rel=\"external\" href=\"https://www.digitaltrends.com/computing/remember-microsoft-reader-now-you-can-forget-about-it/\">Digital Trends</a> reported the news on 16 August 2011. Microsoft's <a rel=\"external\" href=\"https://web.archive.org/web/20110927022927/http://www.microsoft.com/reader/info/support/faq/closing.aspx\">closing FAQ</a> promised that the licence on purchased books would not expire. It also said Microsoft would not supply a replacement copy of Reader to anyone who lost theirs after the cut-off, and that it would provide no way to move books to another reading app.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/move-highlights-between-reading-apps/","url":"https://readwyn.com/blog/move-highlights-between-reading-apps/","title":"Why your highlights don't move from one reading app to another","summary":"Each reading app pins a highlight to a position in its own copy of the book. Why exports carry the text but not the place, and how quote anchoring finds it again.","content_html":"<p>A highlight is stored as an address in one particular file, written in the notation of the app that made it. Another app has a different copy of the book and reads a different notation, so it can't follow the address. What you can carry over is the highlighted text, and the second app then has to find that passage again in its own copy.</p>\n<h2 id=\"each-app-writes-the-address-its-own-way\">Each app writes the address its own way</h2>\n<p>Amazon has used its own numbering from the start. The <a rel=\"external\" href=\"http://g-ecx.images-amazon.com/images/G/01/digital/fiona/general/Kindle_User_Guide.pdf\">user guide for the first Kindle</a>, from 2007, calls locations the digital answer to page numbers: pages change when you change the text size, but a location stays put. A bookmark there belongs to the first location on the page, and the list of notes and marks shows each one under its location number.</p>\n<p>A Kobo keeps highlights as rows of the <code>Bookmark</code> table in its database, and the form of a row depends on the file. A <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?t=358401\">MobileRead forum post</a> from someone writing an importer shows both forms. For a plain EPUB, the start of the highlight is a file path inside the book followed by a step path such as <code>point(/1/4/2/4/38/1:6)</code>. For a KEPUB it's <code>span#kobo\\.153\\.1</code>, the ID of a sentence span that exists only in the converted file. One highlight in one book would therefore be written two different ways, depending on whether the Kobo got the book as EPUB or as KEPUB (the difference between the two is covered in <a href=\"/blog/kepub-or-epub-for-kobo/\">KEPUB or EPUB for a Kobo</a>).</p>\n<p>KOReader ties highlights to where the file lives. Its <a rel=\"external\" href=\"https://koreader.rocks/user_guide/\">user guide</a> says it keeps progress, highlights, and notes in a folder named after the book with <code>.sdr</code> on the end, next to the book by default. Move or rename the book with any other file manager and, by the guide's own account, those notes are gone without a warning. A third option names the folder by a partial MD5 hash of the book's contents instead, which survives a move but not a change to the file itself.</p>\n<h2 id=\"why-a-position-breaks\">Why a position breaks</h2>\n<p>The EPUB standard has its own address for a point in a book. <a rel=\"external\" href=\"https://idpf.org/epub/linking/cfi/\">EPUB Canonical Fragment Identifiers 1.1</a>, which the IDPF published on 5 January 2017, walks the document tree in numbered steps: an even number for each child element, an odd number for each run of text between them, and a colon with a character offset at the end. A CFI is precise, but only for that exact file. If a paragraph is added before the highlight or the book is converted again, the step numbers point somewhere else.</p>\n<p>The specification allows for this: a CFI can carry an element's ID and a few characters of the text around the point in square brackets, and an app should use them to correct an address that no longer fits. If it can't derive a correction, the CFI counts as an invalid reference. The W3C's model for annotations is blunter about plain character offsets, describing them as <a rel=\"external\" href=\"https://www.w3.org/TR/annotation-model/#text-position-selector\">\"very brittle with regards to changes to the resource\"</a>.</p>\n<p>The same goes for two editions of a book, even in one app. A second edition is a different file, and so is a publisher's corrected copy, or the store's copy next to one you added yourself. Every stored position is relative to the file it was made in.</p>\n<h2 id=\"what-the-exports-give-you\">What the exports give you</h2>\n<p>An export is the usual way out, and most of the formats on offer are made for a person to read. <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/29991333812631-Export-annotations-from-your-books\">Kobo's help page</a> describes Export Annotations on Kobo.com, for books bought from Kobo or read through Kobo Plus, as a PDF, HTML, plain text, or Markdown file. It says the export may not work for every book because it depends on the author's approval, that markups can't be exported, and that a generated file is deleted after 30 days.</p>\n<p>Google Play Books writes each book's notes to its own document in Google Drive, in a folder called \"Play Books Notes\" unless you rename it, once you turn on \"Save notes, highlights, and bookmarks\" in the app's settings. <a rel=\"external\" href=\"https://support.google.com/googleplay/answer/3165868\">Google's help page</a> warns that highlighted text may be missing from the document when the publisher has turned off copying or when you pass its copy limit; the page number in the document then takes you back to the passage. It also says your own edits to that document are wiped the next time you add a note to the book.</p>\n<p>KOReader's export, under Export Highlights in the top menu, writes text, HTML, Markdown, JSON, or a file in the format of Kindle's My Clippings.txt, and can send the highlights to services such as Joplin and Readwise (<a rel=\"external\" href=\"https://koreader.rocks/user_guide/\">user guide</a>). On a Kindle, the clippings file and Amazon's online Notebook are the two ways out, and <a href=\"/blog/export-kindle-highlights/\">where your Kindle highlights are</a> covers both.</p>\n<p>None of these files is meant to be read back in by another app. The IDPF drafted a format for that in 2015, <a rel=\"external\" href=\"https://idpf.org/epub/oa/\">Open Annotation in EPUB</a>, which asked reading systems to import a conformant <code>annotations.json</code>, and it stayed a draft. The W3C has taken the problem up again: <a rel=\"external\" href=\"https://www.w3.org/TR/epub-anno-10/\">EPUB Annotations 1.0</a> was a Working Draft as of 15 October 2026. It asks an app to give each note more than one selector, to read them from the most precise to the broadest, and to fall back on the others when the most precise no longer fits the book's text. Where that file lives, and what a note in it may contain, is in the <a href=\"/blog/marginalia/\">post on marginalia</a>.</p>\n<h2 id=\"finding-the-passage-by-its-words\">Finding the passage by its words</h2>\n<p>An anchor that survives a change of file stores the words themselves. The <a rel=\"external\" href=\"https://www.w3.org/TR/annotation-model/\">W3C Web Annotation Data Model</a>, a Recommendation since 23 February 2017, calls this a TextQuoteSelector: the exact highlighted text, plus a prefix and a suffix taken from just before and after it, so that a phrase which appears twice in a chapter can still be told apart.</p>\n<p>Hypothesis, the web annotation tool, set out how it uses this in a <a rel=\"external\" href=\"https://web.hypothes.is/blog/fuzzy-anchoring/\">2013 post on fuzzy anchoring</a>. Each annotation keeps three selectors, an element path, character offsets, and the quote with 32 characters of context on either side, and they're tried in that order. If the page has changed, it searches for the prefix and suffix near where they used to be and checks the text between them against the quote within a set tolerance. Only after that does it search for the quote alone. The matching uses a modified version of Google's diff-match-patch library.</p>\n<p>Readwyn anchors its highlights on the same idea. Each one records the chapter file, the offsets, and 32 characters of text before and after the highlight. If the EPUB has changed, it searches the chapter for the three together and takes the match nearest the old offset. A highlight it can't find stays in the Notebook's list but doesn't show in the book.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/old-kindle-end-of-support/","url":"https://readwyn.com/blog/old-kindle-end-of-support/","title":"Why your old Kindle can't reach the store any more","summary":"Amazon cut the store from non-touch Kindles in 2022 and ended support for every Kindle from 2012 or earlier in May 2026, but books already on them still open.","content_html":"<p>If a Kindle Keyboard or a Kindle 4 says “The Kindle Store is no longer available on this device” when you open the store, the Kindle isn't faulty. That is the wording <a rel=\"external\" href=\"https://liliputing.com/amazon-is-pulling-the-plug-on-the-kindle-store-for-older-kindle-ereaders/\">Liliputing</a> recorded in 2022. Amazon switched the store off on its Kindles without a touchscreen on 17 August 2022, and on 20 May 2026 it ended support for every Kindle e-reader released in 2012 or earlier. Those devices can't buy, borrow or download books any more, but the books already on them still open, <a rel=\"external\" href=\"https://techcrunch.com/2026/04/08/amazon-to-end-support-for-older-kindle-devices-2012/\">TechCrunch</a> reported. (If you aren't sure which model you have, <a href=\"/blog/which-kindle-do-i-have/\">the model guide</a> tells them apart.)</p>\n<h2 id=\"the-2022-cut-off\">The 2022 cut-off</h2>\n<p>Owners got an email in May 2022 that named five models: the Kindle (2nd Gen) International, the Kindle DX International, the Kindle Keyboard, the Kindle (4th Gen) and the Kindle (5th Gen), <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2022/05/24/kindle-store-getting-disabled-on-kindles-without-touchscreens-soon/\">as The eBook Reader reported</a> at the time. The email described them as devices brought out more than ten years earlier, but the real dividing line was the touchscreen. The <a rel=\"external\" href=\"https://press.aboutamazon.com/2011/11/all-new-kindle-family-kindle-kindle-touch-kindle-touch-3g-and-kindle-fire-coming-to-over-16-000-retail-locations-across-the-u-s\">Kindle Touch</a>, on sale from late 2011, wasn't on the list and kept its store; the $69 Kindle <a rel=\"external\" href=\"https://press.aboutamazon.com/2012/9/introducing-the-new-kindle-paperwhite-the-most-advanced-e-reader-ever-constructed\">of 2012</a> lost it. The same email said that books bought on a phone or a computer would still arrive on the old Kindles, and that items in the archive could still be downloaded. Its sweetener was 30% off a new Kindle and $40 of ebook credit, valid until 5 July 2022.</p>\n<h2 id=\"the-2026-cut-off\">The 2026 cut-off</h2>\n<p>In April 2026 Amazon wrote to owners again. This time the rule was age, and the list in <a rel=\"external\" href=\"https://www.theregister.com/2026/04/08/amazon_kindle_support_discontinued/\">The Register's report</a> runs from the <a href=\"/blog/first-kindle-2007/\">first Kindle</a> and the Kindle 2 through the DX and DX Graphite, the Keyboard, the Kindle 4, the Touch and the Kindle 5 to the first Paperwhite. After 20 May they couldn't get new content from the Kindle Store.</p>\n<p>Amazon's tablets of the same years went too. <a rel=\"external\" href=\"https://hothardware.com/news/amazon-ends-support-older-kindles\">HotHardware</a> lists the first and second Kindle Fire and the 2012 Fire HD 7 and Fire HD 8.9, and gives the offer that came with the email: 20% off a new Kindle and a $20 ebook credit, to be spent by 20 June 2026. An Amazon spokesperson told <a rel=\"external\" href=\"https://techcrunch.com/2026/04/08/amazon-to-end-support-for-older-kindle-devices-2012/\">TechCrunch</a> that these models had been supported \"for at least 14 years\".</p>\n<h2 id=\"the-3g-networks-went-first\">The 3G networks went first</h2>\n<p>Some of these Kindles had lost their wireless years earlier. Amazon told owners that US carriers would start shutting down their 2G and 3G networks in December 2021, <a rel=\"external\" href=\"https://www.pcworld.com/article/3629405/which-amazon-kindles-will-lose-internet-in-december.html\">PCWorld reported</a>. The original Kindle of 2007, the Kindle 2 and the Kindle DX of 2009 had mobile data and no Wi-Fi, so they could no longer download anything over the air. The 3G editions from the Kindle Keyboard of 2010 to the first Kindle Oasis of 2016 fell back to Wi-Fi, and Amazon said its cellular Kindles released after 2016 use 4G and weren't affected. Amazon told <a rel=\"external\" href=\"https://www.androidpolice.com/2021/07/29/amazon-comes-clean-about-how-the-3g-shutdown-will-affect-older-kindles/\">Android Police</a> that owners of a first Kindle would get a free Kindle Oasis and cover, and that owners of a Kindle 2, a DX or a Keyboard got $70 off a new Paperwhite or Oasis and $25 of ebook credit, if their device had been in use between January and June 2021.</p>\n<h2 id=\"keeping-an-old-kindle-useful\">Keeping an old Kindle useful</h2>\n<p>An unsupported Kindle still works as a reader, as long as you leave its registration alone:</p>\n<ol>\n<li>Don't deregister it or reset it to factory settings. Amazon's email of 2026 warned that after either one it can't be registered again or used at all, <a rel=\"external\" href=\"https://www.theregister.com/2026/04/08/amazon_kindle_support_discontinued/\">The Register reported</a>.</li>\n<li>Copy your own files across by cable. <a rel=\"external\" href=\"https://liliputing.com/amazon-is-ending-support-for-kindle-devices-from-2012-and-earlier/\">Liliputing</a> notes that personal documents can still go over USB. For books bought in the Kindle Store, see <a href=\"/blog/kindle-download-and-transfer-removed/\">what happened to Download &amp; Transfer via USB</a>.</li>\n<li>Read the rest of your library elsewhere. TechCrunch says it stays open on newer Kindles, in the Kindle app and on Kindle for Web.</li>\n</ol>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/public-domain-day/","url":"https://readwyn.com/blog/public-domain-day/","title":"Public Domain Day: which books become free each 1 January","summary":"Which books entered the public domain on 1 January 2026 in the US and in life plus 70 countries, and which ones follow on 1 January 2027.","content_html":"<p>On 1 January 2026, books published in 1930 became public domain in the United States, among them <em>As I Lay Dying</em>, <em>The Maltese Falcon</em>, <em>Swallows and Amazons</em> and the first four Nancy Drew mysteries. On the same morning, in the UK, most of the EU and the other countries that protect a book for its author's life plus 70 years, everything by writers who died in 1955 became free. The next round is fixed by the same arithmetic: US books of 1931 and writers who died in 1956 follow on 1 January 2027.</p>\n<p>Every batch arrives on 1 January because copyright terms run to the end of a calendar year, not to the anniversary of a publication or a death. US law says so in <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/305\">17 U.S.C. § 305</a>. In the UK, section 12 of the Copyright, Designs and Patents Act 1988 ends the copyright in a book 70 years <a rel=\"external\" href=\"https://www.legislation.gov.uk/ukpga/1988/48/section/12\">from the end of the calendar year in which the author dies</a>.</p>\n<h2 id=\"the-two-clocks-side-by-side\">The two clocks side by side</h2>\n<p>The US counts older books from the year they came out: for a work published before 1978 whose copyright was <a href=\"/blog/when-do-books-enter-the-public-domain/\">renewed</a>, <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/304\">17 U.S.C. § 304</a> gives 95 years from the date it was first secured. Most of the rest of the world counts from the year the author died. The <a rel=\"external\" href=\"https://publicdomainreview.org/features/entering-the-public-domain/2026/\">Public Domain Review's page for 2026</a> gives the regions and the years of the two life-based rules; the US row comes from § 304:</p>\n<table><thead><tr><th>Rule</th><th>Where it applies</th><th>Freed on 1 January 2026</th><th>Due on 1 January 2027</th></tr></thead><tbody>\n<tr><td>95 years from publication</td><td>US books published before 1978</td><td>Books of 1930</td><td>Books of 1931</td></tr>\n<tr><td>Life plus 70 years</td><td>The UK, most of the EU, Russia, much of South America</td><td>Authors who died in 1955</td><td>Authors who died in 1956</td></tr>\n<tr><td>Life plus 50 years</td><td>Most of Africa and Asia</td><td>Authors who died in 1975</td><td>Authors who died in 1976</td></tr>\n</tbody></table>\n<p>Duke University's Center for the Study of the Public Domain publishes a list for the US each year. Its <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2026/\">2026 list</a> adds Agatha Christie's <em>The Murder at the Vicarage</em>, Dorothy L. Sayers's <em>Strong Poison</em>, Evelyn Waugh's <em>Vile Bodies</em>, Olaf Stapledon's <em>Last and First Men</em>, the original German of Freud's <em>Das Unbehagen in der Kultur</em> (<em>Civilization and Its Discontents</em>) and T. S. Eliot's <em>Ash Wednesday</em>. Only the 1930 text is free. Duke points out that a later version, such as an adaptation or a translation, may still be in copyright, though that copyright covers only what the later version added. A book on Duke's list whose author died after 1955 is still in copyright in Britain, which is how a book comes to be <a href=\"/blog/public-domain-differs-by-country/\">public domain in one country and not another</a>.</p>\n<p>Some free editions are ready on the morning itself. <a href=\"/blog/standard-ebooks/\">Standard Ebooks</a> marked the day with <a rel=\"external\" href=\"https://standardebooks.org/blog/public-domain-day-2026\">20 new ebooks</a>, among them Langston Hughes's <em>Not Without Laughter</em>, Edna Ferber's <em>Cimarron</em> and all four Nancy Drew titles, from <em>The Secret of the Old Clock</em> to <em>The Mystery at Lilac Inn</em>.</p>\n<h2 id=\"years-when-nothing-arrives\">Years when nothing arrives</h2>\n<p>The US had a long gap. A 1998 law added 20 years to the term of older works, and Duke's page for 2019 says that year was <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2019/\">the first in over 20 years</a> in which published works entered the US public domain. Canada is in one now: its term went from life plus 50 to life plus 70 at the end of 2022, so no author's books come free there before 1 January 2043. <a href=\"/blog/public-domain-differs-by-country/\">The post on how terms differ by country</a> gives the law, the date, and the authors the change left free.</p>\n<h2 id=\"writers-of-1955-and-1956\">Writers of 1955 and 1956</h2>\n<p>Thomas Mann <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Thomas_Mann\">died in Zürich on 12 August 1955</a>, so in life plus 70 countries his own German texts, from <em>Buddenbrooks</em> (1901) to <em>The Magic Mountain</em> (1924) and beyond, have been free since January. The English versions most people know are another matter. H. T. Lowe-Porter, who translated both, <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/H._T._Lowe-Porter\">died in 1963</a>, so her translations stay protected in those countries until the end of 2033. A. A. Milne <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/A._A._Milne\">died on 31 January 1956</a>, so his own writing comes out of copyright in Britain on 1 January 2027. Americans got <em>Winnie-the-Pooh</em> five years earlier: it was on Duke's <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2022/\">list of 1926 works</a> that became free in 2022. Duke is careful to say it was only the 1926 book that entered the US public domain, not the later Pooh stories.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/public-domain-differs-by-country/","url":"https://readwyn.com/blog/public-domain-differs-by-country/","title":"Why a book can be public domain in one country and not another","summary":"Why a book is public domain in the US but not the UK, or the other way round: each country sets its own copyright term and counts it from a different date.","content_html":"<p>W. Somerset Maugham's <em>Cakes and Ale</em> came out in 1930, so it entered the US public domain on <a href=\"/blog/public-domain-day/\">1 January 2026</a> with every other American book of that year. In Britain it stays in copyright until 2036. Maugham died in 1965, and the UK protects a written work for <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">70 years after the author's death</a>, which is why <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2026/\">Duke's Center for the Study of the Public Domain</a> picks the novel as its example of a book free in one country and not another. Each country sets its own term, and the terms don't start from the same event.</p>\n<p>The United States counts most older books from the year they came out, not from the author's death. According to the <a rel=\"external\" href=\"https://guides.library.cornell.edu/copyright/publicdomain\">copyright term chart of Cornell University Library</a>, everything published before 1931 is free, and a book published from 1931 to 1977 with a copyright notice is protected for 95 years after publication, as long as the copyright of a book from 1963 or earlier was <a href=\"/blog/when-do-books-enter-the-public-domain/\">renewed</a>. It cuts the other way too. A novel published in 1940 by an author who died in 1945 has been free in the UK since 2016, while a renewed US copyright on it runs 95 years, to the end of 2035.</p>\n<p>The same Duke page says that, as a general matter, courts have held you to the law of the place where you use the work. That's why the <a rel=\"external\" href=\"https://www.gutenberg.org/policy/license.html\">Project Gutenberg licence</a> tells anyone outside the US to check the laws of their own country, and says the foundation makes no claim about a book's status anywhere else. It's also how Germany lost the whole site from 2018 to 2021. A court in Frankfurt ruled for the publisher S. Fischer Verlag in February 2018 over 18 ebooks by Heinrich Mann, Thomas Mann and Alfred Döblin, and Project Gutenberg blocked every German visitor on 28 February 2018, as its <a rel=\"external\" href=\"https://cand.pglaf.org/germany/index.html\">own account of the case</a> records. The two sides settled in October 2021. The country-wide block ended, and the settlement keeps the three authors' books blocked in Germany until their German copyright runs out. Heinrich Mann's had already run out at the end of 2020, Thomas Mann's lasted to the end of 2025, and Döblin's lasts until the end of 2027.</p>\n<p>A country can also change its term. Canada protected books for the author's life plus 50 years until the 2022 budget law rewrote <a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/acts/C-42/section-6.html\">section 6 of its Copyright Act</a> to make it 70, and an <a rel=\"external\" href=\"https://orders-in-council.canada.ca/attachment.php?attach=42930\">order in council</a> brought the change into force on 30 December 2022. Section 280 of the <a rel=\"external\" href=\"https://laws-lois.justice.gc.ca/eng/AnnualStatutes/2022_10/page-21.html\">Budget Implementation Act, 2022, No. 1</a> says the new term doesn't revive a copyright that had already expired. Authors who died in 1971 or earlier stay free in Canada, Maugham among them (his books lost their Canadian copyright at the end of 2015). Someone who died in 1972 would have joined them two days after the change; their books now wait until 2043.</p>\n<p>Sites run under other laws fill some of the gaps. <a rel=\"external\" href=\"https://www.fadedpage.com/\">Faded Page</a> offers books that are public domain in Canada and says it follows Canadian copyright law. <a rel=\"external\" href=\"https://gutenberg.net.au/licence.html\">Project Gutenberg Australia</a> checks status under Australian law only and asks readers elsewhere to check their own before they share its files; its <a rel=\"external\" href=\"https://gutenberg.net.au/\">front page</a> says it stopped adding ebooks on 1 January 2025.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-in-a-foreign-language-on-an-e-reader/","url":"https://readwyn.com/blog/read-in-a-foreign-language-on-an-e-reader/","title":"Learning a language by reading ebooks","summary":"Reading books in a language you're learning works best when you know most of the words. What studies show on the 98% rule, repetition and dictionary use.","content_html":"<p>Reading books in the language you're learning does build vocabulary and reading skill, provided the books are already mostly easy for you and you read a lot of them. The <a rel=\"external\" href=\"https://erfoundation.org/guide/ERF_Guide.pdf\">Extensive Reading Foundation's guide for teachers</a> says that learners who know about 98% of the words on a page can read it quickly and with high comprehension, though a 2025 replication found no clear difference in comprehension between 90%, 95% and 98%.</p>\n<h2 id=\"you-need-to-know-98-of-the-words\">\"You need to know 98% of the words\"</h2>\n<p>The figure goes back to a 2000 paper by Marcella Hu and Paul Nation in <a rel=\"external\" href=\"https://nflrc.hawaii.edu/rfl/item/43\">Reading in a Foreign Language</a>. As a later replication describes it, they swapped the rarest words of an easy story for made-up ones at different rates and tested how well learners understood each version. Ten years later, Batia Laufer and Geke Ravenhorst-Kalovski <a rel=\"external\" href=\"https://nflrc.hawaii.edu/rfl/item/206\">tied coverage to vocabulary size</a>: 98% of a text took about 8,000 word families, and their minimum of 95% took 4,000 to 5,000. A word family is a base word with its forms, so \"read\", \"reads\" and \"reader\" count once.</p>\n<p>The line isn't sharp, though. In 2025 Stuart Webb, Ana Pellicer-Sánchez and Andi Wang <a rel=\"external\" href=\"https://nflrc.hawaii.edu/rfl/item/654\">repeated the test</a> with 94 advanced learners and Hu and Nation's own story. Readers at 90%, 95% and 98% coverage showed no statistically significant difference in comprehension; only those who knew every word scored significantly higher than the 90% and 95% groups. Coverage explained 11% of the variation in scores. The paper also notes that in the original study, three quarters of the readers at 90% coverage got at least 9 of the 14 questions right.</p>\n<h2 id=\"any-novel-in-the-language-will-do\">\"Any novel in the language will do\"</h2>\n<p>Not if the novel was written for native readers. Books like that, the foundation's guide (2011) says, are full of words that turn up only once or twice. The guide works with three bands. Below 90% known words, reading is slow and frustrating and comprehension suffers badly. Between 90% and 98% the text is \"instructional\": fine for learning, with a dictionary to hand. From 98% up, a reader can go fast enough to enjoy it, and the guide's target for fluent reading is 150 to 200 words a minute. Graded readers remove the rare words and repeat the useful ones, and a series may have six to eight levels, from Starter to Advanced.</p>\n<h2 id=\"you-pick-up-words-just-by-meeting-them\">\"You pick up words just by meeting them\"</h2>\n<p>You do, but slowly, and mainly the words that keep coming back. In one test, 15 Japanese learners read <a rel=\"external\" href=\"https://nflrc.hawaii.edu/rfl/item/78\">a 400-headword graded reader of A Little Princess</a> in which Rob Waring and Misako Takaki had respelt 25 words, so nobody could know them beforehand. Some appeared 15 to 18 times, others once. After three months the average reader remembered the meaning of about one of the 25, and of no word met fewer than eight times.</p>\n<p>Full-length novels are where a learner meets the rarer words that graded readers leave out. Pellicer-Sánchez and Norbert Schmitt had advanced learners read <a rel=\"external\" href=\"https://nflrc.hawaii.edu/rfl/item/207\">Things Fall Apart</a>, unabridged. For words met more than 10 times, readers recognised the meaning of 84% and could recall it for 55%. Their paper cites Nation's estimate that reading independently takes 8,000 to 9,000 word families, well beyond the <a rel=\"external\" href=\"https://erfoundation.org/guide/ERF_Guide.pdf\">foundation's table of graded reader series</a>, whose top band starts at 3,601 headwords and runs to 4,500 or more.</p>\n<h2 id=\"look-up-every-word-you-don-t-know\">\"Look up every word you don't know\"</h2>\n<p>The research that Caleb Prichard reviewed for his 2008 <a rel=\"external\" href=\"https://nflrc.hawaii.edu/rfl/item/177\">study of dictionary use</a> favours looking words up selectively, and his own results show that many learners don't. Among his Japanese university students reading English nonfiction, a third used the dictionary excessively, and a quarter of the words they checked were neither needed for the article's main points nor common enough to be worth learning.</p>\n<p>Never looking anything up isn't the answer either. When 135 Taiwanese students read an English passage with <a rel=\"external\" href=\"https://nflrc.hawaii.edu/rfl/item/196\">glosses in Chinese</a>, the glosses helped them learn new words, though their reading comprehension didn't improve significantly. The foundation's guide gives a plain test: if you're reaching for the dictionary all the time, the book is too hard.</p>\n<p>An e-reader makes the selective lookup cheap, and it keeps a record. Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/EM10/Kindle_User_Guide_en-US.pdf\">Kindle Paperwhite User's Guide</a> (2nd edition) says every word you look up goes into Vocabulary Builder automatically, where you can quiz yourself with flashcards; that guide turns it off under Settings › Reading Options › Language Learning. A Kobo has a My Words list, and if you also read on a Mac, Readwyn syncs the words you save in its reader <a href=\"/blog/kobo-my-words-on-mac/\">with it</a>. Both devices can look words up in a bilingual dictionary instead of a monolingual one; where that switch lives is in <a href=\"/blog/how-e-reader-dictionaries-work/\">how the dictionary on a Kindle or Kobo works</a>.</p>\n<h2 id=\"reading-for-fun-won-t-raise-your-level\">\"Reading for fun won't raise your level\"</h2>\n<p>It does, by a modest amount. A <a rel=\"external\" href=\"https://nflrc.hawaii.edu/rfl/item/354\">meta-analysis</a> by Eun-Young Jeon and Richard Day pooled 49 studies of extensive reading published from 1980 to 2014, with 5,919 participants, and found a small to medium effect on reading proficiency. The effect was larger for adults than for children and teenagers. It was also larger for learners of English as a foreign language than for those living where English is spoken.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/reading-on-an-e-reader-before-bed/","url":"https://readwyn.com/blog/reading-on-an-e-reader-before-bed/","title":"Does reading on an e-reader before bed hurt your sleep?","summary":"The sleep study that everyone quotes used an iPad. A second team measured the light of a front-lit Kindle: about a ninth of a tablet, and it peaks in the blue.","content_html":"<p>With its front light off, an e-ink reader sends no light of its own to your eyes. It reflects the lamp in the room, as a paper page does. With the light on, it becomes a light source, and when researchers measured a front-lit Kindle, its light peaked in the blue. Nobody in the studies below measured sleep on an e-reader, though: the sleep study that people quote used an iPad.</p>\n<h2 id=\"the-ipad-study\">The iPad study</h2>\n<p>At Brigham and Women's Hospital in Boston, 12 young adults read for about four hours before bed on five evenings in a row from an iPad, and on another five from printed books, in a room lit to about 3 lux. After the iPad evenings they took nearly 10 minutes longer to fall asleep, their melatonin was suppressed by about 55%, its evening rise came about 90 minutes later, and they felt sleepier the next morning (<a rel=\"external\" href=\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4313820/\">Chang and colleagues, PNAS, 2015</a>).</p>\n<p>In Boston the iPad sat on a stand 30 to 45 cm from the reader at full brightness, and staff moved it to keep 30 to 50 lux at the eye; the printed page reflected less than 1 lux. The authors name their own limits: the long sessions, the maximum brightness, and the lack of a light-emitting screen with a warmer colour, so they couldn't say whether the colour or the amount of light caused the effect.</p>\n<h2 id=\"how-much-light-a-kindle-gives\">How much light a Kindle gives</h2>\n<p>In a paper published in October 2015, a team from King's College London and the University of Surrey measured a first-generation Kindle Paperwhite in Surrey's chronobiology lab (<a rel=\"external\" href=\"https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2015.00233/full\">Gringras and colleagues, Frontiers in Public Health, 2015</a>). They worked in a completely dark room, at 35 cm and half brightness, a night-time level chosen after asking ten Paperwhite owners what they used. Its light peaked at 455 nm, in the blue, much like an iPad Air and an iPhone 5s showing the same page. The Kindle gave about 35 melanopic lux, a measure weighted for the receptor that drives the melatonin response, against 302 for the iPad Air at its automatic brightness. That is roughly a ninth of the tablet's figure, and close to the <a rel=\"external\" href=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC4313820/\">31 melanopic lux that the Boston team measured</a> from their iPad. For contrast, the authors describe the light of a paper book under a dim bedside lamp as yellow to red, with little impact on sleep. The Surrey team measured the light only; nobody slept in their study.</p>\n<p>E Ink says its front lights <a rel=\"external\" href=\"https://www.eink.com/tech/detail/Front_Light\">don't shine directly into your eyes</a>: the light runs down the face of the display and bounces back off the ink (<a href=\"/blog/front-light-and-warm-light/\">how a front light works</a>). The Surrey figures show that some of it still reaches the eye.</p>\n<h2 id=\"the-advice\">The advice</h2>\n<p>Professional advice covers screens in general and doesn't single out e-ink. The American Academy of Ophthalmology says blue light affects the body's circadian rhythm and <a rel=\"external\" href=\"https://www.aao.org/eye-health/tips-prevention/should-you-be-worried-about-blue-light\">suggests limiting screen time</a> for two to three hours before bed. Harvard Health gives the same window for <a rel=\"external\" href=\"https://www.health.harvard.edu/staying-healthy/blue-light-has-a-dark-side\">bright screens</a>.</p>\n<p>A warm front light came too late for either study. The Kindle Oasis of June 2019 was <a rel=\"external\" href=\"https://www.androidcentral.com/amazon-kindle-oasis-2019-adjustable-color-temperature\">the first Kindle with one</a>, and when Engadget <a rel=\"external\" href=\"https://www.engadget.com/2019/06/19/the-new-kindle-oasis-adjusts-its-screen-color-to-the-time-of-day/\">looked at it that month</a>, no other Kindle could change the colour of its light. Kobo's ComfortLight PRO moves the screen towards orange through the evening, and it can follow a time that you set next to “Bedtime” under the brightness icon on the Home screen. Kobo's help page says this “<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017481174\">can help you fall asleep more easily</a>”.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/reading-with-low-vision-on-an-e-reader/","url":"https://readwyn.com/blog/reading-with-low-vision-on-an-e-reader/","title":"Reading with low vision on an e-reader","summary":"Which Kindle and Kobo settings help with low vision (bigger and bolder text, white on black, larger menus) and when each one arrived, from 2009 to 2025.","content_html":"<p>An e-reader helps most inside a book. A Kindle or a Kobo lets you make the text much larger than the default, make it heavier, and turn the page to white text on black, and those are the settings to look for. The rest of the device has been slower to follow: menus and the library got larger type years after book text did, and four Kobo models gained a screen reader only in July 2025, and only in the EU.</p>\n<p>The RNIB's <a rel=\"external\" href=\"https://media.rnib.org.uk/documents/Confident_Living_-_Reading_2022.pdf\">guide to reading with sight loss</a> names Kindle and Kobo as e-book readers and gives their advantage as control over the size, shape, colour, and contrast of the text. For print, it treats 16 to 18 point as large print and anything above that as giant print, and its own giant-print books are 24 point. It suggests a local library as a place to find out which size suits you.</p>\n<h2 id=\"the-first-big-font-2009-and-2010\">The first big font, 2009 and 2010</h2>\n<p>In June 2009 the National Federation of the Blind and the American Council of the Blind sued Arizona State University over a pilot that put textbooks on the Kindle DX. The DX could read a book aloud, but its menus were silent, so a blind student couldn't buy a book, open it, or switch the speech on without help, as <a rel=\"external\" href=\"https://nfb.org/node/5058\">the NFB described it in July 2010</a>.</p>\n<p>On 7 December 2009 Amazon <a rel=\"external\" href=\"https://press.aboutamazon.com/2009/12/blind-and-vision-impaired-readers-to-benefit-from-new-kindle-features-in-2010\">announced two changes for the next year</a>: menus that speak, and a seventh font size, \"super size\", twice the height and width of the largest of the Kindle's six sizes at the time. The lawsuit <a rel=\"external\" href=\"https://nfb.org/node/5014\">settled on 11 January 2010</a>. The university denied any breach of the law and agreed that if it used e-readers in classes over the next two years, it would strive to choose ones that blind students could use. In late July 2010 the NFB <a rel=\"external\" href=\"https://nfb.org/node/5058\">welcomed a new Kindle</a> whose voice guide read every menu option aloud.</p>\n<h2 id=\"bold-text-and-white-on-black-2017-to-2021\">Bold text and white on black, 2017 to 2021</h2>\n<p>The next large step came with the all-new Kindle Oasis of 2017. Amazon's <a rel=\"external\" href=\"https://press.aboutamazon.com/uk/2017/10/meet-the-all-new-kindle-oasis-featuring-a-7-inch-300-ppi-display-waterproof-design-and-audible\">press release of 11 October 2017</a> promised more font sizes, five levels of boldness for any font, an option to invert black and white for readers with light sensitivity, and an enlarged display mode for the text of the home screen and the library. A <a rel=\"external\" href=\"https://www.aboutamazon.com/introducing-new-accessibility-features-on-kindle-and-fire\">later Amazon post</a> said the inversion covered the whole interface as well as the book, and named the new Oasis alone for it and for the larger display. The bold options also went to the Kindle apps for iOS, Android, and Fire OS.</p>\n<p>Kobo's version of the bigger interface was Large Print Mode. In February 2018 it sat under Beta Features in the settings and <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2018/02/11/large-print-mode-on-kobo-ebook-readers-makes-ui-text-larger/\">enlarged the menus</a>, the headers and footers, the home screen, the library, and the dictionary, while leaving the font size inside books alone. Firmware 4.19.14123, in December 2019, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2019/12/12/new-kobo-firmware-update-4-19-14123-released/\">gave it a slider</a> for the scale.</p>\n<p>White text on black came to Kobo later. In June 2020 its dark mode was still a hidden, experimental setting that The eBook Reader found <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/?p=21174\">not officially supported</a> and prone to more ghosting than the normal page. Kobo's <a rel=\"external\" href=\"https://www.media-outreach.com/news/singapore/2021/10/05/97469/best-of-both-worlds-rakuten-kobo-introduces-kobo-sage-and-kobo-libra-2/\">press release for the Sage and the Libra 2</a>, dated 5 October 2021, listed a Dark Mode for white text on black. As of November 2026, <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360062231213\">Kobo's help page</a> names eight models that have it: the Sage, Libra 2, Elipsa, Clara 2E, Elipsa 2E, Libra Colour, Clara Colour, and Clara BW. While reading, tap the middle of the page, tap the gear icon, and tick “Dark Mode” under “Page appearance”. Pictures in the book keep their original colours. What the research says about light and dark pages is in <a href=\"/blog/dark-mode-for-reading-research/\">dark mode for reading</a>.</p>\n<h2 id=\"screen-readers-and-system-text-2025-and-2026\">Screen readers and system text, 2025 and 2026</h2>\n<p>Since 29 July 2025, readers in the European Union with a Clara BW, Clara Colour, Libra Colour, or Elipsa 2E can install <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/32246787707799-Kobo-eReader-Accessibility-Features-Support\">an optional update</a> that adds Settings › Accessibility › “Display &amp; Text Size”, which enlarges text outside books, and a screen reader. The update followed the European Accessibility Act, which has applied since 28 June 2025 (<a href=\"/blog/european-accessibility-act-ebooks/\">what that law asks of ebooks</a>). The post on <a href=\"/blog/text-to-speech-on-e-readers/\">text-to-speech on e-readers</a> covers what the screen reader needs and what installing the update resets. Kobo's page, as of version 5.18, also lists what the accessibility software can't do yet. For low vision the item that matters is pinch-to-resize text.</p>\n<p>On a Kindle, you change the size of the book text by tapping the top of a page, then Aa, or by pinching the page, as <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=GFXBHXY4YXFNFPUE\">Amazon's help page</a> shows. An 11th generation or later Kindle, a Scribe, or a Colorsoft also has a font size for the whole device, under Settings › Screen and brightness › Display &amp; text size. Amazon's <a rel=\"external\" href=\"https://www.amazon.com/b?ie=UTF8&amp;node=21346183011\">accessibility page</a> in November 2026 lists VoiceView, a screen reader that you hear through a Bluetooth audio device, and a large display option for the main parts of the interface. It lists dark mode too, with one exception among the e-readers: the Kindle Colorsoft doesn't support it.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/screen-or-paper-comprehension/","url":"https://readwyn.com/blog/screen-or-paper-comprehension/","title":"Do you understand less when you read on a screen?","summary":"Two reviews of the research put paper a little ahead, mostly for texts that explain or argue, read against the clock. A novel at your own pace reads much the same, even on a Kindle DX.","content_html":"<p>Skim a long report on your laptop with a meeting in ten minutes, and you'll probably take in a little less than you would from the same pages printed. Read a novel at your own pace, and the screen makes almost no measurable difference. Two large reviews of the research show that pattern. The gap is small, about a fifth to a quarter of a standard deviation, and it sits mostly in texts that explain or argue, read against the clock.</p>\n<h2 id=\"the-two-reviews\">The two reviews</h2>\n<p>The larger of the two reviews is by <a rel=\"external\" href=\"https://doi.org/10.1016/j.edurev.2018.09.003\">Delgado, Vargas, Ackerman and Salmerón (2018)</a> in <em>Educational Research Review</em>. They gathered studies from 2000 to 2017 that gave people the same text on paper or on a digital device: 38 compared two groups, 16 tested the same people on both, and together they covered 171,055 participants. Paper came out ahead in both kinds of design, with an effect size (Hedges' g) of 0.21. The gap grew when reading was timed rather than self-paced. It held for informational texts and for mixes of informational and narrative ones, but not for studies that used narrative texts alone. And it was larger in the later studies than in the earlier ones.</p>\n<p><a rel=\"external\" href=\"https://doi.org/10.1111/1467-9817.12269\">Virginia Clinton's 2019 review</a> in the <em>Journal of Research in Reading</em> was stricter. She kept only experiments from 2008 to 2018 that assigned readers to paper or screen at random: 33 studies with 2,799 people. The pooled effect was close to Delgado's (0.25 in favour of paper), and the split by genre was sharper. Expository texts gave 0.32; narrative texts gave 0.04, which is next to nothing. Screen readers weren't faster, either: across 14 studies, reading time came out about the same.</p>\n<h2 id=\"a-novel-on-an-e-ink-reader\">A novel on an e-ink reader</h2>\n<p>An e-ink screen behaves differently from a monitor in other ways (the <a href=\"/blog/is-e-ink-better-for-your-eyes/\">post on e-ink and your eyes</a> covers the light), so one experiment with a real e-reader is worth a close look, and it fits Clinton's narrative figure. <a rel=\"external\" href=\"https://doi.org/10.3389/fpsyg.2019.00038\">Mangen, Olivier and Velay (2019)</a> had 50 young adults read a 28-page Elizabeth George mystery of about 10,800 words, half on a Kindle DX and half in a printed pocket book laid out from the same PDF. The two groups took the same time (about 58 minutes), recognised words and sentences from the text equally well, remembered the characters and places equally well, and felt equally absorbed in the story.</p>\n<p>Where they parted was time and order. Asked when things happened, the Kindle readers got 44% right against 57% for the print readers, and they made more mistakes when asked to put 14 events of the plot in sequence. The authors suggest that the thickness of a book and the shift of pages from one hand to the other help a reader place events, while an e-reader gives mostly visual cues. They call the work largely exploratory. Their set-up also took away some of the help an e-reader normally gives: page numbers were removed from both versions and the Kindle's progress bar was hidden, and only two of the 50 readers were experienced Kindle users.</p>\n<h2 id=\"why-the-clock-matters\">Why the clock matters</h2>\n<p>The timed reading in Delgado's review is where the screen loses most, and two experiments explain some of that. <a rel=\"external\" href=\"https://doi.org/10.1037/a0022086\">Ackerman and Goldsmith (2011)</a> fixed the study time in one experiment and let people choose it in a second. With the time fixed, people did as well on screen as on paper. Once they chose how long to study, the screen group scored lower, predicted their own results less accurately and regulated their time more erratically. The authors put the difference down to metacognition, how people judged and paced their own study, rather than to how well they processed the text. Clinton later pooled this kind of self-prediction (calibration) across 11 studies and found it better on paper too.</p>\n<p><a rel=\"external\" href=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC7463273/\">Delgado and Salmerón</a> then tested the clock directly in a study published in 2020. Undergraduates at a Spanish university read a 3,010-word magazine article, either in print or as a PDF on a 17 inch desktop monitor, with free time or with 16 minutes 30 seconds. With free time, the two groups scored about the same on a 14-question test (67% in print, 68% on screen). Under the limit, print readers slipped to 64% while screen readers fell to 50%. Calibration didn't differ between the groups this time, but attention did: probes during the reading showed that print readers' minds wandered less when they were hurried, and screen readers' minds didn't.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/send-epub-to-kindle/","url":"https://readwyn.com/blog/send-epub-to-kindle/","title":"Can you send an EPUB to a Kindle now?","summary":"Send to Kindle has accepted EPUB files since 2022 and converts them on the way; here are the dates, the size limits of email, the web page and the Windows app, and why a file can fail.","content_html":"<p>Yes. Amazon's Send to Kindle service has accepted EPUB files since 2022, by email, from its web page, and from its desktop apps. Amazon converts each EPUB into a Kindle format on the way, so the book on your Kindle is no longer an EPUB, and the file has to be free of DRM.</p>\n<h2 id=\"when-it-changed\">When it changed</h2>\n<p>Before that, the service took MOBI and refused EPUB. On 29 April 2022, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2022/04/29/kindle-ereaders-now-support-epubs-sent-by-email/\">The eBook Reader passed on a detail</a> that had turned up on MobileRead: EPUB was now on Amazon's list of supported personal document types. Its test file, a DRM-free EPUB emailed to a Paperwhite, showed up on the device as an AZW3. Amazon's email to customers that spring said MOBI \"won't support the newest Kindle features for documents\", and set August 2022 as the end for sending .mobi and .azw files (<a rel=\"external\" href=\"https://www.fivefilters.org/?p=4148\">FiveFilters quotes it</a>). The Send to Kindle apps for Mac and Windows took EPUB from <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2022/12/20/send-to-kindle-apps-for-pc-and-mac-now-support-epub/\">20 December 2022</a>.</p>\n<p>The MOBI deadline slipped. A later email, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2023/09/25/send-to-kindle-losing-mobi-support-probably-for-real-this-time\">reported in September 2023</a>, gave the final dates: the wind-down began on 1 November 2023, and Send to Kindle dropped .mobi, .azw and .prc files completely by 20 December 2023, by email, in the Kindle apps for iOS and Android, and in the desktop apps. MOBI documents already in a library stayed where they were, and a Kindle still opens a MOBI file copied to it over USB. A shelf of old MOBI or AZW3 files can still go through Send to Kindle once they're EPUBs; <a href=\"/blog/convert-mobi-to-epub-on-mac/\">converting MOBI to EPUB on a Mac</a> covers that step.</p>\n<h2 id=\"the-routes-and-their-limits\">The routes and their limits</h2>\n<table><thead><tr><th>Route</th><th>Limit</th><th>Where the figure comes from</th></tr></thead><tbody>\n<tr><td>Email to your Send to Kindle address</td><td>25 attachments and 50 MB in total per email</td><td><a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G7NECT4B4ZWHQ8WV\">Amazon's email help page</a></td></tr>\n<tr><td>The web page, amazon.com/sendtokindle</td><td>200 MB per file</td><td><a rel=\"external\" href=\"https://www.amazon.com/sendtokindle\">The page itself</a></td></tr>\n<tr><td>Send to Kindle for Windows</td><td>50 MB per document in December 2022</td><td><a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2022/12/20/send-to-kindle-apps-for-pc-and-mac-now-support-epub/\">The eBook Reader's report</a></td></tr>\n</tbody></table>\n<p>The email route has two more rules. You can only send from addresses on your Approved Personal Document Email List, which holds up to 15 addresses, and Amazon keeps trying to deliver a document for up to 60 days. A ZIP file gets around the 50 MB cap: Amazon's conversion service opens the archive and converts what's inside. As of November 2026 the web page lists EPUB beside PDF, DOC, DOCX, TXT, RTF, HTML and the common image formats; MOBI and AZW3 aren't on it.</p>\n<h2 id=\"what-can-stop-an-epub\">What can stop an EPUB</h2>\n<p>DRM stops it outright. An EPUB bought from another shop and locked with Adobe's DRM <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2022/04/29/kindle-ereaders-now-support-epubs-sent-by-email/\">won't go through</a>, and Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TWXpUGw76dtEg2VD9P\">conversion tips</a> name password protection and security settings as causes of a failed conversion. If the download from your shop was a small .acsm file rather than a book, <a href=\"/blog/what-is-an-acsm-file/\">that file is a licence ticket</a>, not an EPUB at all.</p>\n<p>Those tips also list the languages the conversion accepts. The list starts at Afrikaans, takes in Arabic, Russian and Chinese (Simplified), and ends with five Indic languages, but in November 2026 it has no Japanese or Korean. It also asks you to remove large tables and tables with complex formatting. When a file fails for no reason you can see, <a href=\"/blog/check-epub-with-epubcheck/\">running EPUBCheck on it</a> shows whether the EPUB itself is broken.</p>\n<h2 id=\"what-arrives-on-the-kindle\">What arrives on the Kindle</h2>\n<p>The converted file is a personal document, not a store purchase. On a Kindle e-reader you can highlight it and add typed notes; Amazon's <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TNtSm13k3txFI4EvDV\">table of note types</a> says Kindle Scribe adds handwritten notes, but not in EPUB comics, manga or children's books. The same table says Kindle for Web doesn't open personal documents yet. Documents sent before 11 November 2022 don't get the newer Kindle features, and <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TWXpUGw76dtEg2VD9P\">Amazon's advice</a> is to send them again through Send to Kindle.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/serif-or-sans-serif-for-reading/","url":"https://readwyn.com/blog/serif-or-sans-serif-for-reading/","title":"Serif or sans serif: which is easier to read on a screen?","summary":"Studies find little or no reading-speed difference between serif and sans serif fonts on screens; familiarity and the individual reader matter more.","content_html":"<p>Open the font menu of almost any reading app and you'll find both kinds: faces with small finishing strokes at the ends of the letters, such as Georgia or Literata, and faces without them, such as Helvetica or Avenir. Choose the one you find comfortable. Controlled experiments have found little or no difference in reading speed between serif and sans serif type, and the old case against serifs on screens was made for monitors that drew each letter with a handful of pixels.</p>\n<h2 id=\"what-the-experiments-found\">What the experiments found</h2>\n<p>Comparing one whole typeface with another, say Times with Arial, tells you little about serifs, because the two also differ in letter width, stroke contrast and spacing. Aries Arditi and Jianna Cho, at Lighthouse International in New York, avoided that by building lower-case fonts that were identical apart from the serifs: none, serifs 5% of the cap height long, and serifs 10% long. They measured the smallest size at which people could read each font, and their reading speed both in ordinary text and with words flashed one at a time. Serifs changed neither speed. The 5% font was slightly more legible at the smallest sizes, a gain the authors suspected came from the extra space that serifs push between letters, and they concluded that fonts differing only in serifs <a rel=\"external\" href=\"https://pubmed.ncbi.nlm.nih.gov/16099015/\">showed no difference in legibility</a> (Vision Research, November 2005).</p>\n<p>The font you read fastest in is a personal matter. In a 2022 paper in ACM Transactions on Computer-Human Interaction, Shaun Wallace and ten co-authors compared fonts on hundreds of crowdsourced readers, with sizes matched by how big the text looks rather than by point size. On average, each reader was <a rel=\"external\" href=\"https://doi.org/10.1145/3502222\">35% faster in their own fastest font than in their slowest</a>, with comprehension unchanged. Readers' fastest fonts weren't the same, which is why the authors argued that no single font suits everyone.</p>\n<p>Familiarity matters as well. Sofie Beier and Kevin Larson had people read fonts with common and with unusual letter shapes, tested them, let them read a story in the font for 20 minutes, and tested them again. The <a rel=\"external\" href=\"https://doi.org/10.1075/idj.20.1.02bei\">20 minutes of exposure</a> had an effect on reading speed and the unusual shapes did not, though readers disliked them (Information Design Journal, 2013). So for speed, what counted was the time spent reading a face, not how ordinary its letters looked, and the style you read most easily may be the one you read most.</p>\n<p>Since the best face varies from person to person, read a chapter in each and notice which one you stop seeing. (Readwyn's Text panel offers four serifs and two sans serifs, and the choice applies to every book on that Mac.) Two well-known faces drawn for reading on screens happen to be serifs: <a href=\"/blog/literata-font/\">Literata</a>, which TypeTogether made for Google Play Books, and <a href=\"/blog/bookerly-kindle-font/\">Bookerly</a>, the Kindle's default.</p>\n<h2 id=\"where-the-rule-about-screens-came-from\">Where the rule about screens came from</h2>\n<p>The idea that screens need sans serif comes from monitors whose letters were only a few pixels tall. Microsoft's notes on Verdana, which Matthew Carter designed for screen display and which was first released on 8 July 1996, explain the problem: with so few pixels, letters set close together collide, and <a rel=\"external\" href=\"https://learn.microsoft.com/en-us/typography/font-list/verdana\">an \"r\" and an \"n\" can merge into an \"m\"</a>. Verdana's answer was wide letters and generous spacing.</p>\n<p>Yet Carter also drew a serif face for the same screens in the same year. Microsoft describes <a rel=\"external\" href=\"https://learn.microsoft.com/en-us/typography/font-list/georgia\">Georgia</a>, released in September 1996, as made for clarity at low resolutions: compared with the Scotch Roman types it descends from, it has lighter capitals and a larger x-height. Microsoft added both to its Core fonts for the Web that year, Verdana in July and Georgia on 1 November.</p>\n<p>Pixel density has since climbed well past those monitors. In September 2022 Amazon moved its cheapest Kindle <a rel=\"external\" href=\"https://www.engadget.com/amazon-kindle-2022-upgrade-300-ppi-display-usb-c-16-gb-storage-130030447.html\">from 167 to 300 pixels per inch</a>. Type set at 10 point has an em of 10/72 of an inch, top to bottom: about 23 pixels at 167 ppi and about 42 at 300 ppi, so every stroke and serif gets nearly twice as many pixels in each direction (<a href=\"/blog/e-reader-pixels-per-inch/\">more on pixels per inch</a>).</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/social-drm-watermarks/","url":"https://readwyn.com/blog/social-drm-watermarks/","title":"What is social DRM?","summary":"Social DRM marks an ebook with the buyer's name or an order number instead of locking it: what the mark holds, who uses it, and why it doesn't stop copying.","content_html":"<p>Open some DRM-free ebooks and you'll find your own name inside: on a bookplate near the front, in a line at the end of each chapter, or on a notice at the back. That is social DRM. Instead of encrypting the file, the shop marks every copy with details of the person who bought it, in plain sight, hidden in the file, or both, so that a copy found on a file-sharing site leads back to one order. Nothing stops you copying the book or reading it in any app on any device. The mark relies on people not wanting to pass round a file with their name in it, which is the reasoning <a rel=\"external\" href=\"https://www.booxtream.com/\">BooXtream</a>, one of the companies that does the marking, gives on its home page.</p>\n<h2 id=\"harry-potter-2011-and-2012\">Harry Potter, 2011 and 2012</h2>\n<p>J.K. Rowling kept the digital rights to her books and for years held out against ebook editions. On 23 June 2011 she announced Pottermore, a website that would sell the seven Harry Potter ebooks itself from that October, as <a rel=\"external\" href=\"https://archive.nytimes.com/artsbeat.blogs.nytimes.com/2011/06/23/rowling-releases-harry-potter-into-the-ether-on-pottermore/\">The New York Times reported that day</a>. In the <a rel=\"external\" href=\"https://www.theguardian.com/books/booksblog/2011/jun/23/pottermore-jk-rowling-marketing-genius-harry-potter\">Guardian</a>, Sam Jordison picked out the plan for copy protection: no DRM software, but a digital watermark tying each ebook to the identity of its buyer.</p>\n<p>October came and went. The ebooks finally <a rel=\"external\" href=\"https://www.theguardian.com/books/2012/mar/27/pottermore-harry-potter-ebooks\">went on sale on 27 March 2012</a>, at £4.99 for the first three books and £6.99 for the last four. They worked on the Kindle, the Nook, Sony's Reader and Google Play, but Pottermore was the only place to buy them. The next day Philip Jones of The Bookseller <a rel=\"external\" href=\"https://www.theguardian.com/books/booksblog/2012/mar/28/pottermore-ebook-amazon-harry-potter\">set out in the Guardian</a> what the missing DRM meant for buyers: a title could go onto eight devices, or you could take the plain EPUB and load it onto a Kobo. Amazon listed the books, yet its buy link sent Kindle owners to Pottermore to pay, and Jones called that perhaps the biggest climbdown in Amazon's history.</p>\n<h2 id=\"what-the-mark-holds\">What the mark holds</h2>\n<p>In 2012 watermarking was already a service that could be bought. That July the Firsty Group partnered with the Dutch company Icontact for the right to offer its watermarking system, BooXtream, in the UK, as the news list on <a rel=\"external\" href=\"https://www.booxtream.com/\">BooXtream's site</a> records. The same site, headed \"Social DRM To The Max\", spells out how it works. For each sale the shop sends three things: the customer's name, their email address and the order number. From those, BooXtream adds visible text, all of it optional: a bookplate image with the customer's name on page 2, a personal footer at the end of every chapter, and a disclaimer on the last page with its own entry in the table of contents. It also writes several invisible watermarks into every data file of the EPUB (the <a href=\"/blog/what-is-inside-an-epub-file/\">post on what's inside an EPUB</a> lists those files), encoding the publisher, the customer and the shop. For Kindle owners it turns the marked EPUB into a Mobi file.</p>\n<h2 id=\"publishers-drop-adobe-drm-2015-onwards\">Publishers drop Adobe DRM, 2015 onwards</h2>\n<p>On 22 June 2015 <a rel=\"external\" href=\"https://www.heise.de/news/Verlagsgruppe-Bonnier-gibt-hartes-DRM-in-E-Books-auf-2720431.html\">heise online reported</a> that the German publishers of the Swedish Bonnier group, Piper, Ullstein and Carlsen among them, would replace Adobe DRM with watermarks in their EPUBs on 1 July 2015, and called Bonnier the first large publishing house to make the switch. Adobe's DRM, the kind that usually arrives as an <a href=\"/blog/what-is-an-acsm-file/\">ACSM file</a>, lets a book open only on devices signed in with the reader's Adobe ID. Bonnier said it was too complicated for readers and that its own tests had shown that even hard copy protection was easy to get round. The number of illegal downloads, it argued, depended on how much people wanted a title, not on how hard the protection was. Its books would carry an invisible watermark that traced back to the owner, and a page of warnings for what the company called an extra psychological effect. A spokeswoman told heise the next day that the whole list would change, older books too, unless an author or agent objected, and that the watermark was an encrypted transaction number. Bonnier also promised to chase pirated copies online harder than before.</p>\n<p>The help pages of the publisher <a rel=\"external\" href=\"https://help.versobooks.com/en-US/how-can-i-buy-verso-ebooks-193664\">Verso</a> said in November 2026 that books bought on its own website are watermarked and DRM-free: you may put them on any of your devices, but not upload them to websites or file-sharing networks. The same Verso books bought from Amazon, Apple, Barnes &amp; Noble or Google come with those stores' DRM. The page still names Sony too, although Sony's Reader Store <a rel=\"external\" href=\"https://www.shelf-awareness.com/theshelf/2014-02-07/sony_s_reader_store_customers_heading_to_kobo_ebookstore.html\">closed in the US and Canada in March 2014</a> and handed its customers to Kobo. Defective by Design's <a rel=\"external\" href=\"https://www.defectivebydesign.org/guide/ebooks\">guide to DRM-free ebook sellers</a> notes that the Finnish bookseller Suomalainen Kirjakauppa puts a watermark on the first page of each ebook in place of DRM. More shops of this kind are in the <a href=\"/blog/where-to-buy-drm-free-ebooks/\">list of DRM-free ebook sellers</a>.</p>\n<p>On the reading side, a watermarked book is an ordinary EPUB. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">file formats page</a> says its e-readers support both Adobe DRM and social DRM, and that a social DRM file opens without Adobe Digital Editions, which an Adobe-protected book needs before it can reach the e-reader.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/standard-ebooks/","url":"https://readwyn.com/blog/standard-ebooks/","title":"What is Standard Ebooks, and how do its books differ from Gutenberg's?","summary":"Standard Ebooks remakes public-domain texts, often from Project Gutenberg, with proofreading, modern spelling, real typography, painted covers and a CC0 dedication.","content_html":"<p>Standard Ebooks is a volunteer project that makes new editions of public-domain books. Put the same Victorian novel from its site and from Gutenberg on an e-reader and the differences show early. The <a rel=\"external\" href=\"https://standardebooks.org/ebooks/george-eliot/middlemarch/text/single-page\">Standard Ebooks copy</a> of <em>Middlemarch</em> has a painted cover and spells “tonight” where the printed book, and <a rel=\"external\" href=\"https://www.gutenberg.org/cache/epub/145/pg145.txt\">Gutenberg's transcription</a> of it, have “to-night”.</p>\n<p>The project is run as a low-profit limited liability company (L3C), with Alex Cabal as editor-in-chief and <a rel=\"external\" href=\"https://standardebooks.org/about\">donations from readers</a> as its only income. Many of its texts start life as <a href=\"/blog/what-is-project-gutenberg/\">Project Gutenberg</a> transcriptions. Gutenberg keeps the archive; Standard Ebooks proofreads a text against scans of a printed edition, sets it to a house style and rebuilds the files.</p>\n<p>Its <a rel=\"external\" href=\"https://standardebooks.org/about/what-makes-standard-ebooks-different\">comparison page</a> says that Gutenberg's EPUB and Kindle files sometimes lack curly quotes, covers and title pages, and that some of them are generated automatically. On that second point Gutenberg agrees: almost every book since 2004 is released as plain text and HTML, and software makes the EPUB and Kindle versions from those (<a rel=\"external\" href=\"https://www.gutenberg.org/help/file_formats.html\">file formats help</a>).</p>\n<h2 id=\"where-the-text-comes-from\">Where the text comes from</h2>\n<p>The Standard Ebooks <a rel=\"external\" href=\"https://standardebooks.org/contribute/producing-an-ebook-step-by-step\">production guide</a> calls Gutenberg the best place to look for a book to produce. The producer pulls in the Gutenberg file by its book number and cuts away Gutenberg's header, footer and licence. The colophon of the finished book credits the original transcribers and links back to the Gutenberg text.</p>\n<p>The guide gives no reason for stripping the licence, but <a rel=\"external\" href=\"https://www.gutenberg.org/policy/license.html\">Gutenberg's own licence page</a> shows why it can go. “Project Gutenberg” is a registered trademark, and a file that keeps the name comes with conditions, among them a royalty of 20% of gross profits from anyone who sells copies. Strip the licence and every reference to Gutenberg, and what's left is, in the page's words, “a text unrestricted by U.S. intellectual property law”.</p>\n<p>Next the producer finds page scans, preferring the Internet Archive, then HathiTrust, then Google Books, and picks an edition published before 1 January 1931 so that it's in the US public domain. Scans matter because older transcriptions sometimes dropped italics altogether, and only the printed page shows where they were. Each finished book links both sources; <em>Middlemarch</em>, for example, <a rel=\"external\" href=\"https://standardebooks.org/ebooks/george-eliot/middlemarch\">names</a> its transcription at Gutenberg and its scans at the Internet Archive.</p>\n<h2 id=\"what-changes-in-the-words\">What changes in the words</h2>\n<p>A change that is an editorial decision about the author's text, such as a modernised spelling, goes into its own commit labelled “[Editorial]” so that a purist can undo it. The <a rel=\"external\" href=\"https://standardebooks.org/manual/1.9.1/single-page\">Manual of Style</a> doesn't count a fixed transcriber's slip as editorial, and it puts converting the quotation marks and fixing an obvious printer's typo in the same class. Spelling is the main editorial change. A script closes up compounds that modern English has joined, so “to-night” becomes “tonight” and “some one” becomes “someone”. One case from the guide shows the risk: in <em>The Picture of Dorian Gray</em> the script turned Wilde's “gold-thread” into “goldthread”, which is a plant, and the producer had to put the hyphen back.</p>\n<p>The rule is to modernise spelling that sounds the same, never grammar. “Mak” can become “make”; “maketh” stays. British and American spellings are left as the author had them, even when a book mixes the two.</p>\n<p>Punctuation follows American practice. Full stops and commas go inside the quotation marks, as the <a rel=\"external\" href=\"https://standardebooks.org/manual/1.9.1/8-typography\">Manual of Style</a> requires, and the guide converts books printed with single quotation marks for dialogue to double ones, because a program can turn American style into British more reliably than the reverse. Ligatures are expanded too: “Archæology” becomes “Archaeology”, though French and other languages keep theirs where the ligature changes a word.</p>\n<h2 id=\"the-typography\">The typography</h2>\n<p>In November 2026 the <a rel=\"external\" href=\"https://standardebooks.org/manual\">Manual of Style</a> stood at version 1.9.1; anything it doesn't cover goes by the latest <em>Chicago Manual of Style</em>. The typography chapter is precise to the character. An ellipsis is the single glyph U+2026, never three full stops. Two quotation marks that sit side by side get a hair space between them. A speaker cut off mid-sentence gets one em dash, a partly hidden name such as “D⸺” gets a two-em dash (U+2E3A), and a word hidden completely gets a three-em dash (U+2E3B). Acronyms like NASA are set in small caps, and section numbers in the text are Roman numerals typed as ordinary capital letters, not the special Unicode numeral characters.</p>\n<p>Footnotes move into a single endnotes file, which becomes popup notes in the finished EPUB. The book page states the length up front: for <em>Middlemarch</em>, 318,047 words, or 19 hours 17 minutes of reading.</p>\n<h2 id=\"covers-and-the-cc0-dedication\">Covers and the CC0 dedication</h2>\n<p>Every cover is a public-domain painting in a “fine art” style, cropped to 1400 × 2100 pixels with the title and author laid over it. Paintings are harder to clear than text under US copyright law, so a producer has to show a scan of the painting in a book published before 1931, and the editor-in-chief has the last word. The blurbs and other metadata are written by volunteers, and the guide bans AI tools from every step of production.</p>\n<p>What the volunteers add (the markup, the cover design, the files themselves) is dedicated to the public domain under <a rel=\"external\" href=\"https://standardebooks.org/about/standard-ebooks-and-the-public-domain\">CC0 1.0</a>. The Manual of Style carries the same dedication, so another publisher can reuse both without asking.</p>\n<h2 id=\"the-files-you-can-download\">The files you can download</h2>\n<p>A Standard Ebooks page offers four files. The “compatible epub” is for every device and app except Kindles and Kobos; azw3 is for Kindle, with a separate cover thumbnail because <a rel=\"external\" href=\"https://standardebooks.org/contribute/producing-an-ebook-step-by-step\">a bug in Amazon's software</a> stops Kindles showing covers of side-loaded books; kepub is for Kobo; and the “advanced epub” uses features that most e-readers don't support yet. Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/145\">page for the same novel</a> has no Kobo-specific file among its downloads; <a href=\"/blog/kepub-or-epub-for-kobo/\">KEPUB or EPUB</a> explains what the Kobo format adds. Zip files of whole collections are a perk of Standard Ebooks' paid <a rel=\"external\" href=\"https://standardebooks.org/bulk-downloads\">Patrons Circle</a>; single books stay free.</p>\n<h2 id=\"how-big-each-library-is\">How big each library is</h2>\n<p>Standard Ebooks is small by comparison. Its <a rel=\"external\" href=\"https://standardebooks.org/ebooks\">catalogue</a> shows 12 books a page, and when this post was written, on 1 November 2026, page 129 was the last and held four: about 1,540 titles, or one for every 50 or so at Gutenberg. The gap follows from the work: a first book should be straightforward prose of no more than about 100,000 words, because long first attempts are often abandoned, the <a rel=\"external\" href=\"https://standardebooks.org/contribute/producers\">page for producers</a> warns, and every book goes to a reviewer before release.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/text-to-speech-on-e-readers/","url":"https://readwyn.com/blog/text-to-speech-on-e-readers/","title":"Can a Kindle or a Kobo read books aloud?","summary":"Which Kindles and Kobos can read an ebook aloud, what they need to do it, which books are left out, and the phone apps that read aloud too.","content_html":"<p>A recent Kindle can, as long as you pair Bluetooth headphones or a speaker and the book allows it. A Kobo e-reader can do it only in the European Union, on four models, after an optional update; elsewhere the Kobos with Bluetooth play audiobooks but don't turn the text of an ebook into speech.</p>\n<h2 id=\"kindle\">Kindle</h2>\n<p>Kindle e-readers have two ways to speak. VoiceView, a screen reader that Amazon <a rel=\"external\" href=\"https://aboutamazon.com/news/books-and-authors/kindle-accessibility-features-for-all-readers\">added in 2016</a> for blind and low-vision readers, reads everything on the screen, menus included, and you work it with gestures. On an 11th generation or newer Kindle you put the headphones into pairing mode, press the power button three times within two seconds, and hold two fingers apart on the screen once you hear sound, <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=G202051830\">Amazon's pairing page</a> says; on an older Kindle with Bluetooth you hold the power button for nine seconds and then press two fingers on the screen for a second, in place of the three presses. Open a book and VoiceView reads straight through, <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=202051910\">turning the pages itself</a>. In November 2026 it spoke English, French, German, Italian, Portuguese and Spanish.</p>\n<p>Assistive Reader does without the gestures. It turned up in software 5.18.4 at the end of July 2025, vanished from some Kindles a few days later, and <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2025/08/19/another-kindle-software-update-released-brings-back-tts/\">came back in 5.18.4.0.1</a>; it too needs Bluetooth speakers or headphones. As of November 2026 it runs on <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TdHCSXXqsdhaVKt3Tc\">11th generation and later Kindles</a>, the Kindle Scribe and the Kindle Colorsoft, and reads English, Spanish, French, German and Italian. In a book, tap the top centre of the screen, then Aa, then More, and switch on Assistive Reader. A player bar gives play, pause, speed and a 30 second rewind, and each word is underlined as it's spoken.</p>\n<p>Not every book qualifies. Assistive Reader works only with titles that show “Enhanced Typesetting: Enabled” in the Kindle Store (<a href=\"/blog/kindle-enhanced-typesetting/\">what that label means</a>). In the Kindle apps for iPhone, iPad, Android, Mac and Fire tablets, <a rel=\"external\" href=\"https://www.amazon.com/gp/help/customer/display.html?nodeId=TqLHvK6eo6O7DJVQoZ\">Amazon adds another condition</a>: the book must be bought from the Kindle Store, borrowed through Prime Reading or another official route, or shared through Family Library. A document you sent to your Kindle yourself isn't on that list.</p>\n<p>Publishers have had a say since the Kindle 2, which went on sale in February 2009 with a read-aloud feature. The Authors Guild said it ate into the audiobook market, and within weeks <a rel=\"external\" href=\"https://phys.org/news/2009-03-amazon-authors-mute-kindle-read-aloud.html\">Amazon agreed</a> to let rights holders switch it off book by book, while still calling the feature legal. On 7 June 2009 a Google search by Knowledge Ecology International <a rel=\"external\" href=\"https://www.keionline.org/?p=20936\">found about 13,500 Amazon pages</a> marked \"Text-to-Speech: Not enabled\", up from 343 on 17 May; its examples included Kindle editions of <em>The Communist Manifesto</em> and <em>Crime and Punishment</em>.</p>\n<h2 id=\"kobo\">Kobo</h2>\n<p>An <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/32246787707799\">optional update</a> released in July 2025 for the Clara BW, Clara Colour, Libra Colour and Elipsa 2E is offered only to readers in the European Union, where the <a href=\"/blog/european-accessibility-act-ebooks/\">European Accessibility Act</a> had begun to apply to e-readers a month earlier. It adds Read Aloud for books in EPUB and Kobo EPUB (tap the middle of the page, then More, then Read Aloud) and a Screen Reader under Settings › Accessibility, which needs a Bluetooth audio device and a Bluetooth keyboard to move around the device. Before it, Kobo e-readers <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2025/07/29/kobo-ereaders-screen-reader/\">had no text-to-speech at all</a>.</p>\n<p>Installing it wipes the device: every book on it goes (Kobo purchases can be downloaded again, files copied over from a computer can't), along with any bookmarks and annotations made since the last sync, your sign-in and your settings. Going back (Settings › Device Information › Revert) resets the device again.</p>\n<p>Outside the EU, as of November 2026, the Clara 2E, Sage, Libra 2, Elipsa, Elipsa 2E, Clara BW, Clara Colour and Libra Colour <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/4406292712471\">play audiobooks</a> over Bluetooth headphones or speakers. Those are recordings sold as separate titles, from about 50 MB to 14 GB or more each, and Kobo warns you may have to remove ebooks to make room. (Whether listening counts as reading is <a href=\"/blog/audiobooks-vs-reading/\">another question</a>.)</p>\n<table><thead><tr><th></th><th>Feature</th><th>Needs</th><th>Which books</th></tr></thead><tbody>\n<tr><td>Kindle, 11th generation and later, Scribe, Colorsoft</td><td>Assistive Reader, VoiceView</td><td>Bluetooth audio</td><td>“Enhanced Typesetting” for Assistive Reader</td></tr>\n<tr><td>Earlier Kindles with Bluetooth</td><td>VoiceView</td><td>Bluetooth audio</td><td>Not stated by Amazon</td></tr>\n<tr><td>Kindle app (iPhone, iPad, Android, Mac, Fire)</td><td>Assistive Reader</td><td>The device's own voices</td><td>“Enhanced Typesetting”, from the Kindle Store or an official loan</td></tr>\n<tr><td>Kobo Clara BW, Clara Colour, Libra Colour, Elipsa 2E (EU only)</td><td>Read Aloud, Screen Reader</td><td>Bluetooth audio; a keyboard for the Screen Reader</td><td>EPUB and Kobo EPUB</td></tr>\n<tr><td>Kobo app for iOS (version 11.0 and later)</td><td>Read Aloud</td><td>Nothing extra</td><td>All ebooks in the EU, some titles elsewhere</td></tr>\n<tr><td>Google Play Books on Android</td><td>Read aloud</td><td>Internet for the more natural voice</td><td>Those the publisher allows</td></tr>\n</tbody></table>\n<h2 id=\"phones-tablets-and-android-e-readers\">Phones, tablets and Android e-readers</h2>\n<p>Google Play Books on Android has Read aloud in the menu at the top right of a book, and a setting, Automatically read aloud, that starts it whenever a book opens. Publishers <a rel=\"external\" href=\"https://support.google.com/googleplay/answer/11938821\">decide which books get it</a>, and the More natural reading voice works only online.</p>\n<p>An <a href=\"/blog/android-e-readers/\">Android e-reader</a> can install Play Books from the Play Store, and some don't need headphones. In November 2026 the BOOX Go 7, for one, <a rel=\"external\" href=\"https://shop.boox.com/products/go7\">shipped with</a> a built-in speaker, Android 13 with the Google Play Store, and a free text-to-speech feature of its own.</p>\n<p>Kobo's iOS app has had Read Aloud <a rel=\"external\" href=\"https://help.kobo.com/hc/articles/32430599153687\">since version 11.0</a>, with a choice of speed, language and voice, and the <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/35996348493207\">web reader</a> has it too. In the EU both will read any ebook aloud. Elsewhere they read only some titles, which on the web means those whose publisher has switched it on.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/vertical-text-in-ebooks/","url":"https://readwyn.com/blog/vertical-text-in-ebooks/","title":"Can ebooks show vertical Japanese text?","summary":"Yes. The EPUB 3 settings that set Japanese in vertical columns with right-to-left page turns and ruby, and what Apple Books and Kindle expect.","content_html":"<p>Open a Japanese novel in Apple Books and the characters run down the page in columns, the first column sits at the right edge, and the next page is to the left. Ebooks can do this because EPUB 3 supports it, and two things in the file switch it on. The publisher's stylesheet sets <code>writing-mode: vertical-rl</code> on the text, and the package file puts <code>page-progression-direction=\"rtl\"</code> on its spine so that the pages turn the Japanese way. If a Japanese book shows ordinary horizontal lines, the stylesheet doesn't set the writing mode, or the app ignores it. If the columns are vertical but the pages turn the wrong way, the spine lacks the direction.</p>\n<p>The page direction is a property of the whole book. <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/\">EPUB 3.3</a> gives the spine three choices, <code>ltr</code>, <code>rtl</code> and <code>default</code>, and with <code>default</code> the reading app decides. A single chapter can still go its own way through <code>writing-mode</code>, a CSS property that became a <a rel=\"external\" href=\"https://www.w3.org/TR/css-writing-modes-3/\">W3C Recommendation in December 2019</a>. <code>vertical-rl</code> puts each new line to the left of the last, as in a printed Japanese book; <code>vertical-lr</code> stacks them the other way, which Mongolian uses. EPUB keeps a prefixed twin, <code>-epub-writing-mode</code>, with the same values and behaviour.</p>\n<p>Apple's <a rel=\"external\" href=\"https://help.apple.com/itc/booksassetguide/en.lproj/itc7e9ec52f7.html\">Books Asset Guide</a> adds its own rules. The writing mode has to be set on the <code>html</code> or <code>body</code> element, and each content file can have only one, so a vertical novel with a horizontal appendix needs the two in separate files. Apple asks for Hiragino Mincho ProN or Hiragino Kaku ProN because they're already on the reader's device; YuMincho has to be downloaded from the Font menu first. Switch to the Scroll theme and a vertical book <a rel=\"external\" href=\"https://help.apple.com/itc/booksassetguide/en.lproj/itccf6b30e09.html\">scrolls sideways</a>, which Apple does by default for Japanese and Chinese books unless the file sets <code>ibooks:scroll-axis</code>.</p>\n<p>Latin letters and numbers lie on their side by default (<code>text-orientation: mixed</code>), turned 90 degrees clockwise, while the Japanese characters stay upright. Dates and other short numbers often get tate-chu-yoko instead: two digits set side by side in the space of one character. The <a rel=\"external\" href=\"https://www.w3.org/TR/css-writing-modes-3/#text-combine-upright\">CSS specification</a> shows it with the date 平成20年4月16日, where <code>text-combine-upright: all</code> squeezes \"20\" and \"16\" into single cells. Apple's guide wants the WebKit form, <code>-webkit-text-combine: horizontal</code>, and EPUB has its own <code>-epub-text-combine-horizontal</code>.</p>\n<p>Then there's ruby, the small kana printed beside a hard kanji to give its reading. In vertical text it goes to the right of the column, and in horizontal text above the line; the W3C's <a rel=\"external\" href=\"https://www.w3.org/TR/jlreq/\">Requirements for Japanese Text Layout</a> calls the other sides very rare. A 2010 draft from the Japan Electronic Publishing Association, <a rel=\"external\" href=\"https://www.jepa.or.jp/press_release/reqEPUBJ.html\">Minimal Requirements on EPUB for Japanese Text Layout</a>, set the ruby at about half the size of the base characters and wanted a fallback: the reading in parentheses after the word, for apps that couldn't draw ruby. The same draft asked for page numbers and running heads to stay horizontal in a vertical book, and for readers to be able to show a vertical book horizontally, which it said many Japanese viewers already allowed. The current EPUB 3.3 specification tells publishers not to use the <code>rp</code> element that held those parentheses, because EPUB 3 reading systems already handle ruby. When Rakuten <a rel=\"external\" href=\"https://global.rakuten.com/corp/news/press/2012/0702_01.html\">opened Kobo in Japan in July 2012</a>, its announcement named vertical text and ruby among what the Kobo Touch's EPUB 3 support covered.</p>\n<p>Amazon has a switch of its own. The <a rel=\"external\" href=\"https://kindlegen.s3.amazonaws.com/AmazonKindlePublishingGuidelines.pdf\">Kindle Publishing Guidelines</a> (version 2026.3) describe a <code>&lt;meta name=\"primary-writing-mode\"&gt;</code> tag in the package file, and its value <code>vertical-rl</code> gives right-to-left page turns for Chinese, Japanese and Korean books. The same tag is required for Virtual Panels in a manga that reads right to left.</p>\n<p>In a <a href=\"/blog/fixed-layout-vs-reflowable/\">fixed-layout</a> manga made of page images, the writing mode doesn't matter, only the page direction. A manga in a CBZ archive has no spine at all; if the archive has a <code>ComicInfo.xml</code>, its <code>Manga</code> field set to <code>YesAndRightToLeft</code> <a rel=\"external\" href=\"https://anansi-project.github.io/docs/comicinfo/documentation\">gives the direction</a>, as the post on <a href=\"/blog/read-manga-right-to-left-on-mac/\">reading manga right to left</a> explains.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/waterproof-e-readers-ipx8/","url":"https://readwyn.com/blog/waterproof-e-readers-ipx8/","title":"What does IPX8 mean on a waterproof e-reader?","summary":"IPX8 on a Kindle or Kobo means a maker's test in 2 metres of fresh water for 60 minutes, with no dust rating. What the code promises and what it leaves out.","content_html":"<p>On a Kindle or a Kobo, IPX8 means the reader passed 60 minutes under 2 metres of fresh water, a depth and time that the maker chose. Amazon's spec sheet for the <a rel=\"external\" href=\"https://images-na.ssl-images-amazon.com/images/I/812Wt8KavHS.pdf\">2018 Kindle Paperwhite</a> gives exactly those figures, and PocketBook's page for the <a rel=\"external\" href=\"https://pocketbook.ch/en-ch/catalog/e-readers/inkpad-4-ch\">InkPad 4</a> says the same thing in its own words: fresh water, 2 metres, up to 60 minutes. The code itself comes from IEC 60529, the International Electrotechnical Commission's standard for the enclosures of electrical equipment; the edition the IEC sells today is the 1989 second edition with amendments from 1999 and 2013 (<a rel=\"external\" href=\"https://webstore.iec.ch/en/publication/2452\">IEC webstore</a>). The first digit after \"IP\" rates protection against solids such as dust, and the second rates protection against water.</p>\n<p>Amazon called the Oasis of October 2017 <a rel=\"external\" href=\"https://press.aboutamazon.com/2017/10/meet-the-all-new-kindle-oasis-featuring-a-7-inch-300-ppi-display-waterproof-design-and-audible\">its first waterproof Kindle</a>, and <a rel=\"external\" href=\"https://www.engadget.com/2017-10-11-amazon-new-kindle-oasis.html\">Engadget described</a> it as IPX8 and good for two metres for an hour. The cheaper Paperwhite <a rel=\"external\" href=\"https://fortune.com/2018/10/16/amazon-kindle-paperwhite-waterproof/\">got the same treatment</a> a year later.</p>\n<h2 id=\"what-the-x-means\">What the X means</h2>\n<p>It's tempting to read the X as a zero, a failed dust test. It only means that no rating is given for that digit (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/IP_code\">IP code summary</a>): nobody claims a result for solids, good or bad. For e-readers the practical answer is the same, though, and Kobo writes it out on its <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763773-Waterproof-Kobo-eReaders\">waterproof help page</a>: its readers aren't dustproof or sandproof. At the beach, sand in the charging port is the risk the rating says nothing about.</p>\n<h2 id=\"how-deep-and-how-long\">How deep and how long</h2>\n<p>A 7 in the second place means one thing everywhere: the standard sets a 30-minute immersion with the bottom of the device 1 metre down. An 8 has no set depth or time. It requires only conditions agreed with the manufacturer that are harsher than those for 7 (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/IP_code\">IP code summary</a>). Two IPX8 gadgets can have passed quite different tests, so the figures that matter are the depth and minutes printed next to the code.</p>\n<p>Kobo's own range shows the difference. The first Aura H2O (model N250) was rated IP67, good for 1 metre and 30 minutes, with a cover over its port that had to stay shut. The later models that the <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763773-Waterproof-Kobo-eReaders\">help page</a> gives figures for, from the Aura H2O Edition 2 to the Clara 2E, are IPX8 at 2 metres for an hour. The page gives each model with its model number, so it helps to <a href=\"/blog/which-kobo-do-i-have/\">know which Kobo you have</a>.</p>\n<h2 id=\"salt-water-spray-and-a-wet-screen\">Salt water, spray and a wet screen</h2>\n<p>The test is done in fresh water. Kobo tells owners not to submerge a reader in salt water, soapy water or chlorinated water, and to rinse it in fresh water straight away whenever it does get wet. When the 2017 Oasis came out, Amazon told the Associated Press that pool and sea water would still get some protection, but less than fresh water (<a rel=\"external\" href=\"https://www.foxbusiness.com/features/you-can-shower-with-the-new-kindle-you-just-cant-read\">Fox Business, reprinting the AP review</a>).</p>\n<p>Nor does a high number cover a jet of water. Above 6, the water digits aren't cumulative: a device rated IPX7 or IPX8 hasn't necessarily passed the jet tests behind IPX5 and IPX6, and one that has carries both codes, such as IPX5/IPX7 (<a rel=\"external\" href=\"https://en.wikipedia.org/wiki/IP_code\">IP code summary</a>). Kobo's help page asks readers to keep their e-readers away from fast, forceful water such as waves and waterfalls, and adds that the devices don't float.</p>\n<p>The rating covers water getting inside, not whether the screen works while wet. Anick Jesdanun, who <a rel=\"external\" href=\"https://www.foxbusiness.com/features/you-can-shower-with-the-new-kindle-you-just-cant-read\">reviewed the 2017 Oasis for the AP</a>, found its touchscreen erratic once wet: when he held it under in a cold bath, the font size changed and settings popped up, and in a hot shower the pages turned by themselves. Kobo's help page warns that a wet touchscreen may not respond as expected, tells owners to dry the USB port before plugging in a charger, and says not to charge near water or in high humidity.</p>\n<p>A seal that has been opened is no longer the seal that was tested. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/21137184146071-Repair-your-Kobo-eReader\">repair page</a> says opening an eReader for any reason may compromise its waterproofing, and Rakuten Kobo disclaims liability for water damage to a device whose seal has been broken.</p>\n<h2 id=\"which-readers-are-rated\">Which readers are rated</h2>\n<p>In <a rel=\"external\" href=\"https://www.which.co.uk/news/article/which-ebook-readers-will-survive-a-dip-in-the-pool-aClKE9Q4XICF\">August 2023</a> the consumer magazine Which? wrote that it submerges each waterproof e-reader in 2 metres of water for 60 minutes, the makers' own figures, and recommends only the ones that pass. Its IPX8 models were the Kindle Paperwhite from the 10th generation on, the Oasis from the 9th (<a href=\"/blog/which-kindle-do-i-have/\">which Kindle do I have</a> tells the generations apart), and the Kobo Clara 2E, Sage and Libra 2. Which? listed the basic Kindle of 2022, the Kobo Elipsa, the Clara HD and the Nia as not waterproof.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-are-bisac-codes/","url":"https://readwyn.com/blog/what-are-bisac-codes/","title":"What are BISAC codes, and how do stores sort books by genre?","summary":"BISAC subject codes such as FIC009000 tell US retailers where a book belongs. How the codes are built, who picks them, how stores use them, and Thema.","content_html":"<p>When the subjects of an ebook read like FICTION / Fantasy / General, or its record carries a string such as FIC009000, you're looking at a BISAC subject heading. BISAC (Book Industry Standards and Communications) subject headings are the subject list of the US book trade, kept by the Book Industry Study Group. BISG asks publishers to give each title no more than three of them (<a rel=\"external\" href=\"https://www.bisg.org/BISAC-FAQ\">BISAC FAQ</a>), and <a rel=\"external\" href=\"https://www.bisg.org/BISAC-Subject-Codes-main\">a heading can decide</a> where a shop shelves the book and which genres it's found under in a search.</p>\n<h2 id=\"how-a-code-is-built\">How a code is built</h2>\n<p>Every heading has a text part of two to four levels, separated by slashes, and a nine-character code: three letters, then six digits. The list has 54 major sections, from ANTIQUES &amp; COLLECTIBLES to YOUNG ADULT NONFICTION, and the example in <a rel=\"external\" href=\"https://www.bisg.org/BISAC-FAQ\">BISG's FAQ</a> is TRV025070, \"TRAVEL / United States / South / General\". The letters are usually a mnemonic, so FIC is fiction and TRV is travel.</p>\n<p>The digits no longer mean anything. The FAQ says the codes of the first version were hierarchical and today's should be treated as plain identifiers, and the <a rel=\"external\" href=\"https://www.bisg.org/fiction\">Fiction section</a> shows why. FICTION / Thrillers / General is FIC031000, yet Thrillers / Espionage is FIC006000. FICTION / Time Travel is FIC028080, in the FIC028 run that otherwise belongs to Science Fiction. In 2025 the section BODY, MIND &amp; SPIRIT was renamed MIND, BODY, SPIRIT, and its codes kept the prefix OCC (<a rel=\"external\" href=\"https://www.bisg.org/changed-from-the-previous-edition\">changes in the 2025 edition</a>).</p>\n<p>Books for children and teenagers have sections of their own. A fantasy novel is FIC009000 for adults, JUV037000 under <a rel=\"external\" href=\"https://www.bisg.org/juvenile-fiction\">JUVENILE FICTION</a> and YAF019000 under <a rel=\"external\" href=\"https://www.bisg.org/young-adult-fiction\">YOUNG ADULT FICTION</a>. Juvenile headings cover ages 0 to 11 and young adult headings ages 12 to 18. BISG's FAQ asks publishers not to combine the two, or either of them with an adult section.</p>\n<h2 id=\"who-picks-the-codes-and-how-many\">Who picks the codes, and how many</h2>\n<p>BISG doesn't assign codes. The author or publisher does, usually an editor or someone in marketing, with the main subject first. <a rel=\"external\" href=\"https://www.bisg.org/BISAC-FAQ\">BISG's worked example</a> is a biography of Charles Lindbergh: BIO034000 (aviation and nautical biography) as the main code, plus a code for the history of aviation and TRU006000 for kidnappings. If the book is mainly about the kidnapping, the true crime code moves to the front.</p>\n<p>BISG also wants the most specific heading rather than a General one, fiction and nonfiction headings kept apart in all but rare cases, and the same headings on every format of a book, print or digital. Anything with an ISBN should have a code, so a blank notebook gets NON000000, Non-Classifiable. The codes travel in ONIX, the XML feed that publishers send to their trading partners. The nine characters shouldn't be printed on the book, the FAQ says; the words of the heading can go on the back cover, bottom left. Anyone can look up the whole list on BISG's site for free. A company that loads it into a database <a rel=\"external\" href=\"https://www.bisg.org/BISAC-Subject-Codes-main\">needs a licence</a>, and non-members pay according to their annual book revenue.</p>\n<h2 id=\"how-a-shop-turns-codes-into-shelves\">How a shop turns codes into shelves</h2>\n<p>A code is the publisher's suggestion, and what a shop does with it is up to the shop. Some retailers have mapped the headings to their own store sections, <a rel=\"external\" href=\"https://www.bisg.org/BISAC-FAQ\">BISG's FAQ</a> says, and others use the terms as they are: as search terms, on browsable subject pages, for discounts on a subject, and for bestseller lists by heading. The FAQ names Amazon, Barnes &amp; Noble, Indigo, Baker &amp; Taylor, and NPD BookScan among the users of the list.</p>\n<p>Amazon's self-publishing service asks for something else. In November 2026, the <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200652170\">KDP help page on categories</a> asks for three categories from Amazon's own store tree, written as paths such as \"Fiction &gt; Historical Fiction\", and never mentions BISAC. Amazon may change a category that doesn't fit the book, and a change can take up to 72 hours to show.</p>\n<h2 id=\"a-new-edition-every-year\">A new edition every year</h2>\n<p>The Subject Codes Committee meets monthly, and new headings take effect only in the yearly edition. The 2025 edition came out in December 2025, and it was still the current list on BISG's site in November 2026. Its <a rel=\"external\" href=\"https://www.bisg.org/changed-from-the-previous-edition\">summary of changes</a> lists FICTION / Cozy (FIC155000) and four seasonal romance headings, Winter, Spring, Summer, and Autumn. Codes that are retired are never given to a new heading: a <a rel=\"external\" href=\"https://bisg.org/page/InactivatedCodes\">cumulative appendix</a> keeps every inactivated code with a replacement, or the section to choose one from, so an old record can still be read. A receiver on an older edition has to update or ignore the headings it doesn't know.</p>\n<h2 id=\"thema-the-international-scheme\">Thema, the international scheme</h2>\n<p>The international counterpart is Thema, which EDItEUR manages and anyone may use free of charge. The project started at the Frankfurt Book Fair in 2012, version 1.0 was presented at the 2013 fair, and from 31 December 2013 publishers could put Thema codes in their ONIX feeds (<a rel=\"external\" href=\"https://www.editeur.org/206/About-Thema/\">About Thema</a>). The UK dropped its own BIC scheme for it: BIC froze its development in February 2017 and set its retirement for the end of February 2024 (<a rel=\"external\" href=\"https://www.alpsp.org/news-publications/industry-news/bic-reminds-industry-date-bic-standard-february2023/\">BIC press release, via ALPSP</a>).</p>\n<p>Thema is built differently. A subject is a short letter code, and <a rel=\"external\" href=\"https://ns.editeur.org/thema/en\">qualifiers that start with a digit</a> add the place (1), language (2), time period (3), educational purpose (4), interest such as reading age (5), or style (6). A police procedural is <a rel=\"external\" href=\"https://ns.editeur.org/thema/en/FFP\">FFP</a>, and its note suggests adding 1DDU for one set in Britain. In BISAC the book is FICTION / Mystery &amp; Detective / Police Procedural (FIC022020), and its setting can take a second heading, such as FICTION / Places / United States (FIC152000). Version 1.7 came out on 28 October 2026 with 44 new subject categories and 217 new qualifiers (<a rel=\"external\" href=\"https://www.editeur.org/151/Thema/\">EDItEUR</a>), and a count of its <a rel=\"external\" href=\"https://www.editeur.org/files/Thema/1.7/en/20261020_Thema_v1.7_en.json\">English file</a> gives 3,466 categories and 5,982 qualifiers. EDItEUR's <a rel=\"external\" href=\"https://ns.editeur.org/thema/en\">online browser</a> offers Thema in English and 27 other languages and regional variants, some only partly translated. The official BISAC list is English only. Its licensed download includes a map from BISAC to Thema, according to the FAQ.</p>\n<h2 id=\"why-the-genres-in-your-own-library-are-a-mix\">Why the genres in your own library are a mix</h2>\n<p>Inside an EPUB, subjects go in <code>dc:subject</code>. <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-opf-dcsubject\">EPUB 3.3</a> lets the file name the scheme with an <code>authority</code> property and the code with <code>term</code>, and its own example is FICTION / Occult &amp; Supernatural with the authority BISAC and the term FIC024000. Without an authority, the subject is free text. So one book can carry BISAC headings, another Thema labels, and a third the publisher's own words, all in the same field (see <a href=\"/blog/epub-metadata-dublin-core/\">what metadata is inside an EPUB</a>). That's how a shelf of fantasy ends up split between books tagged plain \"Fantasy\" and books tagged FICTION / Fantasy / Epic, BISAC's FIC009020.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-happened-to-the-nook/","url":"https://readwyn.com/blog/what-happened-to-the-nook/","title":"What happened to the Nook?","summary":"Barnes & Noble's Nook never went away: what changed with Microsoft, Samsung, Lenovo and the UK store, and the 7.8-inch GlowLight 4 Plus that B&N still lists in 2026.","content_html":"<p>It shrank, but it was never discontinued. Barnes &amp; Noble still sells Nook e-readers in the US, and its help centre had a page on the 7.8-inch GlowLight 4 Plus updated in July 2026. What changed is the size of the business around it: B&amp;N sold a stake to Microsoft and bought it back, had its tablets made by Samsung and then Lenovo, and closed its British store in 2016. In November 2025 it was still selling the 6-inch GlowLight 4 at $149.99, and added an <a rel=\"external\" href=\"https://www.androidauthority.com/nook-glowlight-4-ocean-teal-launch-3614108/\">Ocean Teal</a> version.</p>\n<h2 id=\"a-colour-strip-under-the-e-ink\">A colour strip under the e-ink</h2>\n<p>The first Nook, reviewed by <a rel=\"external\" href=\"https://www.engadget.com/2009/12/07/barnes-and-noble-nook-review/\">Engadget in December 2009</a>, had a 6-inch, 16-shade grey E Ink screen above a 3.5-inch colour touchscreen, and that small strip handled nearly everything, from the store to the keyboard to the music player. It had Wi-Fi and free AT&amp;T 3G for shopping and downloads. Scrolling through lists on it gave the reviewers trouble: a downward swipe often didn't register, and they suspected the strip was too short for the gesture.</p>\n<p>In April 2012 B&amp;N put a light in its e-ink reader. The <a rel=\"external\" href=\"https://www.engadget.com/2012-04-12-nook-simple-touch-with-glowlight-hands-on.html\">Simple Touch with GlowLight</a> cost $139, $40 more than the plain model, and lit the page from above the top edge of the screen; you held the \"n\" button for two seconds to switch it on. In August 2013, in a <a rel=\"external\" href=\"https://www.engadget.com/2013/08/18/nook-simple-touch-with-glowlight-gets-another-20-price-drop/\">price cut announcement</a>, B&amp;N called it \"the first Reader to hit the market with built-in illumination\". Our post on <a href=\"/blog/front-light-and-warm-light/\">front lights and warm light</a> explains how such lights work.</p>\n<h2 id=\"microsoft-s-minority-stake\">Microsoft's minority stake</h2>\n<p>On 30 April 2012 Microsoft agreed to put <a rel=\"external\" href=\"http://www.bgr.com/2012/04/30/microsoft-barnes-noble/\">$300 million</a> into a new Barnes &amp; Noble subsidiary that combined the Nook digital business with its college business. That bought about 17.6% at a post-money valuation of $1.7 billion, and B&amp;N kept roughly 82.4%. The partnership ended in December 2014, when Barnes &amp; Noble <a rel=\"external\" href=\"https://www.entrepreneur.com/business-news/barnes-noble-buys-microsofts-stake-in-nook-media-ending/240512\">bought Microsoft's stake back</a> for $62.4 million in cash plus shares, a deal it valued at about $125 million.</p>\n<h2 id=\"tablets-made-by-samsung-then-lenovo\">Tablets made by Samsung, then Lenovo</h2>\n<p>By 2014 the Nook tablet was an Android tablet made with Samsung, the Galaxy Tab 4 Nook. A <a rel=\"external\" href=\"https://www.theverge.com/2014/10/22/7038699/samsung-nook-tablet-now-comes-in-10-inch-size\">10.1-inch version</a> came out that October at $299.99, with a 1280 × 800 screen (149 ppi). In 2021 B&amp;N turned to Lenovo. The <a rel=\"external\" href=\"https://news.lenovo.com/pressroom/press-releases/barnes-noble-introduces-new-nook-10-hd-tablet-designed-with-lenovo/\">Nook 10\" HD tablet</a>, announced on 19 March 2021 for early April at $129.99, had a 10.1-inch screen; Lenovo's press release called it the first major collaboration of the two companies. The <a rel=\"external\" href=\"https://www.barnesandnobleinc.com/press-release/barnes-noble-introduces-new-nook-9-tablet-designed-lenovo/\">Nook 9\" tablet</a>, announced on 5 February 2024 for release that May, cost $149.99 and came with 64 GB of storage and a microSD slot. The latest as of November 2026, the <a rel=\"external\" href=\"https://www.barnesandnobleinc.com/press-release/barnes-noble-introduces-new-nook-reading-tablet-8-7-designed-lenovo/\">Nook Reading Tablet 8.7</a>, was announced on 23 February 2026 for mid-March, also at $149.99.</p>\n<h2 id=\"the-uk-store-closed-in-2016\">The UK store closed in 2016</h2>\n<p>British readers are the ones for whom the Nook really did disappear. In March 2016 B&amp;N told users it would close the Nook App Store, Nook Video and the <a rel=\"external\" href=\"https://www.theverge.com/2016/3/4/11160942/barnes-noble-nook-app-store-shutting-down-march-15th\">UK Nook Store on 15 March</a>. Customers in Britain who wanted to keep their ebooks had to open an account with Sainsbury's, the supermarket chain, and a book that couldn't move to Sainsbury's platform was to be replaced with a Sainsbury's Entertainment voucher, according to <a rel=\"external\" href=\"https://diginomica.com/2016/03/07/bn-nukes-the-nook-with-a-15-march-deadline-for-customers-to-save-their-content/\">diginomica</a>. The same report gives the figures behind the retreat. In the quarter that <a rel=\"external\" href=\"https://www.cityam.com/?p=335857\">ended in January 2016</a>, Nook sales fell 33% year on year to about $52 million, and device and accessory sales fell 44% to $21 million.</p>\n<h2 id=\"the-e-readers-since-2017\">The e-readers since 2017</h2>\n<p>The 6-inch mid-range line went four years without a new model after the GlowLight 3 of 2017, though a 7.8-inch <a rel=\"external\" href=\"https://www.engadget.com/2019-05-23-barnes-and-noble-nook-ereader-glowlight-plus.html\">GlowLight Plus</a> went on sale in May 2019 at $200. When the <a rel=\"external\" href=\"https://www.engadget.com/barnes-and-noble-nook-glowlight-4-e-reader-191526217.html\">GlowLight 4 arrived in December 2021</a>, Engadget called it the first new mid-range Nook since the GlowLight 3. It cost $150, kept a 6-inch 300 dpi screen, went up to 32 GB of storage, and was the first Nook e-reader with USB-C. A cheaper <a rel=\"external\" href=\"https://www.barnesandnobleinc.com/press-release/barnes-noble-introduces-nook-glowlight-4e/\">GlowLight 4e</a> was announced on 25 May 2022 at $119.99, for sale from 7 June, with a 6-inch 212 dpi screen and physical page-turn buttons.</p>\n<p>The <a rel=\"external\" href=\"https://help.barnesandnoble.com/hc/en-us/articles/16957536716443-Introducing-the-NOOK-GlowLight-4-Plus\">GlowLight 4 Plus</a> went back to 7.8 inches, at $199.99, with an <a href=\"/blog/waterproof-e-readers-ipx8/\">IPX7 rating</a> (30 minutes in up to 3 feet of fresh water) and audiobook playback through a headphone jack or Bluetooth, according to the help page that B&amp;N updated on 21 July 2026.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-happens-when-an-ebook-store-closes/","url":"https://readwyn.com/blog/what-happens-when-an-ebook-store-closes/","title":"What happens to your ebooks when the store closes?","summary":"When an ebook shop closes, your books move to another store or are deleted and refunded. What Sony, Nook UK, Sainsbury's and Microsoft did with their customers' books.","content_html":"<p>In three of the four closures below, the books were handed to another shop, and buyers carried on reading there with a new account. The fourth outcome is the one Microsoft chose in 2019: it deleted every book it had sold and gave the money back.</p>\n<p>You don't get to pick. A book with DRM opens only in the apps the shop supports, under a licence that its servers grant, so when a shop gives up, it decides what becomes of its customers' libraries.</p>\n<table><thead><tr><th>Store</th><th>Closed</th><th>Where the books went</th><th>Books that couldn't follow</th></tr></thead><tbody>\n<tr><td>Sony Reader Store (US and Canada)</td><td>Late March 2014</td><td>Kobo</td><td>Not stated</td></tr>\n<tr><td>UK Nook Store</td><td>15 March 2016</td><td>Sainsbury's Entertainment on Demand</td><td>A Sainsbury's voucher</td></tr>\n<tr><td>Sainsbury's ebooks</td><td>Sales 1 October 2016, apps 30 November 2016</td><td>Kobo</td><td>Not stated</td></tr>\n<tr><td>Microsoft Store books</td><td>Sales 2 April 2019, books removed in July 2019</td><td>Nowhere</td><td>All were refunded</td></tr>\n</tbody></table>\n<h2 id=\"moving-to-another-shop\">Moving to another shop</h2>\n<p>Sony announced on 6 February 2014 that its Reader Store in the US and Canada would shut in \"late March\", and that every Reader account, library content included, would move to Kobo's platform over the same period, <a rel=\"external\" href=\"https://www.engadget.com/2014/02/06/sony-to-transition-its-reader-base-to-kobo/\">Engadget reported</a>. Sony promised customers an email that walked them through the migration.</p>\n<p>Barnes &amp; Noble's retreat from Britain was messier. The company told users in early March 2016 that the <a rel=\"external\" href=\"https://www.theverge.com/2016/3/4/11160942/barnes-noble-nook-app-store-shutting-down-march-15th\">UK Nook Store would close on 15 March</a>, along with the Nook App Store and Nook Video. British readers who wanted to keep their ebooks had to open an account with Sainsbury's, the supermarket chain, and for any book that couldn't be moved, the buyer was to get a Sainsbury's Entertainment voucher, according to <a rel=\"external\" href=\"https://diginomica.com/2016/03/07/bn-nukes-the-nook-with-a-15-march-deadline-for-customers-to-save-their-content/\">diginomica</a>. (The rest of the Nook story is in <a href=\"/blog/what-happened-to-the-nook/\">what happened to the Nook</a>.)</p>\n<p>The new home lasted about six months. On 20 September 2016 <a rel=\"external\" href=\"https://www.engadget.com/2016/09/20/sainsburys-close-entertainment-on-demand/\">Engadget reported</a> that Sainsbury's was closing Entertainment on Demand, its digital shop for ebooks, music, films, TV and magazines. The ebook store stopped selling on 1 October, the reading apps kept working until 30 November, and the site and apps closed for good on 1 December. Readers could transfer their libraries to Kobo, which the same report says had already taken in customers of Tesco and Sony. A British reader who bought Nook books and followed both transfers ended 2016 with a third account at a third company, holding whatever books had survived both moves.</p>\n<p>That Engadget report also shows what difference a download makes. Music buyers were told they could download their libraries and export them to iTunes or Google Play Music. The digital magazines disappeared altogether, with a refund promised by email.</p>\n<h2 id=\"refunded-and-deleted\">Refunded and deleted</h2>\n<p>On 2 April 2019 Microsoft <a rel=\"external\" href=\"https://www.theverge.com/2019/4/2/18292177/microsoft-ebooks-refund-stops-selling-digital-books-store\">stopped selling ebooks</a> in the Microsoft Store for Windows and said that bought books would be removed from libraries in early July, free ones too. Until then they could still be read in the Edge browser. Microsoft's own FAQ (<a rel=\"external\" href=\"https://web.archive.org/web/20200106061547/https://support.microsoft.com/en-us/help/4497396\">archived copy</a>) put it flatly: you could carry on reading bought books until July 2019, \"when they will no longer be available\", and free books until the same month, when they'd stop being accessible.</p>\n<p>The refunds needed no action from the buyer. Microsoft's FAQ, quoted by The Verge, said they would start rolling out automatically in early July 2019 to the original payment method; anyone who had paid with a gift card, or whose card was no longer valid, got credit on their Microsoft account instead. Highlights and notes went with the books. To cover them, Microsoft paid an extra $25 in account credit to anyone who had annotated a purchased book before 2 April.</p>\n<h2 id=\"files-you-hold-yourself\">Files you hold yourself</h2>\n<p>None of these closures could reach an ebook without DRM that the buyer had already downloaded. Such a file is an ordinary copy on your own disk, and any reader app that knows the format will open it, whatever happens to the shop that sold it. <a href=\"/blog/where-to-buy-drm-free-ebooks/\">Where to buy DRM-free ebooks</a> names the shops that sell that way. Kindle is a case of its own; <a href=\"/blog/do-you-own-your-kindle-books/\">do you own your Kindle books</a> goes through what Amazon's terms grant.</p>\n<p>Kindle buyers lost one way of keeping a copy on their own computer in 2025. From 26 February that year, <a rel=\"external\" href=\"https://www.theverge.com/news/612898/amazon-removing-kindle-book-download-transfer-usb\">Amazon removed the website option</a> to download a bought book to a computer and copy it to a Kindle over USB, so Kindle store books now reach an e-reader only over Wi-Fi. The details are in <a href=\"/blog/kindle-download-and-transfer-removed/\">why Kindle's Download &amp; Transfer via USB is gone</a>.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-an-acsm-file/","url":"https://readwyn.com/blog/what-is-an-acsm-file/","title":"What is an ACSM file, and why won't it open as a book?","summary":"An ACSM file is an Adobe download ticket for an ebook: what is inside it, why an old one stops working, and how Adobe Digital Editions turns it into a book.","content_html":"<p>An ACSM file is a download ticket for an ebook, and none of the book's text is in it. The letters stand for Adobe Content Server Message (<a rel=\"external\" href=\"https://www.fileinfo.com/extension/acsm\">FileInfo</a>), and the file is a short piece of XML that tells Adobe Digital Editions (ADE) where to fetch your ebook. That's why Apple Books or a PDF viewer finds nothing to show when you open it. Some e-readers can use the file themselves: <a rel=\"external\" href=\"https://support.pocketbook-int.com/fw/970/u/6.8.2909/manual/ww/User_Manual_InkPad_Lite_EN.pdf\">PocketBook's manual</a> says you copy it to the reader, open it, sign in with your Adobe account, and the reader downloads the book.</p>\n<p>Open one in a text editor and you can read all of it. Datalogics prints a sample in its <a rel=\"external\" href=\"https://kb.datalogics.com/article/e_adept_request_expired-ebook-download-error-from-ade-or-other-e-readers-73.html\">knowledge base article on expired links</a>: 28 lines of XML, wrapped in a <code>fulfillmentToken</code> element. It gives the address of the shop's fulfilment server, an order number and an expiry time, then the title, author and publisher in Dublin Core fields. A format line follows (<code>application/epub+zip</code> in the sample), then a list of permissions (display, excerpt, print, play) and an <code>hmac</code> check value on the last line before the closing tag. That format line tells you whether you bought an EPUB or a PDF before anything has downloaded.</p>\n<p>ADE takes the order to that server and gets back the actual book, a rights-protected EPUB or PDF that belongs to the ID you authorised ADE with. <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017814074\">Kobo's help page</a> sets out the steps for its readers: in ADE choose Help, then \"Authorize Computer\"; connect the Kobo by USB, find it under Devices and choose \"Authorize Device\" from the gear menu beside it; then drag the book across. Adobe's <a rel=\"external\" href=\"https://www.adobe.com/solutions/ebook/digital-editions/faq.html\">Digital Editions FAQ</a> says one ID can activate up to six computers and devices. You can also authorise a computer anonymously, but then every book you get afterwards belongs to that computer alone. Libraries and shops that use <a href=\"/blog/what-is-readium-lcp/\">Readium LCP</a> instead send a different small file, an <code>.lcpl</code> licence, and ADE plays no part in opening it.</p>\n<p>The expiry time is why an ACSM file that worked last month can fail today. Each shop sets its own link expiry on the server, 60 minutes by default when it is first set up as a distributor; the shop in the Datalogics sample had chosen 20 days. After that the server refuses the order and ADE shows <code>E_ADEPT_REQUEST_EXPIRED</code>. Your purchase is still there; only the ticket has run out. Download a fresh ACSM from your account on the shop or library site. For the same reason, an old ACSM file kept as a backup won't get you the book back.</p>\n<p>In Google Play Books you <a rel=\"external\" href=\"https://support.google.com/googleplay/answer/179863?hl=en\">export a book</a> with Download EPUB or Download PDF, and what arrives may be an ACSM file instead (the help page spells it \"ASCM\"). Some publishers block the export entirely, and rented books can't be exported.</p>\n<p>You may never see an ACSM file at all. <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017814074\">Kobo says</a> most books from its own store don't need ADE, and that many newer Kobo readers borrow library books directly. A shop that sells <a href=\"/blog/where-to-buy-drm-free-ebooks/\">DRM-free ebooks</a> sends a plain EPUB.</p>\n<p>Adobe no longer develops the system itself: its <a rel=\"external\" href=\"https://www.adobe.com/solutions/ebook/content-server.html\">Content Server page</a> says that in July 2025 support and development of Content Server, the Reader Mobile SDK and ADE moved to Wipro Engineering (formerly HARMAN Digital Transformation Solutions). From <a rel=\"external\" href=\"https://support.google.com/googleplay/answer/179863?hl=en\">1 June 2026</a>, ADE signs in with a ByteBooks ID and no longer accepts an Adobe ID. Google and Kobo both tell you to create the ByteBooks ID with the same email address as your Adobe ID so that your existing books stay linked. Kobo warns that a different address may mean downloading them again from the shop or library, and that the error <code>E_ADEPT_RESET_PW_REQUIRED</code> means you haven't switched yet. A computer or Kobo that ADE already authorised keeps working until ADE asks you to sign in again.</p>\n<p>The current build is ADE 4.5.12, which <a rel=\"external\" href=\"https://www.adobe.com/solutions/ebook/digital-editions/download.html\">Adobe's download page</a> offered for Windows and Mac as of November 2026, with links to apps for iOS and Android and no Linux version.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-daisy-format/","url":"https://readwyn.com/blog/what-is-daisy-format/","title":"What is a DAISY book?","summary":"DAISY is a standard for accessible books that hold narration, text, or both, with jumps by heading and page. What a DAISY book contains and how it relates to EPUB.","content_html":"<p>A DAISY book is a digital talking book whose text and narration are marked up by heading, page, and phrase, so a reader who can't use print can jump around it as a sighted reader flips pages. It grew out of a request that the Swedish Library of Talking Books and Braille (TPB) made to its government in 1991, for a three-year grant with two goals: fit more than 20 hours of continuous speech on one CD-ROM, and let a listener jump into the book from its table of contents. For an experienced listener, the consortium says, analogue talking books were very time-consuming to read (<a rel=\"external\" href=\"https://daisy.org/about-us/history/\">DAISY Consortium history</a>). The current version, DAISY 3, is the American standard ANSI/NISO Z39.86, <a rel=\"external\" href=\"https://daisy.org/z3986/2005/Z3986-2005.html\">approved on 21 April 2005</a>.</p>\n<h2 id=\"the-a-in-daisy\">The A in DAISY</h2>\n<p>The A once stood for audio. The software TPB commissioned in July 1993, by the consortium's account, was built on the \"Digital Audio-based Information System\" concept, whose core was storing speech phrase by phrase. RNIB, one of the libraries that founded the DAISY Consortium in Stockholm in May 1996, now spells it out as Digital Accessible Information System in its <a rel=\"external\" href=\"https://www.rnib.org.uk/living-with-sight-loss/independent-living/reading-and-books/daisy-guide\">guide to DAISY</a>.</p>\n<h2 id=\"audio-is-optional\">Audio is optional</h2>\n<p>The <a rel=\"external\" href=\"https://daisy.org/activities/standards/daisy/daisy-2/daisy-format-2-02-specification/\">DAISY 2.02 specification</a>, dated 28 February 2001, names six kinds of book. At one end is narration with only the title as text, which plays straight through and offers no way to jump. At the other is structured text with no audio at all. Between them are full audio synchronised with full text, and a book of full text with a little audio, such as a dictionary that speaks only its pronunciations. The 2005 standard describes the same span, from XML text only to audio with little or no text.</p>\n<p>Bookshare, Benetech's online library for people who are visually impaired, dyslexic, or otherwise <a rel=\"external\" href=\"https://www.bookshare.org/cms/about\">print-disabled</a>, shows the text end in practice: its plain <a rel=\"external\" href=\"https://bookshare.org/cms/help-center/which-formats-are-books-available\">DAISY downloads</a> are text files for text-to-speech players, reading apps, and braille notetakers. Its two audio options, MP3 and DAISY Audio, are both MP3 files made by its own speech engine, and the second is meant for DAISY players.</p>\n<h2 id=\"a-folder-of-files\">A folder of files</h2>\n<p>A DAISY book isn't a single file like an MP3. A DAISY 2.02 book has to contain exactly one navigation file named ncc.html (the Navigation Control Center) and at least one SMIL file. Audio files and XHTML text documents depend on the type: a text-only book has no audio, and a narrated book with only its headings as text needs no XHTML. The NCC lists every place a reader can jump to: headings from h1 to h6, and the page numbers of the printed edition. The SMIL files tie each stretch of text to a clip of audio by its start and end time in the recording.</p>\n<p>In DAISY 3 the navigation file is the NCX, an XML file that also gives direct access to page numbers, notes, and figures. The <a rel=\"external\" href=\"https://idpf.org/epub/20/spec/OPF_2.0.1_draft.htm\">OPF 2.0.1 specification</a> requires every EPUB 2 book to include an NCX, and it takes the format from section 8 of Z39.86-2005 with no change to its DTD. EPUB 3 replaced it with an HTML navigation document, but the <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/\">EPUB 3.3 specification</a> still describes the NCX as the table of contents for EPUB 2 reading systems, so an EPUB 3 book can carry one for older devices alongside its <a href=\"/blog/what-is-inside-an-epub-file/\">navigation document</a>.</p>\n<h2 id=\"no-special-player-needed\">No special player needed</h2>\n<p>You don't necessarily need a DAISY player. <a rel=\"external\" href=\"https://www.rnib.org.uk/living-with-sight-loss/independent-living/reading-and-books/daisy-guide\">RNIB's guide</a> says DAISY software on a computer will do, some CD players that handle MP3 files also understand DAISY, and a cheap USB player will play the sound but can't follow the structure. It still recommends a hardware player, and one of its DAISY CDs holds around 25 hours of audio. In the US, the National Library Service for the Blind and Print Disabled lets readers download its talking books to iOS and Android phones with <a rel=\"external\" href=\"https://www.loc.gov/nls/about/eligibility-for-nls-services/\">BARD Mobile</a>.</p>\n<h2 id=\"not-only-for-blind-readers\">Not only for blind readers</h2>\n<p>DAISY started in libraries for the blind, but the Z39.86 abstract names its readers as people who are blind or otherwise print-disabled. The <a rel=\"external\" href=\"https://www.loc.gov/nls/about/eligibility-for-nls-services/\">NLS eligibility page</a> lists a perceptual or reading disability among its categories, and a certified reading specialist, a school psychologist, or a librarian can confirm it. Cross-border sharing of these books among such readers is the subject of the <a href=\"/blog/marrakesh-treaty/\">Marrakesh Treaty</a>.</p>\n<h2 id=\"daisy-and-epub-3\">DAISY and EPUB 3</h2>\n<p>The DAISY site calls <a rel=\"external\" href=\"https://daisy.org/activities/standards/daisy/\">DAISY 3 the current version</a> and keeps <a rel=\"external\" href=\"https://daisy.org/activities/standards/daisy/daisy-2/\">DAISY 2</a> only for reference, as superseded, together with its recommendation of June 2003 on skippable structures. Production is moving, though: in October 2019 the consortium's board approved <a rel=\"external\" href=\"https://daisy.org/activities/projects/ties\">TIES</a>, a project to plan the move of its members from DAISY systems to EPUB 3.</p>\n<p>EPUB 3 keeps the synchronised narration under the name media overlays. <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-media-overlays\">EPUB 3.3</a>, republished as a W3C Recommendation on 13 January 2026 and co-edited by Matt Garrish of the DAISY Consortium, defines overlay files as a subset of SMIL 3.0 and says plainly that a book with overlays isn't an audiobook. In such a book a reader can switch off footnotes, endnotes, and page-number announcements (skippability) and leave a table, list, figure, or aside partway through (escapability). A reading system that doesn't support overlays shows the book as if they weren't there.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-djvu/","url":"https://readwyn.com/blog/what-is-djvu/","title":"What is a DjVu file, and why are so many scanned books in it?","summary":"DjVu is a format for scanned pages, begun at AT&T Labs in 1996, that keeps book scans small; the Internet Archive made DjVu copies of its books until 2016.","content_html":"<p>Many of the DjVu books you find online are Internet Archive scans. The Archive made a DjVu copy of each book uploaded to it until it <a rel=\"external\" href=\"https://archive.org/post/1053214/djvu-files-for-new-uploads\">announced in February 2016</a> that it would stop, and Wikisource's <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Help:IA-Upload\">help page on Internet Archive uploads</a> still names it as one of the main sources of the DjVu files that Wikisource works from. The format itself is older. A research team at AT&amp;T Labs started DjVu (said like the French déjà vu) in 1996, and its trick is to split each page into a sharp layer for the text and a soft layer for everything else and compress the two in different ways. A 300 dpi book scan then comes out several times smaller than the same pages stored as JPEG or TIFF images.</p>\n<h2 id=\"how-djvu-shrinks-a-scanned-page\">How DjVu shrinks a scanned page</h2>\n<p>The team's 1998 paper in the Journal of Electronic Imaging, <a rel=\"external\" href=\"https://leon.bottou.org/papers/bottou-98\">High Quality Document Image Compression with DjVu</a>, sets out the idea. Letters and line drawings need a high resolution to stay crisp. Photographs, illustrations and the tint of old paper change gently from pixel to pixel, and they survive being stored at a lower resolution. The encoder therefore pulls them apart.</p>\n<p>The text becomes a one-bit mask coded with JB2, which is based on AT&amp;T's proposal for the JBIG2 fax standard. <a rel=\"external\" href=\"http://yann.lecun.com/ex/djvu/index.html\">LeCun's DjVu page</a> describes how it works: JB2 finds shapes that repeat on a page, such as every lowercase \"e\" in one typeface, stores a prototype of each shape, and then stores only the list of places where it appears. The background goes through IW44, a wavelet coder, usually at a lower resolution than the text. Wavelet data is progressive, so a viewer can draw a rough page at once and sharpen it as more bytes arrive, which suited the aim that the 1998 paper states: sending document images quickly over slow connections.</p>\n<p>In the 1998 paper, a colour magazine page scanned at 300 dpi compressed to between 40 and 60 KB, which the authors put at 5 to 10 times better than JPEG for similar visual quality. LeCun's page gives 5 to 30 KB for a black and white page at 300 dpi, 3 to 8 times smaller than TIFF Group 4, the usual format for such scans at the time. At those rates, a 400-page book of black and white pages comes to roughly 2 to 12 MB.</p>\n<h2 id=\"what-is-inside-the-file\">What is inside the file</h2>\n<p>A DjVu file ends in <code>.djvu</code> or <code>.djv</code>. Its media type, <code>image/vnd.djvu</code>, has an <a rel=\"external\" href=\"https://www.iana.org/assignments/media-types/image/vnd.djvu\">IANA registration</a> created on 17 January 2002, with Léon Bottou of the AT&amp;T team as the contact, and it cites the original specification that AT&amp;T Labs dated 29 April 1999. The same record gives the signature that a program checks to recognise the file: the letters FORM at byte 4, then DJVU, DJVM, PM44 or BM44 at byte 12.</p>\n<p>A page can also carry a hidden text layer made by OCR, stored with the image so that a reader can search the book and copy words out of it. Wikisource's help page describes the Archive's processing in those terms: it ran OCR on the scans and turned the result into a DjVu file with embedded text. The page itself stays a picture, so a DjVu book is always fixed layout. It can't rewrap its lines for a small screen the way an EPUB can (see <a href=\"/blog/fixed-layout-vs-reflowable/\">fixed layout and reflowable ebooks</a>).</p>\n<h2 id=\"who-made-it-and-what-opens-it\">Who made it, and what opens it</h2>\n<p>The <a rel=\"external\" href=\"https://djvu.sourceforge.net/credits.html\">credits page of DjVuLibre</a> dates the start of the project to 1996 at AT&amp;T Labs Research, where Yann LeCun began it, with Léon Bottou, Patrick Haffner, Paul Howard and Yoshua Bengio doing much of the research. LeCun's page says LizardTech, a Seattle company, bought the technology from AT&amp;T in April 2000. In November 2000, according to the same credits, LizardTech published version 2.2 of the DjVu Reference Library under the GNU GPL, and later versions 3.0 and 3.5. DjVuLibre is Léon Bottou's clean-up of that released code, and the credits call him the original author of the library. On a computer, Okular, the KDE document viewer, <a rel=\"external\" href=\"https://okular.kde.org/formats/\">opens DjVu through DjVuLibre</a>.</p>\n<p>E-readers are a thinner field. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">list of supported file formats</a> covers PDF, the CBZ and CBR comic archives, and plain image files such as TIFF, with no DjVu. Two makers build it into their own software. The <a rel=\"external\" href=\"https://support.pocketbook-int.com/fw/700K3/u/6.8.3283/manual/User_Manual_Era_Color_EN.pdf\">user manual</a> of the PocketBook Era Color lists DJVU among its book formats, with the caveat that for PDF and DjVu you can only scale the whole page, since its reflow mode, which rewraps the text, works for PDF alone. Onyx's <a rel=\"external\" href=\"https://shop.boox.com/products/go7\">product page for the Boox Go 7</a> (read in November 2026) has DJVU in its list of document formats and says its other Boox devices read the same formats. On a Kindle or a Kobo, install KOReader, a document reader for e-ink devices that <a rel=\"external\" href=\"https://koreader.rocks/\">lists DjVu</a> among its formats and runs on Kindle, Kobo, PocketBook, Android and desktop Linux.</p>\n<h2 id=\"why-the-internet-archive-stopped\">Why the Internet Archive stopped</h2>\n<p>The forum post of 26 February 2016, signed \"brewster\" as the Archive's digital librarian, gave three reasons: use of the format had declined, new files were coming out with errors, and the Java viewer was hard to support. The DjVu files already made would stay, unless a later change to an item made them obsolete. The Archive said it would go on producing its <code>_djvu.xml</code> files, the OCR output with the position of each word, so the DjVu name stayed in the file lists of new items after the format itself had stopped.</p>\n<h2 id=\"djvu-or-pdf-for-a-scanned-book\">DjVu or PDF for a scanned book</h2>\n<p>LeCun's page carries no date, but its copyright line stops at 2004 and it still speaks of LizardTech as the company selling DjVu. It argued that PDF did not handle scanned documents well, colour scans above all. PDF gained the same tools afterwards. The <a rel=\"external\" href=\"https://opensource.adobe.com/dc-acrobat-sdk-docs/pdfstandards/PDF32000_2008.pdf\">ISO 32000 specification of PDF</a> marks the JBIG2Decode filter for one-bit images as a PDF 1.4 feature and the JPXDecode filter for JPEG 2000 as PDF 1.5. With those, a PDF can lay a sharp text mask over a compressed background, the method known as mixed raster content (MRC).</p>\n<p>The Internet Archive moved its own book PDFs to that method. Its open-source package, <a rel=\"external\" href=\"https://archive-pdf-tools.readthedocs.io/\">archive-pdf-tools</a>, splits each page into a background, a foreground and a mask, and compresses the mask without loss, with JBIG2 or CCITT fax coding. In a <a rel=\"external\" href=\"https://blog.archive.org/2020/11/23/foss-wins-again-free-and-open-source-communities-comes-through-on-19th-century-newspapers-and-books-and-periodicals/\">November 2020 blog post</a>, the Archive reported a compression factor of about ten for most of its books with this method.</p>\n<p>None of that touched the files the Archive had already made. Wikisource's help page dates the change to March 2016 and warns that a book uploaded since then may have no DjVu file, but the older ones are still what <a href=\"/blog/what-is-wikisource/\">Wikisource</a> works from most: the same page says DjVu is the file type it uses more than any other.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-fb2/","url":"https://readwyn.com/blog/what-is-fb2/","title":"What is an FB2 file, and will your e-reader open it?","summary":"FB2 (FictionBook 2) is a Russian ebook format that keeps a whole book in one XML file; KOReader, FBReader and PocketBook open it, but a Kobo does not.","content_html":"<p>Whether your e-reader opens an FB2 file depends on its reading software. FB2 is FictionBook 2, a format made in Russia that stores a whole book, pictures included, in a single XML file. KOReader, FBReader and PocketBook's readers take it as it is. A Kobo doesn't: its <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">list of supported formats</a> in November 2026 names EPUB, PDF, MOBI, plain text, HTML, RTF and the CBZ and CBR comic archives, and FB2 isn't on it. Nor is FB2 among the file types that Amazon's <a rel=\"external\" href=\"https://www.amazon.com/sendtokindle\">Send to Kindle</a> page accepts (the <a href=\"/blog/send-epub-to-kindle/\">Kindle post</a> has that list).</p>\n<p>Open one in a text editor and the format explains itself. The first line is an ordinary XML declaration, and then comes a <code>&lt;FictionBook&gt;</code> element with the namespace <code>http://www.gribuser.ru/xml/fictionbook/2.0</code>. That address is how software recognises the file. PRONOM, the format registry of the UK National Archives, identifies FB2 by exactly that pattern in an <a rel=\"external\" href=\"https://pronom.nationalarchives.gov.uk/fmt/2106\">entry added on 7 September 2026</a> (release V125, submitted by Preservica).</p>\n<h2 id=\"where-it-came-from\">Where it came from</h2>\n<p>The site in that namespace is run by Dmitry Gribov, who kept the <a rel=\"external\" href=\"http://www.gribuser.ru/xml/fictionbook/index.html.en\">FictionBook 2.0 specification</a> there and pointed readers to Mike Matsnev's HaaliReader for Windows CE. Gribov's <a rel=\"external\" href=\"http://www.gribuser.ru/xml/fictionbook/abstact2.0.html\">Russian overview</a> says about six developers of book software and around a dozen enthusiasts for well-made books took part. As of 12 January 2004, it adds, the active core would fit on the fingers of two hands, and the standard led among Russians who read on Windows CE pocket computers.</p>\n<p>A <a rel=\"external\" href=\"http://www.gribuser.ru/xml/fictionbook/whatfor.html\">second page</a> describes the mess it was meant to clear up. Lib.ru, which Gribov calls the main library of the Russian internet, kept its books as plain text at 80 characters a line, and that reads badly on a handheld screen 300 pixels wide. Books made for handhelds had the opposite problem: on a 1600 by 1200 desktop they showed up as a window of 300 by 600.</p>\n<p>FB2's answer was to record what each part of a book is and leave the look to the reading program. The <a rel=\"external\" href=\"http://www.fictionbook.org/index.php/Eng:FictionBook\">FictionBook community page</a> puts accurate structure first and easy conversion to TXT, DOC, RTF or HTML second. It's also frank about the limits: the format suits fiction, and scientific or technical books are still hard to describe in it. There is a <code>&lt;stylesheet&gt;</code> element for CSS, but MobileRead's <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/FB2\">FB2 article</a> says readers usually apply their own defaults instead.</p>\n<h2 id=\"what-the-xml-holds\">What the XML holds</h2>\n<p>The rules are a W3C XML Schema, which Gribov now publishes <a rel=\"external\" href=\"https://github.com/gribuser/fb2/blob/master/FictionBook.xsd\">on GitHub</a> under a BSD licence. A file has three parts.</p>\n<p>The <code>&lt;description&gt;</code> holds the catalogue data, in more detail than you might expect. Each <code>&lt;genre&gt;</code> must be a code from a separate <a rel=\"external\" href=\"https://github.com/gribuser/fb2/blob/master/FictionBookGenres.xsd\">genre list</a> of 186 entries, such as <code>adv_maritime</code> or <code>det_espionage</code>, and it can carry a <code>match</code> percentage for a book that only partly fits. A translation can describe the original edition in a block of its own, with the source language. The file also keeps its own history: who made this XML version, with which program, a version number to raise at each re-release, and an <code>id</code> that the schema says must never change. One optional element even marks which parts of the book are free and which are for sale, with a price and a currency.</p>\n<p>The <code>&lt;body&gt;</code> holds the text as nested <code>&lt;section&gt;</code> elements. Fiction gets tags of its own: <code>&lt;poem&gt;</code>, <code>&lt;stanza&gt;</code> and <code>&lt;v&gt;</code> for a line of verse, plus <code>&lt;epigraph&gt;</code> and <code>&lt;cite&gt;</code>. Footnotes sit in a second body named <code>notes</code>, and the text links to them.</p>\n<p>Pictures come last, in <code>&lt;binary&gt;</code> elements, each with an <code>id</code> and a <code>content-type</code> and written out as base64 text. Base64 turns every 3 bytes into 4 characters (<a rel=\"external\" href=\"https://www.rfc-editor.org/rfc/rfc4648.txt\">RFC 4648</a>), so an image takes about a third more room than the JPEG it came from. MobileRead notes that an FB2 file isn't compressed at all, which is why books often travel zipped as <code>.fb2.zip</code>.</p>\n<h2 id=\"which-programs-open-it\">Which programs open it</h2>\n<p>Calibre's <a rel=\"external\" href=\"https://manual.calibre-ebook.com/faq.html\">FAQ</a> has FB2 and the zipped FBZ among the formats it converts from and FB2 among those it writes, so a FictionBook file can become an <a href=\"/blog/what-is-inside-an-epub-file/\">EPUB</a> for a Kobo or a Kindle. Without converting, <a rel=\"external\" href=\"https://fbreader.org/\">FBReader</a> lists fb2 beside EPUB and MOBI, with apps for iOS, Android, Windows, macOS, Linux and ChromeOS, and KOReader names FB2 in its <a rel=\"external\" href=\"https://koreader.rocks/\">list of supported formats</a>. PocketBook's <a rel=\"external\" href=\"https://products.pocketbook.ch/\">product pages</a> listed both FB2 and FB2.ZIP for the Verse, the Verse Lite and the Verse Pro when I checked them in November 2026.</p>\n<p>In its home market, FB2 now has a successor. MobileRead says LitRes, the Russian ebook seller, used FB2 for years and has mostly used FB3 internally since about 2016. Gribov's <a rel=\"external\" href=\"https://github.com/gribuser/FB3\">FB3 repository</a> describes FictionBook 3 as an open format meant to replace EPUB or sit beside it.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-inside-an-epub-file/","url":"https://readwyn.com/blog/what-is-inside-an-epub-file/","title":"What is inside an EPUB file?","summary":"An EPUB is a ZIP archive: the mimetype file, META-INF/container.xml, the OPF package with its manifest and spine, XHTML chapters, CSS and the nav document.","content_html":"<p>An EPUB is a ZIP archive with a fixed skeleton: a file called <code>mimetype</code>, a <code>META-INF</code> folder with <code>container.xml</code> in it, an OPF file that lists everything else, and the book itself as XHTML chapters with their CSS, images and fonts. The <code>mimetype</code> file is 20 bytes long and holds <code>application/epub+zip</code> and nothing more, not even a line break. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-zip-container-mime\">EPUB 3.3 standard</a> says it must come first in the archive and must not be compressed, so every valid book shows the same bytes near its start: the <code>PK</code> that opens any ZIP, the name <code>mimetype</code> at byte 30 and the media type at byte 38, as the spec's <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#app-media-type\">media type registration</a> lists them. A program can tell an EPUB from an ordinary ZIP without unpacking anything.</p>\n<p>To see the rest, copy a book, change the ending to <code>.zip</code> and open it, or run <code>unzip -l book.epub</code> in a terminal. Only two names at the top level are reserved, <code>mimetype</code> and <code>META-INF</code>, and <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-container-metainf-files\"><code>container.xml</code></a> is the one file that <code>META-INF</code> must hold. Its job is small: a <code>rootfile</code> element gives the path of the OPF file. In the spec's <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#h-3-packaged-epub\">worked example</a> that's <code>EPUB/As_You_Like_It.opf</code>, but the folder can have any name. <code>META-INF</code> may also hold <code>encryption.xml</code>, which names any encrypted files and any embedded fonts whose first 1040 bytes have been scrambled so that they can't be lifted out and reused as they are; the standard calls this <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-font-obfuscation\">obfuscation</a> and says plainly that it isn't encryption.</p>\n<p>The OPF file does three jobs. Its metadata section must give a title, an identifier, a language and a <code>dcterms:modified</code> timestamp, and the <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-pkg-metadata\">spec</a> makes everything else optional, the author included (the fields are covered in <a href=\"/blog/epub-metadata-dublin-core/\">EPUB metadata and Dublin Core</a>). The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-manifest-elem\">manifest</a> lists every file the book uses, each with its media type. One item must carry <code>properties=\"nav\"</code>, and the cover picture can be flagged with <code>cover-image</code>. Then the <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-spine-elem\">spine</a> puts manifest items in reading order. An entry marked <code>linear=\"no\"</code> belongs to the book but sits outside the main run of text: the spec's examples are notes, descriptions and the answers to exercises, which a reader app may show in a pop-up or leave until the end.</p>\n<p>Chapters are XHTML, which is HTML written with XML's stricter syntax, so every tag has to be closed. Next to them sit CSS style sheets, images and fonts, in whatever folders the maker chose. The navigation document is one more XHTML page, and it must contain exactly one <code>nav</code> element of type <code>toc</code>: that's the table of contents. Books made for EPUB 2 used an NCX file for this instead, and EPUB 3 still allows one as a <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-opf2-ncx\">legacy feature</a>, so a book can carry both. Readwyn's reader builds its contents panel from whichever of the two the book has.</p>\n<p>A book that you've unpacked and edited has to go back together by the same rules: <code>mimetype</code> first and uncompressed, and every other file either stored or compressed with Deflate, the only two methods <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-zip-container-zipreqs\">the spec allows</a>. The standard itself names <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-conformance-checking\">EPUBCheck</a> as the checker the publishing industry relies on (<a href=\"/blog/check-epub-with-epubcheck/\">how to run it</a>).</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-project-gutenberg/","url":"https://readwyn.com/blog/what-is-project-gutenberg/","title":"What is Project Gutenberg?","summary":"Project Gutenberg is a free library of nearly 80,000 public-domain ebooks, begun by Michael Hart in 1971: its history, its copyright rules, and its formats.","content_html":"<p>Project Gutenberg is a volunteer-run library of public-domain ebooks, and on 1 November 2026 its <a rel=\"external\" href=\"https://www.gutenberg.org/\">home page</a> counted 79,708 of them. The books are free, and legally so in the United States. The project can give them away because it adds a book only after checking that its copyright has run out under US law. That's also why the notice at the top of each text, such as <a rel=\"external\" href=\"https://www.gutenberg.org/cache/epub/1342/pg1342.txt\">Pride and Prejudice</a>, tells readers outside the US to check the law of their own country first.</p>\n<h2 id=\"a-mainframe-account-in-1971\">A mainframe account in 1971</h2>\n<p>The project began at the University of Illinois. Michael Hart was given an operator's account on the Xerox Sigma V mainframe of the Materials Research Lab, with computer time that the project's <a rel=\"external\" href=\"https://www.gutenberg.org/about/background/history_and_philosophy.html\">history page</a> values at $100,000,000. He decided the most useful thing he could do with it was to type in a text and send it out over the network. The text he chose was the US Declaration of Independence. He named the project after <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Johannes_Gutenberg\">Johannes Gutenberg</a>, the printer.</p>\n<p>The date depends on which Gutenberg page you read. Hart's <a rel=\"external\" href=\"https://web.archive.org/web/20110917035457/http://www.gutenberg.org/wiki/Michael_S._Hart\">obituary on the project's site</a> puts the decision on 4 July 1971, after a free printed copy of the Declaration gave him the idea. The catalogue entry for <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1\">eBook #1</a> gives a release date of 1 December 1971, while <a rel=\"external\" href=\"https://www.gutenberg.org/cache/epub/1/pg1.txt\">Hart's own note</a> at the top of the text says \"early in 1971\". The same note describes the hardware: a disk pack the size of a cake in its carrier, which cost $1,500 and held 5 megabytes, of which the Declaration took 1 to 2 per cent.</p>\n<p>The Bill of Rights came next, then the Constitution, parts of the Bible, and Shakespeare one play at a time, according to the <a rel=\"external\" href=\"https://www.gutenberg.org/about/background/history_and_philosophy.html\">history page</a>. The catalogue dates the King James Bible, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/10\">eBook #10</a>, to August 1989 and The Complete Works of William Shakespeare, <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/100\">eBook #100</a>, to January 1994. Number 1000 went to Dante's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1000\">Divina Commedia</a>, in Italian, in August 1997. Hart's note in eBook #1 set the target of 10,000 files online by the end of 2001.</p>\n<p>Hart died at home in Urbana, Illinois, on 6 September 2011, aged 64. By then, his obituary says, the library had books in 60 languages.</p>\n<h2 id=\"which-books-it-may-publish\">Which books it may publish</h2>\n<p>Each book needs a copyright clearance, and the project publishes the <a rel=\"external\" href=\"https://www.gutenberg.org/help/copyright.html\">rules its volunteers follow</a>. The one it uses most is also the simplest: as of 2026, anything published in 1930 or earlier qualifies, wherever it first appeared, because the US term for those books is at most 95 years (<a href=\"/blog/when-do-books-enter-the-public-domain/\">when does a book enter the public domain?</a> has the history of that number). A second rule, Rule 6, covers US books published up to the end of 1963 whose copyright was never renewed. From the Library of Congress renewal records, the project estimates that over 85% of registered books were never renewed. It accepts Rule 6 research only from volunteers who have shown they can do it.</p>\n<p>US law isn't the law everywhere, and at the end of 2015 the difference reached a court. S. Fischer Verlag sued the Project Gutenberg Literary Archive Foundation in Frankfurt over 18 ebooks by Heinrich Mann, Thomas Mann and Alfred Döblin. All were in the public domain in the US, but Germany protects a work until 70 years after its author's death, and the three men died in 1950, 1955 and 1957. The court ruled on 9 February 2018 that blocking German addresses would do, and on 28 February the foundation blocked all of gutenberg.org for Germany. Under a settlement of October 2021 the site-wide block ended, and only the books by those three authors stay blocked in Germany until their German copyright expires: after 2020 for Heinrich Mann, after 2025 for Thomas Mann, and after 2027 for Döblin (<a rel=\"external\" href=\"https://cand.pglaf.org/germany/index.html\">the foundation's page on the case</a>). In November 2026 that leaves only Döblin's books.</p>\n<h2 id=\"what-you-download-today\">What you download today</h2>\n<p>Most new books now arrive from <a href=\"/blog/distributed-proofreaders/\">Distributed Proofreaders</a>, a volunteer site where people check a scanned book a page at a time; the Gutenberg home page calls it the main source.</p>\n<p>Every book page offers the same text in several files. The page for Pride and Prejudice (<a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1342\">eBook #1342</a>) counted 212,068 downloads in the 30 days before 1 November 2026. That page recommends EPUB3 for most devices and also lists plain text, two EPUB files for older e-readers (one of them without images), a KF8 file for Kindles, a file for older Kindles, and a zipped HTML copy. The sizes show what the illustrations weigh: the EPUB3 is 23.7 MB, the EPUB without images 595 kB, and the plain text 754 kB. If you're unsure which EPUB your reader wants, <a href=\"/blog/epub-2-or-epub-3/\">EPUB 2 or EPUB 3</a> explains the difference. Kindle owners are sent to Amazon's Send to Kindle site with the EPUB3 (the steps are in <a href=\"/blog/send-epub-to-kindle/\">how to send an EPUB to a Kindle</a>); Kobo and Nook owners are told to plug the reader in and copy the file over USB.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-readium-lcp/","url":"https://readwyn.com/blog/what-is-readium-lcp/","title":"What is Readium LCP, the newer ebook DRM?","summary":"Readium LCP is a DRM for ebooks that opens a book with a passphrase from the shop or library: what the .lcpl file is, how loans work, and how it differs from Adobe's.","content_html":"<p>Readium LCP (Licensed Content Protection) is a DRM system for ebooks, PDFs and audiobooks. The bookshop or library that sells or lends you a book gives you a passphrase, and that passphrase opens the book in any app that supports LCP. <a rel=\"external\" href=\"https://www.edrlab.org/readium-lcp/\">EDRLab</a>, a non-profit in Paris, certifies every app and server that uses it, and the system is published as the international standard ISO/IEC 23078-2:2024.</p>\n<h2 id=\"the-licence-file\">The licence file</h2>\n<p>What the shop or library hands you is usually a file ending in <code>.lcpl</code>: the licence, written as JSON, with the address of the book inside it. Under <a rel=\"external\" href=\"https://readium.org/lcp-specs/releases/lcp/latest\">the LCP specification</a>, the reading app downloads the book from that address and then puts the licence into the downloaded file at <code>META-INF/license.lcpl</code>, so you end up with one EPUB that carries its own licence. The licence also holds the hint that the app shows when it asks for your passphrase, and the licence's own ID encrypted with your key, which lets the app see at once whether the passphrase you typed is right. Limits are optional: start and end dates for a loan, the most pages you may print, and the most characters you may copy to the clipboard over the life of the licence.</p>\n<h2 id=\"the-passphrase\">The passphrase</h2>\n<p>You normally have one passphrase for each bookshop or library, either chosen by you or made by the shop; a library may use the number on your library card (<a rel=\"external\" href=\"https://www.edrlab.org/readium-lcp/principles/\">EDRLab's principles page</a>). Its example of the worst hint is \"Enter your passphrase\". A good hint is precise, such as \"Enter your favorite adventure book\", or one that tells you to type the 11 digits of the ID on your New York Public Library card. You type it the first time you open a book on a device, the app stores it, and the same passphrase then opens that provider's other books (<a rel=\"external\" href=\"https://www.edrlab.org/readium-lcp/faq/\">EDRLab's FAQ</a>). With automatic key retrieval, the app fetches the passphrase itself and you never type it. Checking it needs no connection, so a book still opens offline, even after the shop has closed.</p>\n<p>There's no account to open with anyone else: EDRLab's <a rel=\"external\" href=\"https://www.edrlab.org/readium-lcp/\">LCP page</a> says readers \"don't need to create a third-party account\". That's the main difference from Adobe's DRM, where an ID activates each computer or e-reader, up to a fixed number of them (the <a href=\"/blog/what-is-an-acsm-file/\">post on ACSM files</a> covers that system). EDRLab's FAQ calls a fixed limit such as six devices a bad idea, because people replace phones and share books with family. An LCP app that opens a book while online may instead register the device with the licence server, and the distributor can revoke a licence that shows an unusually large number of devices. Adobe's products, Adobe Digital Editions among them, don't recognise LCP-protected content (<a rel=\"external\" href=\"https://www.edrlab.org/readium-lcp/faq/\">FAQ</a>).</p>\n<h2 id=\"loans\">Loans</h2>\n<p>A loan runs on a second document, the <a rel=\"external\" href=\"https://readium.org/lcp-specs/releases/lsd/latest\">License Status Document</a>, which gives each licence a state (ready, active, returned, revoked, cancelled or expired) and the links that an app uses to register a device, renew a loan or return a book early. A return sets the licence's end date to now, and the book locks.</p>\n<h2 id=\"what-stays-closed\">What stays closed</h2>\n<p>The specification's Basic profile, with AES-256 encryption and SHA-256 hashing, is meant for testing only. Because it's openly documented it isn't secure, so real books use closed \"production\" profiles that only certified software gets (<a rel=\"external\" href=\"https://www.edrlab.org/readium-lcp/faq/\">FAQ</a>). Certified companies pay EDRLab a yearly fee set by their digital revenue (a server that issues over a million licences a year pays by volume instead, if that comes to more), and must also be members of EDRLab. In November 2026, the <a rel=\"external\" href=\"https://www.edrlab.org/readium-lcp/pricing/\">pricing page</a> asks €1,200 a year for a licence server run by an organisation with up to €300,000 of digital sales, and €300 for a mobile app, before the membership fee.</p>\n<h2 id=\"where-it-came-from\">Where it came from</h2>\n<p>EDRLab's <a rel=\"external\" href=\"https://www.edrlab.org/readium-lcp/overview/\">history of the project</a> traces it to a 2012 paper by Bill Rosenblatt for the IDPF, then the body behind EPUB, titled \"EPUB Lightweight Content Protection: Use Cases and Requirements\", where the L was for lightweight. The work moved to the Readium Foundation, where people from De Marque, Feedbooks and Learning Ally drafted a specification, and the New York Public Library helped design the part that handles loans. EDRLab took charge of the implementation in December 2015.</p>\n<p>In October 2024, EDRLab <a rel=\"external\" href=\"https://www.edrlab.org/2024/10/25/more-than-12-million-ebooks-in-84-countries-are-available-with-lcp-protection/\">counted 54 licence servers</a> and estimated that at least 12 million titles were available with LCP in 84 countries. The users it named include tolino in Germany, Bokus in Sweden, NetGalley, the National Library of Finland and the University of Chicago Press, and it said every publisher in France had adopted LCP. Its count of compatible readers came to 92 mobile apps, 13 desktop apps such as Thorium Reader and FBReader, one Braille device, and 21 e-ink readers from five brands: tolino, PocketBook, Vivlio, inkBook and Bookeen.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-tolino/","url":"https://readwyn.com/blog/what-is-tolino/","title":"What is Tolino, and is it a German e-reader?","summary":"Tolino is the e-reader brand of German booksellers, with technology from Rakuten Kobo since 2017: where it came from, what the readers do, and how Onleihe library loans work on them.","content_html":"<p>A tolino is bought in a German bookshop and opens on the ebook store of the shop that sold it, which is as German as an e-reader gets. The software and the cloud behind it are another matter. They have belonged to Rakuten, the owner of <a href=\"/blog/history-of-kobo/\">Kobo</a>, since January 2017, when Deutsche Telekom <a rel=\"external\" href=\"https://www.faz.net/aktuell/wirtschaft/netzwirtschaft/telekom-trennt-sich-ueberraschend-vom-e-reader-tolino-14601426.html\">sold the tolino platform to Rakuten</a> for a price that wasn't made public. Tolino's own <a rel=\"external\" href=\"https://mytolino.de/geschichte-der-tolino-allianz/\">history page</a> now names Kobo as its technology and innovation partner.</p>\n<h2 id=\"where-it-came-from\">Where it came from</h2>\n<p>Tolino <a rel=\"external\" href=\"https://mytolino.de/geschichte-der-tolino-allianz/\">reached the market in March 2013</a>, and in 2017 the FAZ recalled that it had been launched as Germany's <a rel=\"external\" href=\"https://www.faz.net/aktuell/wirtschaft/netzwirtschaft/telekom-trennt-sich-ueberraschend-vom-e-reader-tolino-14601426.html\">answer to the Kindle</a>. <a rel=\"external\" href=\"https://www.test.de/E-Book-Reader-Tolino-Shine-Allianz-gegen-Amazon-4523157-4523160/\">Stiftung Warentest's data sheet</a> for the tolino shine, dated 25 March 2013, lists its makers as Deutsche Telekom, Club Bertelsmann, Hugendubel, Thalia and Weltbild. It cost 99 euros, had a 1024 × 758 screen, and read EPUB, PDF and plain text, with Adobe DRM for protected books. Telekom ran the shared system for all the partners, and the wholesaler Libri later <a rel=\"external\" href=\"https://web.archive.org/web/20141026200024/http://home.libri.de/de/home/presse/pressemitteilungen/tolino-fuer-den-ganzen-buchhandel.html\">joined the alliance</a>, which brought the readers into the web shops of the more than 1,000 bookshops for which Libri ran the shop technology.</p>\n<p>The partners have changed since then. In November 2026 the <a rel=\"external\" href=\"https://mytolino.de/geschichte-der-tolino-allianz/\">tolino site</a> lists Thalia, Hugendubel, Osiander and Libri, with about 1,300 independent bookshops, and says the readers are also sold in Austria, Switzerland and the Netherlands. Weltbild went insolvent and <a rel=\"external\" href=\"https://www.netzwelt.de/news/233576-insolvenz-trifft-kindle-konkurrenten-tolino-muss-31-august-handeln-2208.html\">stopped trading on 31 August 2024</a>. Its tolino customers had to download the ebooks they'd bought there before that date, because Weltbild's servers were then switched off; later purchases could go through another tolino shop.</p>\n<h2 id=\"the-readers-today\">The readers today</h2>\n<p>You sign in on the reader with your account at any tolino bookseller, and the tolino cloud holds your purchases; tolino describes this as an <a rel=\"external\" href=\"https://mytolino.de/geschichte-der-tolino-allianz/\">open system</a> that doesn't bind you to one bookseller. The tolino shine color, announced on <a rel=\"external\" href=\"https://mytolino.de/news/innovationstreiber-im-buchmarkt-die-tolino-allianz-bringt-farbe-ins-digitale-lesen/\">10 April 2024</a> together with the black-and-white tolino shine and the 7-inch tolino vision color, was the first tolino with a colour screen. In that announcement the alliance said its partners had sold well over 5 million tolino readers since the start and put its market share at about 40 per cent. Weltbild was still listed as a partner, less than five months before it closed. Its <a rel=\"external\" href=\"https://mytolino.de/produkte/tolino-shine-color/\">product page</a> gives a 6-inch E Ink Kaleido 3 panel at 300 ppi for black-and-white content and 150 ppi for colour (the reason is explained in <a href=\"/blog/e-ink-carta-pearl-kaleido/\">Carta, Pearl and Kaleido</a>). It reads EPUB, MOBI, PDF, TXT, CBR, CBZ, JPEG and PNG, and comes with 25 GB of tolino cloud storage in Germany, Austria and Switzerland, synced to up to five devices.</p>\n<h2 id=\"borrowing-library-ebooks-through-the-onleihe\">Borrowing library ebooks through the Onleihe</h2>\n<p>The Onleihe lends ebooks from public libraries, and tolino readers can borrow from it directly. On the current generation (tolino shine 5th generation, shine color and vision color), the Onleihe has had its own menu entry <a rel=\"external\" href=\"https://mytolino.de/onleihe-neu/\">from software 5.15.253009</a>, and <a rel=\"external\" href=\"https://mytolino.de/anleitung-onleihe/\">tolino's guide</a> gives these steps:</p>\n<ol>\n<li>Connect to Wi-Fi, tap “Mehr” on the home screen, and choose “Onleihe”.</li>\n<li>Tap “Bibliothek auswählen”, type the name of your town or library, and pick it from the list.</li>\n<li>Sign in with your library account and tap “Anmelden!”.</li>\n<li>Find a title with the magnifying glass, tap “Ausleihen”, and confirm with “OK”.</li>\n<li>When you first open a protected loan, type the four-character code (letters, numbers or both) shown under “Buch direkt lesen”. The reader remembers it for later loans.</li>\n</ol>\n<p>Most Onleihe titles use <a href=\"/blog/what-is-readium-lcp/\">Readium LCP</a>, which <a rel=\"external\" href=\"https://mytolino.de/anleitung-onleihe/\">tolino's guide</a> says its readers have supported since mid-2019. An Adobe ID is now needed only for PDF titles, which use Adobe DRM (the system behind <a href=\"/blog/what-is-an-acsm-file/\">ACSM files</a>). To give a book back early, long-press its cover on the home screen or in “Meine Bücher” and choose “Zurückgeben”. Only ebooks can be borrowed on the device: audiobooks, ePaper and eMagazines from the Onleihe don't work there. Older tolinos have no Onleihe menu entry, but <a rel=\"external\" href=\"https://mytolino.de/onleihe-neu/\">their built-in browser</a> can open the library finder at mytolino.de/onleihe/.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-wikisource/","url":"https://readwyn.com/blog/what-is-wikisource/","title":"What is Wikisource, and can you download its books?","summary":"Wikisource is Wikimedia's library of free source texts, many of them proofread against page scans. Every work has a Download button for EPUB, MOBI, or PDF.","content_html":"<p>Wikisource is the Wikimedia Foundation's free library of source texts: books, laws, letters, and speeches that are out of copyright or under a free licence, many of them checked by volunteers against scans of the printed pages. You can download its books. Every work has a Download button that gives you an EPUB, a MOBI, or a PDF.</p>\n<h2 id=\"its-history\">Its history</h2>\n<p>Wikisource grew out of a quarrel on Wikipedia in 2001 over whether primary sources belonged in it, and a separate Project Sourceberg was suggested as their home. It <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Wikisource#History\">launched on 24 November 2003</a> under that name, a pun on Project Gutenberg, and a vote renamed it Wikisource on 6 December that year. By 2005 it had <a rel=\"external\" href=\"https://meta.wikimedia.org/wiki/Wikisource\">split into separate editions for each language</a>. Meta's table now lists more than 80 language editions, and texts in a language without one go to the shared site at wikisource.org. Six more, among them Alemannic and Swahili, live as a namespace inside their own language's Wikipedia, with no site of their own.</p>\n<h2 id=\"how-a-page-gets-checked\">How a page gets checked</h2>\n<p>A library that anyone can edit sounds like one full of typos. Its texts are built with the <a rel=\"external\" href=\"https://www.mediawiki.org/wiki/Extension:Proofread_Page\">Proofread Page extension</a>, which puts the scan of each printed page in one column and the OCR text beside it, so an editor (and later a reader) can compare the two line by line.</p>\n<p>Each page then carries a status. <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Help:Page_status\">The help page on page status</a> sets them out: “Not proofread” is the default, “Proofread” means one contributor has checked the page, and “Validated” means a second contributor has checked it too. You can't proofread and validate the same page yourself, and the Validate button only appears once somebody else has done the proofreading. Editors who aren't logged in can't change a status at all.</p>\n<h2 id=\"what-it-accepts\">What it accepts</h2>\n<p>The English edition's <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Wikisource:What_Wikisource_includes\">inclusion policy</a> says it no longer accepts new texts from Project Gutenberg, or any other second-hand transcription, even when someone has made a scan-backed file from that text. It wants contributors to proofread against a scan of the book instead, so the copy can always be checked against the physical edition.</p>\n<p>The same policy rules out new writing. Original compositions belong on other Wikimedia projects such as Wikibooks. Translations made by users are the main exception: English Wikisource accepts them, alongside published translations that are already in the public domain.</p>\n<h2 id=\"whose-copyright-law\">Whose copyright law</h2>\n<p>Wikisource takes texts that are in the public domain or under a licence that meets its free content definition, and its <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Wikisource:Copyright_policy\">copyright policy</a> says the laws that apply are mainly those of the United States, where the Wikimedia servers are. Fair use is explicitly banned. The inclusion policy lets in most writing published before 1931, as long as it can be verified, and it doesn't host extracts or partial works. A text that passes that American test can still be protected under <a href=\"/blog/public-domain-differs-by-country/\">the term of your own country</a>.</p>\n<h2 id=\"downloading-a-book\">Downloading a book</h2>\n<p>The Download button sits at the top right of the page, and the mobile site shows an EPUB icon in the same place. The <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Help:Reading_offline\">Reading offline help page</a> says to start from the top-level page when a work is spread over several pages, such as one page for each chapter. It suggests EPUB for most readers, MOBI for older Kindles, and PDF for printing or large screens.</p>\n<p>“More formats” in the same dialog opens WS Export. <a rel=\"external\" href=\"https://wikisource.org/wiki/Wikisource:WS_Export\">Its project page</a> says a user called Tpt wrote it for French Wikisource, and every other language edition can now use it. There, the same help page says, you can choose RTF, set the paper size of a PDF, pick another font, or leave out the images for a smaller file.</p>\n<p>Not every download comes out well. The help page warns that some works are unfinished and others have formatting that doesn't survive the move to an ebook. Volunteers collect the works that export cleanly in <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Category:Ready_for_export\">Category:Ready for export</a>, which held 713 pages on 1 November 2026.</p>\n<p>That category also feeds an <a href=\"/blog/what-is-opds/\">OPDS catalogue</a>, so a reader app that supports OPDS can browse and download those books directly. <a rel=\"external\" href=\"https://en.wikisource.org/wiki/Help:OPDS\">The OPDS help page</a>, illustrated with the catalogue open in KOReader, gives the address as <code>https://ws-export.wmcloud.org/opds/en/Ready_for_export.xml</code> and says the catalogue is rebuilt every day.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/when-do-books-enter-the-public-domain/","url":"https://readwyn.com/blog/when-do-books-enter-the-public-domain/","title":"When does a book enter the public domain?","summary":"A US book published before 1978 is protected for 95 years, a newer one until 70 years after its author dies, and many unrenewed books are already free.","content_html":"<p>On 1 January 2026, <em>The Maltese Falcon</em>, <em>As I Lay Dying</em>, <em>The Murder at the Vicarage</em> and the first four Nancy Drew books became free to copy in the United States. All of them came out in 1930, and Duke's Center for the Study of the Public Domain <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2026/\">explains why</a>: American works from that year were protected for 95 years, through the end of 2025. That's the rule for a US book published before 1978 whose copyright was kept up. A book written from 1978 onwards is protected until 70 years after its author dies, which is also the rule for most books in the UK and, under <a rel=\"external\" href=\"https://www.wipo.int/wipolex/en/text/443261\">Directive 2006/116/EC</a>, across the EU. The date always falls on 1 January, because a copyright term runs to the <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/305\">end of the calendar year</a> in which it would otherwise expire.</p>\n<h2 id=\"where-the-95-years-come-from\">Where the 95 years come from</h2>\n<p>The number is the result of three laws stacked on each other. Under the Copyright Act of 1909, a book got 28 years from publication and could be renewed for 28 more. The 1976 Act stretched the renewal term to 47 years, and the Copyright Term Extension Act of 1998 added another 20, so the full term for these older works grew from 56 years to 95. The Copyright Office's <a rel=\"external\" href=\"https://www.copyright.gov/circs/circ15a.pdf\">Circular 15A</a> walks through each step and draws the conclusion: everything published in the US before 1 January 1931 is now in the public domain.</p>\n<p>The 1998 extension also froze the US public domain for 20 years. Works from 1923 would have expired at the end of 1998, and instead stayed protected until the end of 2018. Duke's page for 2019 notes that it was <a rel=\"external\" href=\"https://cspd.law.duke.edu/publicdomainday/2019/\">the first year in over 20 years</a> in which published works entered the US public domain; that year's class included Agatha Christie's <em>The Murder on the Links</em>. Since then each <a href=\"/blog/public-domain-day/\">Public Domain Day</a> has freed another year of books, and renewed books from 1931 follow on 1 January 2027.</p>\n<h2 id=\"books-that-were-never-renewed\">Books that were never renewed</h2>\n<p>The 95 years apply only to a book whose copyright was renewed. Under the old system, the owner had to file a renewal in the 28th year, and if nobody did, the copyright expired at the end of the first term. Cornell's <a rel=\"external\" href=\"https://guides.library.cornell.edu/copyright/publicdomain\">copyright term chart</a> cites a Copyright Office study by Barbara Ringer, published in 1961, which found fewer than 15% of registered works were renewed, and only 7% of books. So a great many American books from 1931 to 1963 are already in the public domain, decades ahead of the 95-year date. The same chart treats a book published from 1931 to 1977 without a copyright notice as public domain too.</p>\n<p>Books from 1964 to 1977 don't have this gap: a law of 26 June 1992 made their renewal automatic. Many foreign books don't either, because the Uruguay Round Agreements Act restored the US copyright of certain foreign works that had lapsed for want of renewal, with effect from 1 January 1996 (both points are in Circular 15A).</p>\n<p>Checking a particular book means looking for its renewal record. The Online Books Page keeps a <a rel=\"external\" href=\"https://onlinebooks.library.upenn.edu/cce/\">guide to the scanned volumes of the Catalog of Copyright Entries</a>, where renewals were listed, and points to the Copyright Office's own online records for registrations from 1978 onwards.</p>\n<h2 id=\"books-from-1978-onwards\">Books from 1978 onwards</h2>\n<p>For works created from 1 January 1978, <a rel=\"external\" href=\"https://www.law.cornell.edu/uscode/text/17/302\">17 U.S.C. § 302</a> sets the term at the life of the author plus 70 years. A book with two authors runs until 70 years after the death of the one who lived longer. A novel by a writer who died in 1990 will therefore be free in the US on 1 January 2061. Anonymous and pseudonymous books, and works made for hire, follow a different clock: 95 years from first publication or 120 years from creation, whichever ends first. Circular 15A adds that if the real author of a pseudonymous book is later named in the Copyright Office's records, the term switches to life plus 70.</p>\n<p>Works created before 1978 but never published or registered were brought under the life plus 70 rule too, with a floor: none of them could expire before the end of 2002, and any that were published before that date stay protected through 2047.</p>\n<h2 id=\"outside-the-united-states\">Outside the United States</h2>\n<p>The UK government's <a rel=\"external\" href=\"https://www.gov.uk/copyright/how-long-copyright-lasts\">guidance on how long copyright lasts</a> gives written works 70 years after the author's death, and Duke notes that in countries with that term, works by authors who died in 1955 became free in 2026. The international floor is lower: <a rel=\"external\" href=\"https://www.wipo.int/wipolex/en/text/283698\">Article 7 of the Berne Convention</a> requires only the author's life plus 50 years. Because the US counts mostly from publication and Europe counts from death, a book can be free in one country and protected in the next, a gap covered in <a href=\"/blog/public-domain-differs-by-country/\">why the public domain differs by country</a>.</p>\n<p>A modern printing of an old text can carry a right of its own in the UK. The same government page lists a separate term for the layout of a published edition, 25 years from first publication, so a fresh typesetting of a Victorian novel is protected for that long even though the words are not.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/where-to-buy-drm-free-ebooks/","url":"https://readwyn.com/blog/where-to-buy-drm-free-ebooks/","title":"Where can you buy DRM-free ebooks?","summary":"Which shops sell only DRM-free ebooks, how Kobo and Amazon handle DRM-free titles book by book, and where free public-domain editions come from.","content_html":"<p>A handful of shops sell nothing but DRM-free ebooks, Smashwords and Baen among them, and they're the easier place to start. The big stores sell DRM-free books too, but title by title. At Kobo, Amazon and the other large stores, the publisher or the self-published author decides on DRM for each book, so one store sells both kinds, and Kobo's own help centre says <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019527954-Download-books-from-your-Kobo-account-to-export-to-another-device-or-app\">most books are protected by DRM</a>.</p>\n<h2 id=\"shops-that-sell-no-drm\">Shops that sell no DRM</h2>\n<p>Smashwords sells self-published books. Its <a rel=\"external\" href=\"https://www.smashwords.com/about\">about page</a> counts over a million DRM-free ebooks, with nearly 100,000 priced at free on any given day, and says the store has no current plans to sell books with DRM. Draft2Digital bought Smashwords in March 2022. Since January 2026 Smashwords has been a store only, its authors publish through Draft2Digital, and the page names Draft2Digital as the one distributor that supplies the store.</p>\n<p>Baen, the science fiction and fantasy publisher, sells its own ebooks direct. The <a rel=\"external\" href=\"https://www.baen.com/faq\">Baen FAQ</a> says every available Baen ebook is DRM-free and available worldwide, and it lists the formats: EPUB and MOBI sit alongside relics such as Sony's LRF and the Ebookwise/Rocket format. <a rel=\"external\" href=\"https://www.humblebundle.com/books\">Humble Bundle</a> sells books in themed bundles at a price you pick above a minimum. In early November 2026 nearly every book and comic bundle on that page carried a \"DRM-free\" label (the exceptions were tabletop role-playing products), but the formats changed from bundle to bundle: some were EPUB only, some PDF only, and a few added MOBI and PRC.</p>\n<p>For a longer list, the campaign site Defective by Design keeps a <a rel=\"external\" href=\"https://www.defectivebydesign.org/guide/ebooks\">guide to DRM-free ebook sellers</a>. It names Weightless Books and Book View Café, technical publishers such as No Starch Press, Manning and The Pragmatic Bookshelf, and shops that sell in French, German, Finnish and Japanese. A few of the sellers in the guide watermark their files rather than lock them; the post on <a href=\"/blog/social-drm-watermarks/\">social DRM</a> tells what such a mark holds.</p>\n<p>If a Baen or Humble download is a MOBI, Readwyn converts it to EPUB 3 when it imports the book and keeps the original file beside it. It doesn't convert a file with DRM.</p>\n<h2 id=\"kobo-and-amazon-one-book-at-a-time\">Kobo and Amazon, one book at a time</h2>\n<p>Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019527954-Download-books-from-your-Kobo-account-to-export-to-another-device-or-app\">download help page</a> sorts the files you can download into two kinds. Under My Account › My Books, the Actions icon beside a book leads to Download and then Download file. A protected EPUB or PDF needs Adobe Digital Editions before it can go onto another maker's device, while a DRM-free one can be copied straight across. Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017763713-File-formats-your-Kobo-eReader-and-Kobo-Books-app-support\">file formats page</a> adds that a book which comes as an <a href=\"/blog/what-is-an-acsm-file/\">ACSM file</a> is protected by DRM.</p>\n<p>Amazon now does something similar for self-published books. According to the KDP <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GDDXGH9VR22ACM8U\">help page on DRM</a>, verified purchasers have been able since 20 January 2026 to download EPUB and PDF files of books whose authors have confirmed them as DRM-free, from the Manage Your Content and Devices page. A book published before 9 December 2025 doesn't get the download automatically; its author has to confirm the choice first. Someone who borrowed the book can't download the files. If the author switches DRM back on, new downloads stop, but files already downloaded stay with the reader. That page is written for KDP authors, so it covers their books, not the catalogues of publishers that sell through the Kindle store. Amazon's website hasn't offered a file of those since it <a href=\"/blog/kindle-download-and-transfer-removed/\">removed Download &amp; Transfer via USB</a>.</p>\n<p>A publisher can make the same decision for its whole list. On 24 April 2012, Tom Doherty Associates <a rel=\"external\" href=\"https://reactormag.com/torforge-e-book-titles-to-go-drm-free/\">announced</a> that every ebook from Tor, Forge, Orb, Starscape and Tor Teen would be DRM-free by early July, sold through the retailers that already carried them. Doherty's reason was short: \"Our authors and readers have been asking for this for a long time.\" Tor UK announced the same move the next day.</p>\n<h2 id=\"free-editions\">Free editions</h2>\n<p><a href=\"/blog/standard-ebooks/\">Standard Ebooks</a> is a volunteer project that produces public-domain books and gives them away. Its <a rel=\"external\" href=\"https://standardebooks.org/about\">about page</a> says it dedicates its own work to the public domain too. Each book comes in four files, which its <a rel=\"external\" href=\"https://standardebooks.org/help/how-to-use-our-ebooks\">help page</a> explains: a compatible EPUB for most readers, an azw3 for Kindles, a kepub for Kobos, and an \"advanced\" EPUB that the project says isn't ready for general use yet. Its <a rel=\"external\" href=\"https://standardebooks.org/ebooks\">catalogue</a> listed 1,540 books on 1 November 2026. <a href=\"/blog/what-is-project-gutenberg/\">Project Gutenberg</a> gives away public-domain books too, in plainer editions.</p>\n<p>Baen gives some of its own books away as well. The <a rel=\"external\" href=\"https://www.baen.com/allbooks/category/index/id/2012\">Baen Free Library</a>, with an introduction by Eric Flint, held 82 titles on the same date, free to read online or to download.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/which-kindle-do-i-have/","url":"https://readwyn.com/blog/which-kindle-do-i-have/","title":"Which Kindle do I have?","summary":"Where a Kindle names its own model, how its serial number identifies older models, and the physical clues that separate Kindles with the same name.","content_html":"<p>A Kindle with recent software names its own model. Open Settings, then Device Options, then Device Info: since software version 5.13.7 of August 2021, the top of that window gives the model and its generation, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2021/08/30/new-quick-way-to-tell-which-kindle-model-you-own/\">The eBook Reader reported</a>. An older Kindle gives no model name there, but it does give its serial number, and a few characters near the start of that number name the model. Before 2021 that was the way to do it: <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2020/09/18/how-to-identify-which-kindle-model-you-own/\">the same blog's guide of 2020</a> sent owners to Device Info for the serial and then to a list of codes.</p>\n<p>The name on the box often won't settle it. Amazon has sold three different e-readers called plain \"Kindle\" since 2019. The one it announced in March 2019 has a 6-inch screen at <a rel=\"external\" href=\"https://press.aboutamazon.com/2019/3/all-new-kindle-now-with-an-adjustable-front-light-for-just-89-99\">167 ppi</a>. Its successor of September 2022 puts three times as many pixels on a screen of the same size, 300 ppi, and charges over <a rel=\"external\" href=\"https://press.aboutamazon.com/2022/9/introducing-the-all-new-kindle-and-kindle-kids-now-with-300-ppi-high-resolution-display-usb-c-charging-and-2x-storage\">USB-C</a>. The third came in <a rel=\"external\" href=\"https://press.aboutamazon.com/2024/10/amazon-launches-entirely-new-kindle-lineup-including-reimagined-kindle-scribe-and-first-ever-color-kindle\">October 2024</a>, again at 300 ppi, in a green that Amazon calls Matcha.</p>\n<p>Kindle owners keep a table of these codes on the <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Kindle_Serial_Numbers\">MobileRead wiki</a>, which also notes that the Your Devices tab of Manage Your Content and Devices in your Amazon account shows the serial, so the Kindle itself needn't be to hand. On older Kindles the serial starts with B0 (or 90), and the third and fourth characters are the model: B0<strong>24</strong> is the first Kindle Paperwhite with Wi-Fi only, B0<strong>D4</strong> the Wi-Fi Paperwhite of 2013. From the 2015 Paperwhite onwards the serial starts with a G, and the code moves to characters four to six: G09<strong>0G1</strong> is that 2015 Paperwhite, G0B<strong>0GC</strong> the first Kindle Oasis, G09<strong>10L</strong> the 2019 Kindle, and G00<strong>2AP</strong> the 2022 one. The table was last edited on 2 April 2024, and its newest e-ink entries are the 2022 Kindle and the first Kindle Scribe, so a model from late 2024 or 2025 isn't listed yet.</p>\n<p>One model can carry several codes, because Amazon gave each region and each radio its own. The Kindle of 2010 went on sale as \"the all-new Kindle\", <a rel=\"external\" href=\"https://press.aboutamazon.com/2010/7/announcing-a-new-generation-of-kindle-the-all-new-kindle-is-smaller-lighter-and-faster-with-50-percent-better-contrast\">$139 with Wi-Fi or $189 with free 3G</a>, and it's now usually called the <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Amazon_Kindle\">Kindle Keyboard</a>. Its Wi-Fi version is B008, the 3G version for the US and Canada B006, and the European 3G version B00A. The 2018 Paperwhite has more than ten codes, for its storage sizes, its 4G versions and its colours.</p>\n<p>A Kindle with a dead battery can still be named from its body. The earliest models (the original Kindle, the Kindle 2 and the DX) have the serial printed on the back near the bottom, and on the first Kindle it sits under the back cover, <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/Kindle_Serial_Numbers\">the MobileRead table notes</a>. That original, launched on <a rel=\"external\" href=\"https://press.aboutamazon.com/2007/11/introducing-amazon-kindle\">19 November 2007 for $399</a>, also has a full keyboard, which Amazon called standard-layout. The <a rel=\"external\" href=\"https://press.aboutamazon.com/2016/4/introducing-kindle-oasis-our-thinnest-and-lightest-kindle-ever\">2016 Oasis</a> is lopsided on purpose: a thick grip on one side with physical page-turn buttons, tapering to 3.4 mm at its thinnest point. For a Paperwhite, measure the screen across the diagonal. A 6.8-inch screen with a USB-C port marks the <a rel=\"external\" href=\"https://press.aboutamazon.com/2021/9/amazon-unveils-the-next-generation-kindle-paperwhite-and-new-kindle-paperwhite-signature-edition\">2021 model</a>, and a 7-inch one the <a rel=\"external\" href=\"https://press.aboutamazon.com/2024/10/amazon-launches-entirely-new-kindle-lineup-including-reimagined-kindle-scribe-and-first-ever-color-kindle\">2024 model</a>. The <a rel=\"external\" href=\"https://press.aboutamazon.com/2018/10/meet-the-all-new-kindle-paperwhite-thinner-lighter-2x-the-storage-and-waterproof-for-just-129\">2018 model</a>, the first waterproof Paperwhite, still has a 6-inch screen. A Kindle that charges on a Qi pad is from 2021 or later, since Amazon called the 2021 Paperwhite Signature Edition the first Kindle with wireless charging.</p>\n<p>The code matters most for the two oldest Paperwhites, which look alike. The first, announced on <a rel=\"external\" href=\"https://press.aboutamazon.com/2012/9/introducing-the-new-kindle-paperwhite-the-most-advanced-e-reader-ever-constructed\">6 September 2012</a> at 212 ppi, is on <a rel=\"external\" href=\"https://www.theregister.com/2026/04/08/amazon_kindle_support_discontinued/\">The Register's list</a> of the Kindles that lost the store on 20 May 2026; the 2013 model came a year too late for that list. So of two Wi-Fi Kindles both sold as \"Kindle Paperwhite\", B024 can no longer buy a book and B0D4 still can (<a href=\"/blog/old-kindle-end-of-support/\">what an unsupported Kindle can still do</a>). The 3G versions have codes of their own in the MobileRead table.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/which-kobo-do-i-have/","url":"https://readwyn.com/blog/which-kobo-do-i-have/","title":"Which Kobo do I have?","summary":"How to find your Kobo's model number in Settings, and how to tell every Kobo e-reader from the N416 of 2010 to the revised Clara BW of 2025 by its buttons and back.","content_html":"<p>Every Kobo e-reader has a model number, usually an N and three digits, and the device will tell you its own: tap More at the bottom of the Home screen, then Settings, then the last entry of the menu, \"About\" followed by the name of your eReader (Kobo gives these steps in its <a rel=\"external\" href=\"https://help.kobo.com/hc/de/articles/360017765753-Finden-Sie-Ihren-Kobo-eReader-Modellnamen-und-Ihre-Modellnummer\">German</a> and <a rel=\"external\" href=\"https://help.kobo.com/hc/fr/articles/360017765753-Rechercher-le-nom-et-le-num%C3%A9ro-de-votre-mod%C3%A8le-de-liseuse-Kobo\">French</a> help). Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019676973-Identify-your-Kobo-eReader-or-Kobo-tablet\">identification page</a> turns that number into a name for 25 e-readers, from N416 (the first Kobo, 2010) to N428 (the Libra Colour, 2024).</p>\n<p>The software version is one menu away, under Settings › Device information, in a form like 4.23.15444, <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360024773213-Find-your-eReader-s-software-version\">according to Kobo</a>. Readwyn's Sync screen also names a connected Kobo and draws it, reading the model from the file <code>.kobo/version</code> on the device. If the Kobo won't switch on, the identification page still helps: it gives a physical clue for each model (a button, a port cover, the finish of the back), and the sections below follow those clues. The release dates below are from Wikipedia's <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Kobo_eReader\">article on the Kobo eReader</a> unless another source is linked.</p>\n<h2 id=\"the-first-kobos-2010-to-2013\">The first Kobos, 2010 to 2013</h2>\n<p>Kobo was a spin-off of the Canadian bookseller Indigo, and it unveiled the Kobo eReader (N416) on 24 March 2010 at $149, <a rel=\"external\" href=\"https://www.cbc.ca/news/science/indigo-targets-amazon-with-kobo-e-reader-1.880917\">CBC News reported</a>. It couldn't download books over the air as the Kindle could: you loaded them over USB or Bluetooth, and 100 public-domain classics came preloaded. You'll know it by the blue navigation pad on the front. The Kobo Wireless (N647 or N47B), which followed in October 2010, has a black or white pad instead. On these two, Kobo says, the model number is on the bottom of the device.</p>\n<p>Kobo launched the Kobo Touch (N905, N905B or N905C) on 23 May 2011, as BookExpo America got under way, <a rel=\"external\" href=\"https://www.engadget.com/2011/05/23/dnp-kobo-unbuttons-for-129-ereader-touch-edition-we-go-hands/\">Engadget reported</a> from the show. It used Neonode's zForce infrared touch layer, and the only button left on its face was a home button under the screen (<a rel=\"external\" href=\"https://newatlas.com/kobo-ereader-touch-edition/18712/\">New Atlas</a>). In September 2012 the front-lit Glo (N613) and the 5-inch Mini (N705) came out together. The Glo has a light button next to the power slider on its top right edge. Then came the Aura HD (N204B), a limited edition with a 6.8-inch screen of 1440 × 1080 pixels, in April 2013, and the Aura (N514), with an <a rel=\"external\" href=\"https://en.wikipedia.org/wiki/Kobo_Aura\">edge-to-edge screen</a> and a flat back with creased edges, later that year.</p>\n<p>The Original, the Wireless and the Mini are on Kobo's list of <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360019690433-Products-no-longer-manufactured-or-supported-by-Kobo\">products it no longer supports</a>, with the Vox and Arc tablets, whose numbers start with K or T (K080 for the Vox, T416 for the Arc 7HD).</p>\n<h2 id=\"the-kobos-of-2014-to-2019\">The Kobos of 2014 to 2019</h2>\n<p>A cover on the bottom edge, hiding a Micro USB port and a microSD slot, means the Aura H2O (N250) of 2014. In 2015 Kobo sold two lookalikes with a silver power button on top, the 300 ppi Glo HD (N437) and the entry-level Touch 2.0 (N587). Turn them over to tell them apart: the Glo HD has a soft, rubbery back and the Touch 2.0 a plain plastic one with a light texture, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2015/09/08/kobo-touch-2-0-review-and-video-demo/\">The eBook Reader's review</a> noted.</p>\n<p>Blue power buttons on the back belong to three models. Two came in September 2016, the 7.8-inch Aura ONE (N709) and the 6-inch Aura Edition 2 (N236). The third, the Aura H2O Edition 2 (N867), also has two screws on its bottom edge; it went on sale on 22 May 2017 at $179, rated IPX8, <a rel=\"external\" href=\"https://www.theverge.com/circuitbreaker/2017/5/2/15520900/kobo-aura-h20-smaller-cheaper-e-reader\">The Verge reported</a>.</p>\n<p>The 6-inch Clara HD (N249), from June 2018, has its power button on the bottom edge. Later that year came the 8-inch Forma (N782), and in September 2019 the 7-inch Libra H2O (N873). Both have page-turn buttons; the power button is on the side of the Forma and on the back of the Libra H2O.</p>\n<p>Nine older models, from the Touch of 2011 to the Aura ONE of 2016, are on the same support page with a warning: if you update one after 25 May 2023, the update has to go through Kobo Desktop, and Device information should then show version 4.35 or newer.</p>\n<h2 id=\"from-2020-onwards\">From 2020 onwards</h2>\n<p>A power button on the bottom also marks the 6-inch Nia (N306) of July 2020, so check the model number to tell it from a Clara HD. The 10.3-inch Elipsa (N604) followed in June 2021 with the Kobo Stylus in the box. That October the 8-inch Sage (N778 or N778K) and the 7-inch Libra 2 (N418) became the first Kobos with Bluetooth, so you could listen to audiobooks on wireless headphones, and the Sage also took the stylus, <a rel=\"external\" href=\"https://www.techradar.com/news/kobos-new-ereaders-show-asymmetry-is-here-to-stay-and-were-all-for-it\">TechRadar wrote</a>. Both have page-turn buttons and a power button on the back (Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/4406285169175-Kobo-Libra-2\">Libra 2 page</a> has a picture of it). The 6-inch Clara 2E (N506) of September 2022 has its power button on the back as well, while the Elipsa 2E (N605) of April 2023 has its power button on the side.</p>\n<p>Kobo brought out the Clara BW (N365), the Clara Colour (N367) and the Libra Colour (N428) together, in shops from <a rel=\"external\" href=\"https://global.rakuten.com/corp/news/press/2024/0410_01.html\">30 April 2024</a>. All three have the power button on the side, which is the quickest way to tell a Clara BW from a Clara 2E. In November 2026 they're also the only models that Kobo lists as <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/21137184146071-Repair-your-Kobo-eReader\">repairable</a> with parts from iFixit. The Libra Colour's E Ink Kaleido 3 screen shows black and white at 300 ppi and colour at 150 ppi, by <a rel=\"external\" href=\"https://ereader.kobo.com/en-us/products/kobo-libra-colour\">Kobo's own specifications</a>.</p>\n<p>P365, the one e-reader number on Kobo's list that doesn't start with N, belongs to a revised Clara BW. Kobo didn't announce it. It went on sale around April 2025 with a bigger battery, and the FCC paperwork shows it was made by Top Victory Electronics rather than Netronix, which had built every Kobo since the first, <a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2025/06/10/kobo-released-revised-kobo-clara-bw-made-by-different-oem/\">The eBook Reader found</a>.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/who-invented-the-ebook/","url":"https://readwyn.com/blog/who-invented-the-ebook/","title":"Who invented the ebook?","summary":"No one person invented the ebook: Bob Brown, Ángela Ruiz Robles, Michael Hart, Alan Kay, and Sony each have a claim, depending on what you count as first.","content_html":"<p>On 7 December 1949, Ángela Ruiz Robles, a teacher in Ferrol in northern Spain, applied for a patent on a mechanical, electric, and air-pressure method for reading books. The <a rel=\"external\" href=\"https://www.oepm.es/export/sites/portal/comun/documentos_relacionados/PDF/AngelaRuizRobles.pdf\">Spanish Patent and Trademark Office</a> calls her the precursor of the ebook, but she's one of several people with a fair claim. Nobody invented the ebook alone, and the name you give depends on what you count: a machine for reading text that isn't printed on pages (Bob Brown in 1930, Ruiz Robles in 1949), a digital text that anyone on a network could copy (Michael Hart in 1971), a portable computer meant to replace books (Alan Kay in 1972), or a reading device that shops sold (Sony in 1990).</p>\n<h2 id=\"bob-brown-s-reading-machine-1930\">Bob Brown's reading machine, 1930</h2>\n<p>Bob Brown, a bestselling writer and publisher then living in France, set out his idea in <em>The Readies</em> in 1930. Craig Saper describes the machine in a <a rel=\"external\" href=\"https://dhq.digitalhumanities.org/vol/5/3/000108/000108.html\">2011 article in Digital Humanities Quarterly</a>. The text was printed very small on a roll, as one continuous line, and ran under a magnifying glass. A regulator set the speed, the reader could wind back or skip ahead, and moving the lens nearer or farther changed the apparent type size: Brown named 6, 8, 10, 12, or 16 point, or whatever suited the reader, which is what the font-size menu of every e-reader does today.</p>\n<p>Saper writes that in 1931 a group of expatriate writers met at Cagnes-sur-Mer to try a prototype built by Ross Saunders to Brown's plans, and that Gertrude Stein, Ezra Pound, and William Carlos Williams were among those who sent him texts written for it. Brown never found a manufacturer.</p>\n<h2 id=\"angela-ruiz-robles-and-the-mechanical-encyclopedia\">Ángela Ruiz Robles and the mechanical encyclopedia</h2>\n<p>Ruiz Robles wanted to spare her pupils the weight of their textbooks. Her design, as the <a rel=\"external\" href=\"https://www.oepm.es/export/sites/portal/comun/documentos_relacionados/PDF/AngelaRuizRobles.pdf\">patent office's profile</a> describes it, held its subjects on reels that passed under a transparent sheet, which could magnify the text, and the reader moved from topic to topic with a finger. A transparent cover inside, which could be lit, held other papers for reading. On the left, buttons formed letters from alphabets in several languages, and a plastic panel at the bottom served as a writing slate.</p>\n<p>She paid the fees on the 1949 patent until 1961 but couldn't afford to build the machine, and the education authorities gave it little support. In 1962 she filed a second patent, for an apparatus for readings and various exercises, which was also meant to help pupils with poor sight. This time a prototype was made at the artillery workshops in Ferrol: 24 by 22 centimetres and 6 centimetres thick. It's built of zinc, wood, and bronze. Where you can see it depends on whom you ask: <a rel=\"external\" href=\"https://www.telefonica.com/es/sala-comunicacion/blog/angela-ruiz-robles/\">Telefónica reports</a> that it's on display at Spain's National Museum of Science and Technology (MUNCYT) in A Coruña, while the patent office's profile says the Museo Pedagógico de Galicia in Santiago de Compostela has shown it since 2006. Her <em>Enciclopedia Mecánica</em> won prizes but never went on sale.</p>\n<h2 id=\"michael-hart-and-the-first-ebook-file-1971\">Michael Hart and the first ebook file, 1971</h2>\n<p>If an ebook has to be digital, the usual first is a plain text file. Michael Hart, a student at the University of Illinois, was given an operator's account on the Xerox Sigma V mainframe of the university's Materials Research Lab. His own <a rel=\"external\" href=\"https://gutenberg.org/about/background/history_and_philosophy.html\">history of Project Gutenberg</a>, written in 1992, puts $100,000,000 of computer time in that account and says he got it because two of the four operators happened to be his best friend and his brother's. He typed in the US Declaration of Independence and tried to send it across the network. He picked it partly for size: the file was only 5K, while a normal 300-page book needed a megabyte, far more storage than anyone could expect to have in 1971.</p>\n<p><a rel=\"external\" href=\"https://www.guinnessworldrecords.com/world-records/684329-first-ebook\">Guinness World Records</a> lists that file as the first ebook, typed on 4 July 1971: about 1,300 words, which anyone on ARPAnet could get once Hart had announced it. That file was the start of <a href=\"/blog/what-is-project-gutenberg/\">Project Gutenberg</a>.</p>\n<h2 id=\"alan-kay-s-dynabook-1972\">Alan Kay's Dynabook, 1972</h2>\n<p>A year later, Alan Kay of Xerox's Palo Alto Research Center described a device he called the DynaBook in \"<a rel=\"external\" href=\"https://www.mprove.de/visionreality/media/kay72.html\">A Personal Computer for Children of All Ages</a>\", a paper for the ACM National Conference in Boston in August 1972. He asked readers to treat it as science fiction. The specification still reads like a tablet's: no bigger than a notebook, under 4 pounds, a screen with at least 4,000 characters of print quality and contrast close to a page's, and a target price of $500, which he compared to a colour television. The drawing in the paper shows a slab 12 by 9 by 1 inches.</p>\n<p>Kay wanted a screen that drew power only when it changed and could be read by ambient light, which is how <a href=\"/blog/how-e-ink-works/\">e-ink</a> works; he admitted that no such flat, low-power screen existed yet. He also imagined vending machines where you could browse anything from encyclopaedias to \"<a rel=\"external\" href=\"https://www.mprove.de/visionreality/media/kay72.html\">the latest adventures of wayward women</a>\", but couldn't take a copy away until you'd paid.</p>\n<h2 id=\"the-first-ebook-reader-in-the-shops\">The first ebook reader in the shops</h2>\n<p>None of the four built a device that a shop sold. The first of those, by the reckoning of the <a rel=\"external\" href=\"https://collections.vam.ac.uk/item/O1325440/\">Victoria and Albert Museum</a>, was Sony's Data Discman of 1990. <a href=\"/blog/first-e-reader-rocket-ebook/\">What was the first e-reader?</a> picks the story up there.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/why-e-ink-flashes/","url":"https://readwyn.com/blog/why-e-ink-flashes/","title":"Why does an e-ink screen flash when you turn the page?","summary":"Why a Kindle or Kobo flashes black between pages, what ghosting is, how the flashing changed after 2011, and where the page refresh settings are.","content_html":"<p>The black flash is a full refresh. An e-ink screen makes its image from real pigment particles, and a fast update straight from one page to the next leaves some of them a little out of place, so a faint grey trace of the old page, called ghosting, shows through the new one. A full refresh drives every pixel through black and white and puts the particles back where they belong. Your reader does one every few pages, more often on pages with pictures, and on every page if you turn on the setting for it.</p>\n<h2 id=\"why-the-old-page-lingers\">Why the old page lingers</h2>\n<p>E Ink's engineers described the trade-off in a <a rel=\"external\" href=\"https://patents.google.com/patent/US7193625B2/en\">US patent filed in May 2003 and granted in March 2007</a>. One way to change a pixel, which the patent calls a slideshow waveform, first drives it to black and only then to its new shade. That keeps the electrical charge balanced, but a change can take up to twice the time the ink needs to switch. The other way goes straight from one grey to the next with no black or white flash. The patent also says where ghosts come from. The ink has a memory, as a <a href=\"/blog/how-e-ink-works/\">bistable screen</a> must: the result of a pulse depends on the states the pixel was in before, as well as the one it's in now, and that memory can lead to ghosting.</p>\n<p>Amazon's <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_colorsoft/v1/en-US/html/kug.html\">user guide for the Kindle Colorsoft</a> puts it more plainly. The particles have to move physically, and the occasional flash of the whole screen is the Kindle refreshing where the ink sits, so that text and images stay sharp. Kobo says much the same in its <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360062226733-Use-your-Kobo-eReader-as-a-notebook\">help page on notebooks</a>: if erased writing is still faintly visible, the “Refresh page” command flashes the screen black and white to clear it.</p>\n<h2 id=\"from-every-page-to-every-few\">From every page to every few</h2>\n<p>Early e-readers refreshed the entire screen on every turn. By 2011 the Kindle 4, like the Nook Touch and the Kobo Touch before it, redrew only the text on most turns; the Kindle flashed black once every six, and its page turns got faster and smoother for it. Not every owner liked the change, and at the start of October 2011 Amazon's firmware 4.0.1 for the Kindle 4 added a switch, on the third page of the settings menu, that brought the flash back to every page (<a rel=\"external\" href=\"https://blog.the-ebook-reader.com/2011/10/02/firmware-update-for-kindle-4-adds-page-refresh-setting/\">The eBook Reader, 2 October 2011</a>). It was only on or off. The Kobo Touch already let its owners pick how often the full refresh came.</p>\n<p>Two years later the ink itself changed. On 4 September 2013 <a rel=\"external\" href=\"https://www.eink.com/news/detail/E-Ink-Introduces-Carta-the-Next-Generation-of-Electronic-Ink-Display-Technology\">E Ink announced Carta</a>, an ink tuned for reading that could use its Regal waveforms, which the company said greatly reduce the need for full refreshes after a few page turns. Amazon chose Carta for that year's Kindle Paperwhite, the first e-reader to use it. Carta also brought more contrast, and E Ink has since put out <a href=\"/blog/e-ink-carta-pearl-kaleido/\">several numbered versions</a> of it.</p>\n<h2 id=\"the-settings-on-a-kindle-and-a-kobo\">The settings on a Kindle and a Kobo</h2>\n<p>A Kindle's default is to keep full refreshes to a minimum. On the Kindle Voyage, <a rel=\"external\" href=\"https://d1ergij2b6wmg5.cloudfront.net/kug/kindle_voyage/v1/en-US/html/kug.html\">Amazon's guide</a> puts the switch at Quick Actions, All Settings, Reading Options, Page Refresh. On the Colorsoft, Amazon's first <a href=\"/blog/colour-e-ink-for-reading/\">colour Kindle</a>, it sits under All Settings, Home and Library, Reading options, Page refresh. With it on, the screen refreshes on every turn. Comics have a separate Page Refresh On/Off switch in the menu you open while reading. The Colorsoft guide also warns that you'll see the flash more often on images, and on colour images in particular.</p>\n<p>Kobo's <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/32868920931991-Accessibility-Features-Design-Support\">accessibility statement</a> says its screens are designed to keep flashing down and refresh only when needed, partly to lower the risk of photosensitive reactions. Kobo's help centre has no page on the page-turn setting that I could find; owners document it. In a <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?p=4421778\">MobileRead thread from May 2024</a>, one owner gives the path as Settings, Reading settings, “Refresh the screen every”, and a Libra Colour owner lists the choices as 1 to 10 pages or every chapter. A <a rel=\"external\" href=\"https://www.mobileread.com/forums/showthread.php?t=370077\">later thread, from September 2025</a>, began with a Libra 2 owner asking what the “Page refresh” setting was for. One reply came from a reader who sets it to every page, because text grows paler over a run of partial refreshes. Another owner had raised the interval in the configuration file so far that the menu read “None selected”, and noted that opening the cover and turning the first page gives a full refresh anyway.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/why-ebook-files-are-large/","url":"https://readwyn.com/blog/why-ebook-files-are-large/","title":"Why are some ebook files so large?","summary":"Even War and Peace is a 1.7 MB EPUB. Big ebook files come from images, embedded fonts and the extra copies inside old Kindle files.","content_html":"<p>Project Gutenberg offers <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/1342\">Pride and Prejudice</a> as an EPUB of 595 kB and as an EPUB of 23.7 MB, with the same words in both. The big one reproduces the 1894 George Allen edition with Hugh Thomson's illustrations, as the title page and picture captions of Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/cache/epub/1342/pg1342-images.html\">illustrated HTML version</a> show, and the pictures account for nearly all of it. Most big ebook files work the same way. Text is small, so a file of 20 MB or more is carrying something besides the words: images most often, sometimes fonts, and in older Kindle files, extra copies of the book.</p>\n<h2 id=\"length-has-little-to-do-with-it\">Length has little to do with it</h2>\n<p>Gutenberg's <a rel=\"external\" href=\"https://www.gutenberg.org/ebooks/2600\">War and Peace</a> is a 1.7 MB EPUB. Its plain text file is 3.2 MB, nearly twice the size of the EPUB, because an EPUB is a ZIP archive and prose compresses well. The text-only Pride and Prejudice shows the same thing at a smaller scale: 754 kB as plain text, about a fifth less as an EPUB. A hundred books the size of War and Peace would take roughly the space of seven copies of the illustrated Austen, which is why the number of books a device holds depends more on what kind of books they are than on how long (see <a href=\"/blog/how-many-books-fit-on-an-e-reader/\">How many books fit on an e-reader?</a>).</p>\n<h2 id=\"where-the-23-7-mb-goes\">Where the 23.7 MB goes</h2>\n<p>The illustrated EPUB, as downloaded on 1 November 2026, is 24.9 million bytes long, which is the 23.7 MB that Gutenberg's page shows. Opened up, its 164 images (100 JPEGs and 64 PNGs) make up 24.5 million of those bytes. The ten XHTML files that hold the novel, George Saintsbury's preface and the contents come to 943 kB before compression and 310 kB inside the archive. The JPEGs average 242 kB each, so one illustration weighs about a third as much as the plain text of the whole novel.</p>\n<h2 id=\"zipping-doesn-t-shrink-pictures\">Zipping doesn't shrink pictures</h2>\n<p>An EPUB is a ZIP file, but ZIP can only squeeze what repeats. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-zip-container-zipreqs\">EPUB 3.3 specification</a> allows only two kinds of entry: stored, which means not compressed at all, and Deflate. Deflate works on text because text repeats itself. A JPEG has already been compressed by its own method, so there's little left for Deflate to find. Running all 100 JPEGs of the Austen file through Deflate at its highest setting saves 1.7 percent, and Gutenberg doesn't bother: every image in that EPUB is stored as it is. The XHTML, by contrast, shrinks to a third of its size.</p>\n<h2 id=\"fonts-can-outweigh-the-book\">Fonts can outweigh the book</h2>\n<p>Fonts for Chinese, Japanese and Korean can be the heaviest thing in a book. Google's Noto Serif CJK comes in <a rel=\"external\" href=\"https://github.com/notofonts/noto-cjk/blob/main/Serif/README.md\">several packagings</a>: the regular weight of <a rel=\"external\" href=\"https://github.com/notofonts/noto-cjk/blob/main/Serif/SubsetOTF/JP/NotoSerifJP-Regular.otf\">Noto Serif JP</a>, cut down to the characters used in Japan, is 6.2 MB, while the regular weight of <a rel=\"external\" href=\"https://github.com/notofonts/noto-cjk/blob/main/Serif/OTF/Japanese/NotoSerifCJKjp-Regular.otf\">Noto Serif CJK JP</a>, with the full character coverage, is 24.6 MB. GitHub shows the subset as 5.92 MB because it counts 1,024 kB to the megabyte. A book that embeds the regular and the bold weights of the subset carries 12.7 MB of font. When I ran both through Deflate at its highest setting, it took only about 15 percent off, so the pair still fills about 10.8 MB of the archive, some 35 times the compressed text of Pride and Prejudice. EPUB 3.3 lists <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-core-media-types\">WOFF and WOFF2</a> among its font formats, and both of those are compressed already: the <a rel=\"external\" href=\"https://www.w3.org/TR/WOFF2/\">WOFF2 specification</a> describes better compression than the Flate compression of WOFF 1.0.</p>\n<h2 id=\"kindle-files-can-hold-the-book-three-times\">Kindle files can hold the book three times</h2>\n<p>At Gutenberg, the Kindle files are the larger ones. Its Kindle files of the illustrated Pride and Prejudice are 25.2 MB (KF8) and 25.1 MB (older Kindles), and its KF8 War and Peace is 2.4 MB against 1.7 MB for the EPUB. Files built with KindleGen, Amazon's old command-line converter, could be far bigger, because it put three things in one file: a KF8 book, an old MOBI book and the source files it was made from. The <a rel=\"external\" href=\"https://wiki.mobileread.com/wiki/KindleGen\">MobileRead wiki</a> says that source record typically doubles the size of the file, and that the <code>-donotaddsource</code> option leaves it out. A file with both books in it is a <a href=\"/blog/mobi-kf8-combo-files/\">combo file</a>.</p>\n<h2 id=\"authors-pay-for-megabytes\">Authors pay for megabytes</h2>\n<p>On Amazon, file size costs money. Under the 70% royalty option, KDP's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200634500\">Digital Book Pricing Page</a>, last updated on 21 April 2026, takes a delivery cost off the list price before it works out the royalty: $0.15 per megabyte on Amazon.com, £0.10 on Amazon.co.uk and €0.12 on the German, French, Spanish, Italian and Dutch stores, with sizes rounded up to the nearest kilobyte. Amazon decides the size itself: the page charges for the megabytes “we determine your Digital Book file contains”. For sales in yen, the same page says Amazon deducts no delivery cost at all once a book reaches 10 MB.</p>\n<p>A file the size of the illustrated Austen would give up more than $3.50 of every US sale under the 70% option, before the royalty is worked out, and the text-only one about nine cents. The 35% option has no delivery cost, but Amazon's <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/G200634560\">list price rules</a> (November 2026) tie its minimum price on Amazon.com to size: $0.99 under 3 MB, $1.99 from 3 MB up to 10 MB, and $2.99 at 10 MB or more.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/why-ebook-fonts-look-different/","url":"https://readwyn.com/blog/why-ebook-fonts-look-different/","title":"Why does this ebook ignore my font setting?","summary":"Your font setting is only a default. If the book's stylesheet names a font on each paragraph, or a publisher font switch is on, the book keeps its own typeface.","content_html":"<p>You pick a new font on your Kobo, open a novel, and the text stays in the publisher's typeface, or the headings change while the paragraphs don't. The book's own stylesheet is the first place to look. If it names a font on the paragraphs themselves, your font setting is only a default for the text where the book names nothing. A \"publisher font\" option that's still switched on has the same effect. A related puzzle, the same book looking different from one app to the next, can come from an embedded font that one app can decode and another can't.</p>\n<h2 id=\"why-a-rule-on-a-paragraph-wins\">Why a rule on a paragraph wins</h2>\n<p>An EPUB is a bundle of web pages styled with CSS. If a reading app sets your font once, at the top of the document, the paragraphs inherit it. Under the <a rel=\"external\" href=\"https://www.w3.org/TR/css-cascade-4/#inheriting\">CSS rules</a>, a paragraph inherits a font only if the book's stylesheet gives it none. If the stylesheet says <code>p { font-family: \"Minion Pro\" }</code>, every paragraph keeps Minion whatever you pick. The same specification <a rel=\"external\" href=\"https://www.w3.org/TR/css-cascade-4/#cascade-origin\">ranks the book's normal rules above the reader's</a>; to beat them, an app has to mark its own rule <code>!important</code> or rewrite the book's CSS.</p>\n<p>Amazon asks publishers not to do this. Its <a rel=\"external\" href=\"https://kdp.amazon.com/en_US/help/topic/GH4DRT75GWWAGBTU\">KDP text guidelines</a> want the main font set at the <code>&lt;body&gt;</code> level and say \"Body text should not have a forced font face\", warning that readers may otherwise lose the choice. When Amazon processes a book, it moves the font family used in most of the content up to the root tag, and if fonts conflict with the reader's override, it strips the font files out.</p>\n<h2 id=\"the-publisher-font-switch\">The publisher font switch</h2>\n<p>Many apps have a setting that hands the choice back to the book. Amazon's <a rel=\"external\" href=\"https://www.glencoelibrary.org/wp-content/uploads/2023/02/Kindle_Paperwhite_Users_Guide.pdf\">Kindle Paperwhite guide from 2012</a> already listed a Publisher Font option in the Aa menu, shown only on books that embed fonts, and the KDP page says readers on KF8 devices and apps can turn publisher fonts on or off. Kobo's Web Reader puts a \"Publisher default\" toggle under Advanced Settings (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/35996239522967-Kobo-Web-Reader-Navigation-Reading-Features\">Kobo help</a>).</p>\n<p>At Apple, the publisher sets the switch inside the book. The <a rel=\"external\" href=\"https://help.apple.com/itc/booksassetguide/en.lproj/itc2cf4d26eb.html\">Apple Books Asset Guide</a> tells publishers that any book with embedded fonts must define <code>ibooks:specified-fonts</code>, and its example sets it to true. A reader can still choose another font, and <a rel=\"external\" href=\"https://help.apple.com/itc/booksassetguide/en.lproj/itc74d42b31e.html\">can always switch back</a> to the book's original ones.</p>\n<h2 id=\"what-to-try\">What to try</h2>\n<ol>\n<li>Pick a named font instead of the publisher option. On a Kobo that's the Font Face dropdown of the Text menu (<a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360017639913-Adjust-font-size-and-change-the-font-style-on-your-Kobo-eReader\">Kobo help</a>), where a font you <a href=\"/blog/add-fonts-to-kobo-and-kindle/\">added yourself</a> also appears. Kobo's help also warns that its OpenDyslexic font <a rel=\"external\" href=\"https://help.kobo.com/hc/en-us/articles/360020048733-Use-the-OpenDyslexic-font-on-your-Kobo-eReader\">may not work in some ebooks or languages</a>.</li>\n<li>If only some text changes, the stylesheet sets a font on the text that stays the same. In KOReader, turn off Embedded style and Embedded fonts: its <a rel=\"external\" href=\"https://koreader.rocks/user_guide/\">user guide</a> calls them on/off switches for the publisher's styles and fonts, with Style Tweaks for finer control.</li>\n<li>If the book is a DRM-free file of your own, open its CSS, delete the <code>font-family</code> lines on <code>p</code> and on paragraph classes, and keep one on <code>body</code>.</li>\n</ol>\n<p>Readwyn's reader takes only a short list of rules from a book's CSS (alignment, indents, margins, italic and bold, small caps, relative size, vertical alignment, and hidden elements). A font family isn't on that list, so the font picked in its Text panel sets the text of every book.</p>\n<h2 id=\"fonts-that-won-t-decode\">Fonts that won't decode</h2>\n<p>Some publishers obfuscate an embedded font so it can't be copied out by unzipping the EPUB. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-33/#sec-font-obfuscation\">EPUB 3.3 specification</a> defines how: the first 1040 bytes of the font are scrambled with a key, a SHA-1 hash of the book's unique identifier, and the file <code>META-INF/encryption.xml</code> marks the font as scrambled. The <a rel=\"external\" href=\"https://www.w3.org/TR/epub-rs-33/#x4-3-font-obfuscation\">reading system specification</a> says apps should undo this but doesn't require it, and the EPUB spec itself warns that such a book may look different from one app to the next. Where the font can't be decoded, the text falls back to another font, which may lack some characters (<a href=\"/blog/ebook-shows-boxes-instead-of-letters/\">why ebooks show boxes</a>). Because the key comes from the identifier, changing a book's unique identifier without obfuscating its fonts again leaves them unreadable in every app.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/why-library-ebooks-have-waitlists/","url":"https://readwyn.com/blog/why-library-ebooks-have-waitlists/","title":"Why do library ebooks have waiting lists?","summary":"Library ebooks are licensed one reader per copy, cost libraries far more than readers, and often expire. How the licences work and how to get a book sooner.","content_html":"<p>A popular library ebook has a waiting list because the library licensed a fixed number of copies, and each copy lends to one reader at a time, as a paper book does. Those copies are dear. In figures that the State of Maryland filed in a federal court, a New York Times best seller in December 2020 cost a reader $22.05 as an ebook and cost a library $95.00 (<a rel=\"external\" href=\"https://www.govinfo.gov/content/pkg/USCOURTS-mdd-1_21-cv-03133/pdf/USCOURTS-mdd-1_21-cv-03133-0.pdf\">AAP v. Frosh, opinion of 16 February 2022</a>).</p>\n<h2 id=\"one-copy-one-reader\">One copy, one reader</h2>\n<p>OverDrive's Libby app says in its help that publishers <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/0720.html\">often require each digital copy to go to one person at a time</a>, and that the library's budget decides how many copies it buys. The trade calls this one copy, one user. Rachel Noorda and Kathi Inman Berens, in a <a rel=\"external\" href=\"https://www.ala.org/sites/default/files/advocacy/content/ebooks/Digital-PL-Ecosystem-Report%20%281%29.pdf\">2023 report that the American Library Association publishes</a>, call OverDrive the dominant player in library ebooks and say libraries typically pay three to four times the consumer price for a licence to a popular title.</p>\n<p>The same limit applies when a library lends scans of print books that it owns, one reader per copy, a practice called <a href=\"/blog/controlled-digital-lending/\">controlled digital lending</a>. In 2024 a US appeals court <a rel=\"external\" href=\"https://storage.courtlistener.com/recap/gov.uscourts.ca2.60988/gov.uscourts.ca2.60988.306.1.pdf\">ruled</a> that the Internet Archive's version of it wasn't fair use.</p>\n<h2 id=\"licences-that-expire\">Licences that expire</h2>\n<p>Many of those licences also expire. The report by Noorda and Berens gives an early metered model: from 2011, HarperCollins licensed ebooks to libraries for 26 circulations, after which the licence ended. Penguin Random House moved to <a rel=\"external\" href=\"https://www.publishersweekly.com/pw/by-topic/industry-news/libraries/article/77904-penguin-random-house-changes-its-library-e-book-terms.html\">licences that expire two years after purchase</a> on 1 October 2018, and lowered its cap on new adult titles from $65 to $55 at the same time. A table in the report, built from OverDrive's data, shows US terms in July 2023: Hachette, Macmillan, Penguin Random House and Simon &amp; Schuster offered ebooks on two-year terms, and HarperCollins on 26 circulations.</p>\n<p>So a queue can come back. When a licence runs out, the library pays again or the copy goes. In Libby, a title that has expired shows <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6248.htm\">“Unknown Wait”</a> on the holds shelf, and the library may keep it visible and use the holds to decide whether to buy more copies.</p>\n<h2 id=\"embargoes-and-the-maryland-law\">Embargoes and the Maryland law</h2>\n<p>From <a rel=\"external\" href=\"https://www.ala.org/news/2012/01/ala-denounces-amazon-macmillan-response-congressional-inquiry-competition-digital\">1 November 2019</a>, Macmillan let each library buy a single copy of a new ebook and made it wait eight weeks for more. Macmillan went back to its earlier terms on <a rel=\"external\" href=\"https://www.ala.org/news/2012/01/ala-welcomes-cancellation-macmillan-embargo\">17 March 2020</a>, a change that Noorda and Berens tie to the Covid-19 lockdowns.</p>\n<p>Maryland tried to change the terms by law. From 1 January 2022, a publisher that licensed an ebook to the public had to offer it to the state's public libraries on reasonable terms. The Association of American Publishers sued, a federal judge blocked the law that February, and on <a rel=\"external\" href=\"https://www.govinfo.gov/content/pkg/USCOURTS-mdd-1_21-cv-03133/pdf/USCOURTS-mdd-1_21-cv-03133-1.pdf\">13 June 2022</a> the court declared it unconstitutional and unenforceable because the Copyright Act preempts it.</p>\n<h2 id=\"lending-without-a-queue\">Lending without a queue</h2>\n<p>Hoopla started a pay-per-use model in 2013, which OverDrive later offered too, according to the same report. The collection a library chooses is open with no waiting, and the library pays each time a patron completes a loan. Noorda and Berens name the drawback: paying per loan is cheap for a book few people borrow and very expensive for one that suddenly becomes popular.</p>\n<h2 id=\"getting-a-book-sooner-in-libby\">Getting a book sooner in Libby</h2>\n<ol>\n<li>Open Shelf &gt; Holds. The <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6104.htm\">estimated wait</a> sits next to each cover, with your place in line and the number of copies in use below the hold.</li>\n<li>If you have a card at a second library, <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6009.htm\">add it</a>. Then tap Available or Unavailable next to a title to see <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6167.htm\">its borrowing options at every library you've added</a>, and place the hold where the list is shortest.</li>\n<li>In Search, turn on the “skip the line” filter under “Catalog Filters”. <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6172.htm\">Skip-the-line copies</a> have no wait list and can't be renewed, and your library may give them a shorter loan.</li>\n<li>When a hold comes in, borrow it <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6198.htm\">within 72 hours</a>, or tap Suspend to keep your place. A <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6007.htm\">suspended hold</a> still moves up the list, and it lapses after 365 days in a row.</li>\n<li>If your library doesn't own the book, save it to a <a rel=\"external\" href=\"https://help.libbyapp.com/en-us/6115.htm\">Notify Me smart tag</a>. Where the library offers this, the tag tells it you want the title, and Libby notifies you if the library buys it.</li>\n</ol>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/why-you-forget-what-you-read/","url":"https://readwyn.com/blog/why-you-forget-what-you-read/","title":"Why you forget most of what you read, and what helps","summary":"Ebbinghaus published how fast memory fades in 1885. Later studies show that recalling beats rereading, and how far apart to review.","content_html":"<p>One day after learning a list, Hermann Ebbinghaus needed about two thirds of his original effort to learn it again. After a month he still saved about a fifth. He published those numbers in 1885, from years of tests on himself, and the shape of his curve is the one that matters for a book you've just finished: the steepest loss comes in the first hours, and what survives fades slowly after that. Reading a book once leaves you on that curve, and reading it again helps less than it feels. Pulling the content back out of your own head, and doing it again a few days later, helps far more.</p>\n<h2 id=\"what-ebbinghaus-measured\">What Ebbinghaus measured</h2>\n<p>He learned series of 13 nonsense syllables until he could recite them twice without a mistake, then learned the same series again after a gap of 20 minutes to 31 days and timed how much work the second learning saved. Chapter 7 of the <a rel=\"external\" href=\"https://psychclassics.yorku.ca/Ebbinghaus/memory7.htm\">1913 English translation of <em>Memory</em></a> puts the result in one sentence:</p>\n<blockquote>\n<p>After 24 hours about one third was always remembered; after 6 days about one fourth, and after a whole month fully one fifth of the first work persisted in effect.</p>\n</blockquote>\n<p>\"Remembered\" there means a saving in relearning time, not the share of the syllables he could still repeat, so a claim such as \"you forget two thirds of a book in a day\" reads more into the curve than it holds.</p>\n<p>The curve has held up. In 2015 Jaap Murre and Joeri Dros of the University of Amsterdam <a rel=\"external\" href=\"https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0120644\">repeated the experiment</a>, with Dros as the only subject, spending about 70 hours on lists relearned at the same intervals. Their results came out close to the original. Comparing Ebbinghaus's data with theirs and with an earlier German replication, they conclude that the curve most probably jumps upwards at the 24 hour point. They note that research on sleep would predict such a jump, though they don't claim sleep as its cause.</p>\n<p>Nonsense syllables are a harsh test, and Ebbinghaus knew it. He also memorised stanzas of Byron's <em>Don Juan</em>, 80 syllables each, and, as he reports in <a rel=\"external\" href=\"https://psychclassics.yorku.ca/Ebbinghaus/memory5.htm\">chapter 5</a>, needed hardly nine readings per stanza, about a tenth of what he expected a nonsense series of that length to take. Meaning, rhythm and rhyme made learning far cheaper. A novel is full of meaning too, but one pass through it with no return is the condition the curve describes.</p>\n<h2 id=\"why-rereading-feels-better-than-it-works\">Why rereading feels better than it works</h2>\n<p>The test case is from Washington University in 2006: 180 undergraduates, two short passages from a TOEFL preparation book, one on the sun and one on sea otters, and two ways of studying them (<a rel=\"external\" href=\"https://learninglab.psych.purdue.edu/downloads/2006/2006_Roediger_Karpicke_PsychSci.pdf\">Henry Roediger and Jeffrey Karpicke, Psychological Science</a>). In the second experiment of the paper, one group read a passage in four study periods of 5 minutes. Another read it once, then spent three periods writing down all they could recall, with no feedback. The rereaders got through the text about 14 times in all, the recall group about 3 times.</p>\n<p>Five minutes later, the rereaders recalled more: 83% of the ideas against 71%. A week later the order had flipped, with the recall group at 61% and the rereaders at 40%. The rereaders had also predicted that they'd remember the passage better. The authors suggest that students base such predictions on what pays off quickly.</p>\n<p>When John Dunlosky and four co-authors rated ten study methods in 2013, they put practice testing and spaced practice in its high-utility group and rereading and highlighting in the low one (<a rel=\"external\" href=\"https://europepmc.org/article/MED/26173288\">abstract on Europe PMC</a>). The evidence on highlighting has a post of its own: <a href=\"/blog/does-highlighting-help-you-remember/\">does highlighting help you remember?</a></p>\n<h2 id=\"how-far-apart-to-review\">How far apart to review</h2>\n<p>Ebbinghaus saw the spacing effect too. In <a rel=\"external\" href=\"https://psychclassics.yorku.ca/Ebbinghaus/memory8.htm\">chapter 8</a> he reports that 38 repetitions spread over three days did as much for a series of 12 syllables as 68 repetitions made on the day before the test.</p>\n<p>How long to wait depends on how long you want to remember. More than 1,350 people each learned a set of facts, got one review after a gap of up to 3.5 months, and took a final test up to a year later (<a rel=\"external\" href=\"https://europepmc.org/article/MED/19076480\">Cepeda and colleagues, Psychological Science, 2008</a>). The best gap grew as the test moved further away, but it shrank as a share of that delay: about 20 to 40% of a week, and about 5 to 10% of a year. For something you want to remember next week, that works out at a review after one and a half to three days; for next year, after roughly 18 to 36 days.</p>\n<h2 id=\"applying-it-to-a-book\">Applying it to a book</h2>\n<p>None of these studies used whole books, so carrying them over to a novel or a history is an inference, though not a long one. At the end of a chapter, close the book and write or say what happened and what the author argued before you look at anything. A list of highlights grouped by chapter makes a fair test sheet: the chapter heading is the prompt, and the highlights under it are the answer. Readwyn's Notebook <a href=\"/blog/all-highlights-in-one-notebook/\">lists them that way</a>, in reading order. Do the recall again two or three days later, and only then check the highlights for what you missed.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/zettelkasten-for-reading-notes/","url":"https://readwyn.com/blog/zettelkasten-for-reading-notes/","title":"How to take reading notes with a Zettelkasten","summary":"How Niklas Luhmann turned his reading into 90,000 linked slips, and how literature notes, permanent notes and ebook highlights fit his method.","content_html":"<p>Niklas Luhmann, the sociologist behind the best-known Zettelkasten, didn't write in it while he read. Between 1952 and early 1997 he filled about 90,000 slips, and the <a rel=\"external\" href=\"https://niklas-luhmann-archiv.de/nachlass/zettelkasten\">Niklas Luhmann Archive</a> in Bielefeld says he made them in a second step, from reading notes he had taken first. Taking notes for a Zettelkasten can follow the same two stages. While you read, note what the text says and where it says it; afterwards, write the cards: one thought each, in your own words, filed beside a card it answers and linked to others.</p>\n<h2 id=\"the-first-box-1952-to-1960\">The first box, 1952 to 1960</h2>\n<p>Luhmann started the box as a trainee lawyer in Lüneburg and kept it going as an official at the education ministry of Lower Saxony in Hanover. His first collection, which <a rel=\"external\" href=\"https://niklas-luhmann-archiv.de/nachlass/zettelkasten\">the archive</a> calls Zettelkasten I, holds about 22,000 slips from roughly 1952 to 1960, though notes on world society kept going into it until about 1973. He wrote on thin DIN A6 paper, often cut from used A4 sheets, and mostly on one side, so a slip could be read without lifting it out of the drawer.</p>\n<p>These early notes stay close to the books. The archive describes them as running text that follows the source, and topic sections often open with a list of things to read. The bibliography of this box was itself a list: just under 900 titles, sorted by author, on about 120 slips. The box had 108 numbered sections, among them organisation and law, authority, and causality.</p>\n<h2 id=\"the-second-box-from-1961\">The second box, from 1961</h2>\n<p>During a stay at Harvard, where he sat in on Talcott Parsons's sociology courses, Luhmann started again. The first box had stayed in Germany. Zettelkasten II grew to about 67,000 slips by early 1997, and its notes got shorter and more like theses.</p>\n<p>His reading notes changed with it. From about 1964 he gave every reference a slip of its own (around 11,000 of them), and on the back of many he wrote a brief excerpt in keywords. From that excerpt he wrote the real cards. If an excerpt ran long, it went on extra slips marked in red with the author and the year.</p>\n<p>Every card got a fixed number that never changed. A card that took up one point of card 1/1 became 1/1a and was filed between 1/1 and 1/2. A point raised on 1/1a became 1/1a1, and so on, until some numbers ran to 13 characters. He also wrote the numbers of related cards in other parts of the box onto a card, often with a link back. The archive's sample count puts these references at 25,000 to 28,000 in the second box alone. The way in was a keyword register of about 3,200 entries, and as a rule it named no more than four cards for a keyword: from there, the links led on.</p>\n<h2 id=\"the-method-after-luhmann\">The method after Luhmann</h2>\n<p>Luhmann wrote about the box himself, in an essay the <a rel=\"external\" href=\"https://niklas-luhmann-archiv.de/nachlass/zettelkasten\">archive</a> dates to the early 1980s, “Kommunikation mit Zettelkästen”. In <a rel=\"external\" href=\"https://luhmann.surge.sh/communicating-with-slip-boxes\">Manfred Kuehn's English translation</a>, bibliographic notes taken from a text go into the box, while the works he has actually read get slips in a separate box, so he can tell what he read from what he only meant to read. The slip for a book he'd read could also carry the numbers of the cards it led to. He thought excerpts were only useful for especially remarkable formulations.</p>\n<p>The labels most people search for came much later. Sönke Ahrens's <em>How to Take Smart Notes</em>, <a rel=\"external\" href=\"https://www.soenkeahrens.de/s/2022-HTTSN-Preview.pdf\">first published in 2017</a>, uses the names literature notes and permanent notes for two kinds of note. In the <a rel=\"external\" href=\"https://www.soenkeahrens.de/s/22-Appendix-1.pdf\">appendix to the 2022 edition</a>, he says one app can hold both, provided the types stay visibly apart and only the permanent notes form the linked network. He also advises writing each new note, where you can, as a direct reply to a note that's already there. Set against Luhmann's box, the literature notes are his keyword excerpts on the back of a reference slip, and the permanent notes are his numbered cards.</p>\n<p>Ebook highlights belong to the first stage. A list of highlights with page numbers gives you the source's words and where to find them again, which is the job Luhmann's excerpts did. In Readwyn, the <a href=\"/blog/all-highlights-in-one-notebook/\">Notebook</a> exports the highlights and notes of one book as a Markdown or plain-text file, with the title, the authors, each chapter title, and every highlight with its page and its note. That file is the raw material for a literature note. Ahrens says literature notes are short (<a rel=\"external\" href=\"https://notes.andymatuschak.org/zKzUzQENhyEDnuwPUhh2EQM\">as Andy Matuschak quotes him</a>), so you'd still cut the list down to what you might use. A highlight can't go into the box as it stands, because it's someone else's sentence: the permanent card restates one idea in your words and carries the number, or the link, of the card it replies to.</p>\n<h2 id=\"reading-luhmann-s-own-cards\">Reading Luhmann's own cards</h2>\n<p>You can look at the original slips. In April 2019 the archive put the first part of the box online, and <a rel=\"external\" href=\"https://www.heise.de/newsticker/meldung/Missing-Link-Luhmanns-Denkmaschine-endlich-im-Netz-4364512.html\">Heise reported</a> that about 4,000 cards of the first box were fully transcribed by then, with about 35,000 more up as images. The work is a project of Bielefeld University and the Cologne Center for eHumanities, planned to run for 15 years.</p>\n","date_published":"2026-11-03T04:32:13+05:30","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/3d-library/","url":"https://readwyn.com/blog/3d-library/","title":"Walk through your books in a 3D library","summary":"Readwyn can show your library as a room you walk through: the four sizes of room, the controls, how books are ordered on the shelves, and how to read a book from a shelf.","content_html":"<p>Press ⇧⌘L in the library and your books become a room you can walk through, every one of them on a shelf. You walk with W, A, S and D, look around with the mouse, and click a book to take it out. It works on macOS and on Linux.</p>\n<h2 id=\"enter-the-library\">Enter the library</h2>\n<p>Next to the grid and list buttons above your books there's a third one with an arch, “Enter the library”. The View menu has “Enter the Library” as well. If the library is empty, the button is dimmed and its tooltip says “Add books to fill your library”.</p>\n<p>The 3D library fills the window, with a small bar at the top left: “‹ Library”, the size of the room, the number of books, the search and “?”. Until the first frames are ready you see “Opening your library…”. A scene of 10,000 books appeared 0.79 to 0.92 s after ⇧⌘L when its lighting came from the cache, and after 1.89 s without it (the release app in a virtual Mac, with the library already loaded). The limit is 2 seconds. The light follows the app's scheme: daylight with Light, and dusk with the lamps lit with Dark.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library-3d-walk.webp?h=95c2ba59920dc6860cb4\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library-3d-walk.webp?h=4f3e17741e79a31319d8\" alt=\"The grand hall of the 3D library: a long hall with tall windows, armchairs on a rug, reading tables, and doorways to book rooms on both sides\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The grand hall of a library of 10,000 books</figcaption>\n</figure>\n<h2 id=\"the-four-sizes-of-room\">The four sizes of room</h2>\n<p>How big the room is depends on how many books you have:</p>\n<table><thead><tr><th>Room</th><th>Books</th></tr></thead><tbody>\n<tr><td>Reading nook</td><td>1 to 200</td></tr>\n<tr><td>Home library</td><td>201 to 1,200</td></tr>\n<tr><td>Gallery library</td><td>1,201 to 5,000</td></tr>\n<tr><td>Grand hall</td><td>5,001 and more</td></tr>\n</tbody></table>\n<p>The nook is a single built-in bookcase by a window, with an armchair. The home library is a room with one wall of shelves and a ladder. The gallery library is one long room lined with shelves on both sides, and the grand hall has book rooms opening off it; 10,000 books fill about 70 bays in 9 rooms. Shelves are only built for the books you have, so the extra floor space of a bigger size is for walking around, not for empty bays.</p>\n<p>When the library grows into the next size up, a card tells you: “Your library has outgrown the gallery library. Welcome to the grand hall.” You can also give a library a bigger room than its count calls for, in Settings › Libraries.</p>\n<h2 id=\"the-controls\">The controls</h2>\n<ul>\n<li>W, A, S, D, and the up and down arrows walk. The left and right arrows turn.</li>\n<li>Shift walks faster. Space jumps.</li>\n<li>C crouches to the low shelves. V rises on tiptoe for the high ones.</li>\n<li>A click locks the pointer and shows a crosshair. The mouse, or one finger on the trackpad, turns the view.</li>\n<li>“/” opens the search. “?” shows the controls again.</li>\n<li>Esc frees the pointer. With the pointer free, Esc or ⌘[ goes back to the library screen.</li>\n</ul>\n<p>“Look speed” in Settings › 3D library sets how fast the view turns.</p>\n<h2 id=\"find-a-book\">Find a book</h2>\n<p>Books are shelved by genre family, then genre, then author surname. An author's series stays together as one block in series order, and the blocks and standalone books run by year of first publication. Signs over the rooms name their genres.</p>\n<p>Press “/” to search titles, authors and series. You get the first 8 matches, each with its section, and Return glides you over to the book and takes it out. Clicking a sign or a line of a directory glides you there too.</p>\n<p>If a book is in the wrong section, click “Edit details” on its book screen and pick a genre under “Section in the 3D library”.</p>\n<h2 id=\"take-a-book-out\">Take a book out</h2>\n<p>Put the crosshair on a book and click. It slides out into your hand, and a panel shows its family and genre, title, authors, series, year, pages, and status with progress.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library-3d-book.webp?h=24fffa38be9a43bc2d3e\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library-3d-book.webp?h=e360d193888acff796b2\" alt=\"A book taken out of a shelf, held with its cover forward, and a panel with Fiction, Literary fiction, the title, the author, the year, the pages, the status To read, and the buttons Read and Details\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A book out of its shelf, with its panel</figcaption>\n</figure>\n<p>“Details” opens the book screen; “Put back”, Esc or a movement key returns the book to its shelf. “Read” doesn't open the reader straight away. First you carry the book to the nearest reading table, desk, or side table by an armchair and sit down. The book opens at the page where you stopped (or at the start, for a new book), and then the reader takes over. With no table nearby, the book opens in your hand. If you don't want to wait for the walk, a click, Return or Esc skips straight to the reader.</p>\n<p>Leave the reader and you're back where you were standing, with the book on its shelf again.</p>\n<h2 id=\"markers-on-the-shelves\">Markers on the shelves</h2>\n<p>You can tell a book's state from the shelf. The book you're reading stands 4 cm forward with a ribbon hanging below it, and the longer the ribbon, the further you've got. An unread book has a paper slip sticking out of its top. A finished book has no marker. Next up books get their slip in the app's accent colour.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/add-books-by-dragging/","url":"https://readwyn.com/blog/add-books-by-dragging/","title":"Add books by dragging them onto the window","summary":"Drag ebook files from Finder onto the Readwyn window: what the overlay tells you before the drop, which files it skips and where the new books show afterwards.","content_html":"<p>Any file that “Add books” accepts can also be dragged from Finder onto the Readwyn window. Before you let go, an overlay says how many books it will add and to which library, and marks the files it will skip. After the drop, the new books sit at the front of the library, each with a ring.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library.webp?h=b388e029cc1f91357982\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library.webp?h=1ae4e63639f580a708ea\" alt=\"The library in the grid view: the status tabs, the Filter and Sort controls, the Add books button, and two rows of covers\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The library, with “Add books” at the top right</figcaption>\n</figure>\n<h2 id=\"what-readwyn-accepts\">What Readwyn accepts</h2>\n<p>You can drop EPUB, MOBI (AZW3, combo files and PalmDOC included), PDF, OEB 1.x packages (a ZIP file with a package file), and the comic archives CBZ, CB7 and CBT. CBR archives aren't read.</p>\n<p>Readwyn reads the metadata embedded in each file. A book in an older format also gets a converted EPUB 3 next to its original, so it opens in the Readwyn reader (<a href=\"/#formats\">which formats convert, and how</a>). A file with DRM doesn't convert, so it keeps only its original file.</p>\n<h2 id=\"while-you-re-dragging\">While you're dragging</h2>\n<p>The overlay names the count and the current library, for example “Drop to add 3 books to Fiction”, and lists the files. Each file that “Add books” would skip is marked “Skipped”, and past five files a line counts the rest (“and 7 more”).</p>\n<p>Sometimes a drop can't do anything, and the overlay tells you so before you let go. If none of the files can be added, it says “Readwyn can't add these files” and lists the formats it takes. With no library open, it says “Open a library to add books”. If the edit details sheet has unsaved changes, it asks you to “Save or cancel your changes first”. In all three cases, dropping changes nothing.</p>\n<h2 id=\"after-the-drop\">After the drop</h2>\n<p>The window switches to the library, and a message with a check says what came in: “Added 2 books to Home library. 1 file skipped.”</p>\n<p>As long as that message shows, the library lists every book, newest first. The search and the filters clear, the line “Sorted by date added, newest first” replaces the Filter, Sort and view controls, and each new book has an accent ring. The library scrolls to the first new book and pulses every book you dropped, even one that was already there. Change the tab, the sort, the search or a filter and the library goes back to normal.</p>\n<p>A comic archive with no page images isn't added, and the message names it: “Skipped “no-pages.cbz”: it has no page images.”</p>\n<p>Files are stored by their content hash and never changed after that, so if the details that came from a file are wrong, you fix them in the library, in the <a href=\"/blog/edit-ebook-metadata-on-mac/\">edit details sheet</a>. On the book's Details tab, “In the library” shows the date it arrived, marked “added”.</p>\n<h2 id=\"into-an-empty-library\">Into an empty library</h2>\n<p>An empty library shows the “Welcome to Readwyn” page, and you can drop files there as well, or click “Choose files…” under “Add books”. If you're coming from calibre, <a href=\"/blog/import-calibre-library/\">import the library</a> instead of dragging its files, so your IDs, covers, ratings and Kobo data come with it.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/all-highlights-in-one-notebook/","url":"https://readwyn.com/blog/all-highlights-in-one-notebook/","title":"All your highlights in one notebook","summary":"Readwyn’s Notebook gathers the highlights and notes of every book in your library by chapter, with colour filters, search and export to Markdown or plain text.","content_html":"<p>Click “Notebook” in the sidebar and every highlight and note in the library is there, book by book. The books with highlights are listed on the left. Click one, and its highlights fill the right side in reading order, grouped by chapter. <a href=\"/blog/highlight-and-take-notes-in-epub-on-mac/\">Highlights you made in the reader</a> and highlights that came in from <a href=\"/blog/export-kobo-highlights-on-mac/\">a Kobo sync</a> sit in the same list.</p>\n<h2 id=\"finding-your-way-around\">Finding your way around</h2>\n<p>The Notebook’s toolbar switches between two views, “Highlights” and “My Words”, each with a count. The number beside “Highlights” covers the whole library, and a highlight with a note counts once.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-notebook.webp?h=c5d907fb6a66e4603293\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-notebook.webp?h=758885220be4b17df8b7\" alt=\"The Notebook: the books with highlights at the left, and the highlights of Middlemarch at the right, by chapter, with colour chips and a note\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The highlights of one book, by chapter, each with its page</figcaption>\n</figure>\n<p>Each book on the left shows its cover and counts such as “14 highlights · 3 notes”. Above its highlights on the right you get the title, the authors, the same counts and three buttons: “Copy all” copies every highlight in the book, “Export” saves them to a file, and “Open book” opens the reader at your saved place.</p>\n<p>Below the title is a row of colour chips, “All 14”, “Yellow 6” and so on. Click a chip to see only that colour.</p>\n<p>Every highlight has a dot in its colour, its text marked in that colour, its note underneath and its page on the right. That’s the printed page if the book has a page list. If it doesn’t, you get an estimate written as “~p. 196”, and a book with no page count at all shows a percentage instead.</p>\n<p>The Highlights tab on each book screen shows the same list for one book, with the same colour chips, “Copy all”, and an “Open in Notebook” button.</p>\n<h2 id=\"one-highlight-at-a-time\">One highlight at a time</h2>\n<p>Hover over a highlight to get its buttons:</p>\n<ul>\n<li>“Copy” copies it together with its page,</li>\n<li>“Add note”, or “Edit note” if it already has one,</li>\n<li>“Show context” shows the paragraphs around it,</li>\n<li>“Show in book” opens the reader at the highlight, and your saved place doesn’t move until you turn a page,</li>\n<li>Delete.</li>\n</ul>\n<p>Delete one by mistake and the “Highlight deleted.” message has an “Undo” that brings it back.</p>\n<p>Notes are edited in place, either from “Add note” or by clicking an existing note. Return saves, Shift+Return starts a new line, and “Cancel” leaves the note as it was. Saving an empty note removes it.</p>\n<h2 id=\"a-highlight-in-its-context\">A highlight in its context</h2>\n<p>“Show context” swaps the highlight and its note for a single “In context” card: the paragraph before, the highlight’s own paragraph with the passage marked, your note right under it, and the paragraph after. Any footnotes in the passage follow below a hairline. “Hide context” puts the highlight back.</p>\n<p>The context is stored with the highlight, so a Kobo highlight keeps it after the Kobo is unplugged.</p>\n<h2 id=\"searching-highlights-and-notes\">Searching highlights and notes</h2>\n<p>Type in “Search highlights and notes” to find every highlight or note in the library that contains your text. Case doesn’t matter, and it matches inside words, so “heart” also finds “Sweetheart”. The left side narrows to the books with matches (“2 matches”), and each of those shows only its matching highlights.</p>\n<h2 id=\"exporting\">Exporting</h2>\n<p>“Export” saves a book’s highlights and notes as a Markdown or plain-text file through the usual macOS save panel. The file starts with the title and authors, then gives each chapter title and, under it, each highlight with its page and note.</p>\n<h2 id=\"highlights-from-another-edition\">Highlights from another edition</h2>\n<p>Sometimes a Kobo highlight can’t be found in the copy of the book on your Mac. Those highlights come last, under “Not found in this copy”, with a count and a note such as “These highlights came from a Kobo. The copy on this Mac is another edition and does not have their text, so they have no page and no context. They stay in the Notebook and on the device.” They have no “Show context” or “Show in book”.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/back-up-ebook-library/","url":"https://readwyn.com/blog/back-up-ebook-library/","title":"How to back up your ebook library","summary":"Which Readwyn folders to back up, which to leave out, and why to quit the app first. Readwyn does not run backups itself, so your own backup tool does the copying.","content_html":"<p>Everything a Readwyn library holds is in its data folder, so that's the folder to back up. Add your My Words folder and the small file <code>owner-id</code>. Leave out the caches, which Readwyn rebuilds, and the Wiktionary download, which you can fetch again. Quit Readwyn before the backup runs, and turn off Readwyn server if it's on. The app doesn't run, schedule, or check backups itself, so the copying is up to your own backup tool.</p>\n<table><thead><tr><th>What</th><th>Where on a Mac</th><th>Back up</th></tr></thead><tbody>\n<tr><td>Each library</td><td><code>~/Library/Application Support/readwyn/Libraries/</code>; the folder button in Settings › Libraries opens it</td><td>Yes</td></tr>\n<tr><td>My Words</td><td><code>~/Library/Application Support/readwyn/My Words/</code></td><td>Yes</td></tr>\n<tr><td>The owner ID</td><td><code>~/Library/Application Support/readwyn/owner-id</code></td><td>Yes</td></tr>\n<tr><td>Dictionaries</td><td><code>~/Library/Application Support/readwyn/Dictionaries/</code></td><td>No</td></tr>\n<tr><td>Caches</td><td><code>~/Library/Caches/readwyn/</code></td><td>No</td></tr>\n</tbody></table>\n<p>If you installed from the Mac App Store, the <code>~/Library/</code> folders in the table sit inside the app's container: <code>~/Library/Containers/com.readwyn.Readwyn/Data/Library/</code>.</p>\n<h2 id=\"finding-the-data-folder\">Finding the data folder</h2>\n<p>Each library has its own data folder. In Settings › Libraries, click the folder button (“Show in Finder”) on a library's row and Finder opens it. When a data folder is on another drive, the line under the library's name ends with that drive (“1,206 books · 41.0 GB · on “Shelf””), so you know which drive your backup has to include. With <a href=\"/blog/more-than-one-ebook-library/\">more than one library</a>, back up each folder.</p>\n<p>Everything that makes up the library is in two subfolders:</p>\n<ul>\n<li><code>log/</code>: every change as one record, such as a book added, a title set, or a book finished.</li>\n<li><code>blobs/</code>: the book files and the covers, each stored under the hash of its content.</li>\n</ul>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book-details.webp?h=3f5c56cb7a1e9927112f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book-details.webp?h=647f1b629b6b477420d1\" alt=\"The Details tab of Middlemarch: the original MOBI file and the EPUB 3 converted from it, each with Show in Finder, and the KEPUB copy on a Kobo\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The original and the converted EPUB 3 are both kept in the library's data folder</figcaption>\n</figure>\n<h2 id=\"my-words-and-the-owner-id\">My Words and the owner ID</h2>\n<p>Words you save from a lookup belong to the app rather than to one library, so they aren't in any data folder. Back up <code>My Words/</code> in the app folder.</p>\n<p><code>owner-id</code> is tiny, but every record in the log names the owner ID of the server that wrote it. Keep it with the libraries.</p>\n<h2 id=\"what-to-leave-out\">What to leave out</h2>\n<ul>\n<li><strong>Caches.</strong> The cache folder holds the SQLite cache and the cover thumbnails, among other things. Readwyn rebuilds the SQLite cache from the log.</li>\n<li><strong>Dictionaries.</strong> After a restore, download Wiktionary again from Settings › Reading › Look up (about 900 MB).</li>\n</ul>\n<h2 id=\"quit-readwyn-first\">Quit Readwyn first</h2>\n<p>If Readwyn is running while the backup takes its copy, the snapshot can end up with a log record whose book file or cover is missing.</p>\n<p>Quitting isn't enough if Readwyn server is on (Settings › Other devices). The server keeps running after you quit, and a paired Mac can still change the library. Turn it off while the backup runs.</p>\n<h2 id=\"why-incremental-backups-stay-small\">Why incremental backups stay small</h2>\n<p>The library is built so that an incremental backup only picks up what changed:</p>\n<ul>\n<li>Each change is appended to the log. Readwyn starts a new log file at every launch and never touches it again after the app quits.</li>\n<li>A book file or cover never changes once it's stored.</li>\n<li>Metadata isn't written into the book files. Editing a title adds one record to the log.</li>\n</ul>\n<p>The exception is “Compact” in Settings › Libraries. It deletes removed books for good, which means rewriting old log files and deleting the files and covers that only those books used. So the first backup after a compaction copies those rewritten log files again.</p>\n<h2 id=\"your-kobo-database\">Your Kobo database</h2>\n<p>Before a sync writes to a Kobo, Readwyn copies the device's database, <code>KoboReader.sqlite</code>. The message after the sync confirms it: “The device database is backed up.” To find those copies, use “Show backups in Finder” under “Backups” in Settings › E-readers.</p>\n<h2 id=\"on-linux\">On Linux</h2>\n<p>The <a href=\"/blog/ebook-manager-for-linux/\">Linux app</a> keeps its app folder in <code>~/.local/share/readwyn</code> and its caches in <code>~/.cache/readwyn</code>, or in the matching folders under <code>$XDG_DATA_HOME</code> and <code>$XDG_CACHE_HOME</code> if you set them. The files inside are the same as on a Mac.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/browse-ebook-library-from-phone/","url":"https://readwyn.com/blog/browse-ebook-library-from-phone/","title":"Browse your ebook library from your phone on the home network","summary":"Browse your Readwyn library from a phone on your home Wi-Fi with an OPDS reader app: the Listen on setting, the warnings, and what happens away from home.","content_html":"<p>Your phone can browse the Readwyn library on your Mac if it has a reader app for OPDS 1.2 catalogues and is on the same home Wi-Fi. Whether the phone gets through at all comes down to one setting, “Listen on”, and until you change it only apps on the Mac itself can reach the catalogue.</p>\n<h2 id=\"let-the-phone-reach-the-mac\">Let the phone reach the Mac</h2>\n<p>In Settings › Other devices on the Mac, turn on “Readwyn server” and look at “Listen on”. It starts as “This Mac only”. If the catalogue is already on, the Reader apps section warns:</p>\n<p>“Only apps on this Mac can reach the catalogue. To reach it from a phone or an e-reader, choose a network in Listen on.”</p>\n<p>Choose the network your phone is on. Readwyn lists the Mac’s networks the way System Settings › Network names them, each with the Mac’s address on it, for example “Wi-Fi · 192.168.1.20”.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-settings-devices.webp?h=5e2bf654f800618406e3\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-settings-devices.webp?h=1263ffab1becc7426e6c\" alt=\"Settings › Other devices: Readwyn server is on, Running with 2 devices connected; Listen on shows Wi-Fi · 192.168.1.20 and port 7313, with a warning that anything on the Wi-Fi can reach this Mac; below are the paired devices and the reader apps\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>Listen on: the network that the phone uses</figcaption>\n</figure>\n<p>macOS may ask whether Readwyn can accept incoming connections. Allow it, or the phone can’t get in.</p>\n<h2 id=\"turn-on-the-catalogue\">Turn on the catalogue</h2>\n<p>Under “Reader apps”, turn on “Share with reader apps (OPDS)”. The address appears with a Copy button, for example <code>http://192.168.1.20:7314/opds</code>. Click “Add a reader app…”, name it after your phone, and type the address, user name, and password from the next sheet into the reader app. Readwyn’s own test used KOReader, so that’s the app <a href=\"/blog/opds-catalogue-for-koreader/\">the step-by-step post</a> follows.</p>\n<p>The phone can then browse your libraries by author, series, or shelf and download books (<a href=\"/blog/what-is-opds/\">every feed it can open</a>). It can’t change anything, though, and your reading status and progress stay on the Mac. If you want to read on another Mac and keep your progress, pair that Mac instead: <a href=\"/blog/share-ebook-library-between-two-macs/\">Share one ebook library between two Macs</a>.</p>\n<h2 id=\"what-the-warnings-mean\">What the warnings mean</h2>\n<p>Once a network is chosen, two warnings appear. The one under “Listen on” is about other Macs: “Anything on your Wi-Fi can reach this Mac. Only paired devices can open the libraries.” The one under the address is about reader apps:</p>\n<p>“Anything on your Wi-Fi can reach the catalogue, and the reader apps send their password in clear. Share only on a network you trust.”</p>\n<p>The catalogue uses plain HTTP with Basic authentication, not HTTPS, because many reader apps refuse a self-signed certificate. Every request has to carry one reader app’s user name and password, and a request without them is refused with HTTP 401. Since the password crosses the network unencrypted, turn this on at home only, never on a café or hotel network.</p>\n<h2 id=\"away-from-home\">Away from home</h2>\n<p>The catalogue works on your home network only. Readwyn offers no access from outside it and runs no server on the internet for you, so download what you want to read before you leave.</p>\n<p>You don’t need the Readwyn window open, though. With “Readwyn server” on, the server keeps running after you quit Readwyn and starts again at login.</p>\n<h2 id=\"check-that-the-phone-got-through\">Check that the phone got through</h2>\n<p>The list of reader apps shows “Never connected” until the phone’s first request, then “Last connected today”. If it stays at “Never connected”, check that the phone is on the same Wi-Fi and that “Listen on” names that network.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/busiest-reading-hour-and-weekday/","url":"https://readwyn.com/blog/busiest-reading-hour-and-weekday/","title":"Find your busiest reading hour and weekday","summary":"Two charts in Readwyn's Stats show the hour of the day and the weekday you read most, counted in your own time zone, with late nights split at midnight.","content_html":"<p>Further down the Stats screen, two charts give the answer: “By hour of the day” and “By day of the week”. Each caption names the winner, for example “Most at 9 pm” and “Most on Fridays”, and the busiest bar is drawn in your accent colour with the rest paler. Both count your reading time in Readwyn's reader for the range shown at the top of Stats.</p>\n<h2 id=\"how-the-charts-are-built\">How the charts are built</h2>\n<p>“By hour of the day” has 24 bars, from midnight to 11 pm, labelled every six hours with “12 am”, “6 am”, “12 pm”, and “6 pm”. Each bar adds up what you read in that hour across every day of the range.</p>\n<p>“By day of the week” has seven bars. They run from Monday to Sunday, since a week in Readwyn starts on a Monday.</p>\n<p>The arrows at the top of Stats step through the ranges, so you can compare January's evenings with July's. All time takes in every hour you've read in Readwyn's reader, but none of a Kobo's undated time and nothing from <a href=\"/blog/track-reading-time-automatically/\">private reading</a>.</p>\n<p>For a whole year, there's also the “When you read” card of <a href=\"/blog/year-in-books/\">year in review</a>. Its headline names your busiest month, and the line under it gives that month's time and your favourite hour, for example “31 hours. And you read most at 10 pm.”</p>\n<h2 id=\"time-zones-and-late-nights\">Time zones and late nights</h2>\n<p>Hours and weekdays are local, with each day starting at midnight in the time zone of the Mac showing the stats. If you're viewing another Mac's libraries, your Mac sends the name of its time zone, so the other Mac counts your hours, not its own.</p>\n<p>Reading that runs over the hour counts partly in each hour. A session that goes past midnight on a Friday night is split between Friday and Saturday. When the clocks go forward in spring, no time lands in the hour that was skipped, and when an hour repeats in autumn, both parts count in that hour.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/calibre-alternative-for-mac/","url":"https://readwyn.com/blog/calibre-alternative-for-mac/","title":"A calibre alternative for the Mac","summary":"What Readwyn does for a Mac user who comes from calibre: a one-time import, conversion to EPUB 3 with the original kept, its own reader, and Kobo sync.","content_html":"<p>Readwyn is an ebook library and reader for the Mac. If you’re coming from calibre, it imports your calibre library once, and <a href=\"/blog/what-the-calibre-import-keeps/\">the ratings, identifiers and Kobo reading data come with it</a>. After that the books live in a Readwyn library, and you don’t need calibre to read them or to sync them with a Kobo.</p>\n<h2 id=\"bringing-the-library-over\">Bringing the library over</h2>\n<p>You start from “Import from calibre” on the welcome page, or from “Import from calibre…” in the library switcher or Settings › Libraries. A scan of the folder comes first, with the number of books, how much space they take and a rough time for the import. Then the books arrive in the background and you can carry on reading (<a href=\"/blog/import-calibre-library/\">every step, with the screens</a>). In the sheet’s own words: “Readwyn copies your calibre library once. calibre stays exactly as it is.”</p>\n<h2 id=\"the-library\">The library</h2>\n<p>You can see the books as a grid of covers, or as a list with columns for status, progress, rating, last read and date added. The status tabs (“Reading”, “Next up”, “To read”, “Read”, “Did not finish”) count the books in each. The Filter menu narrows them by shelf, author, series and language, and by comics or ebooks when a library has both. Titles in other scripts stay as they are: Kalki’s பொன்னியின் செல்வன் sits in the grid in Tamil.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library.webp?h=b388e029cc1f91357982\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library.webp?h=1ae4e63639f580a708ea\" alt=\"The Readwyn library in the grid view, with covers, the status tabs, and the Kobo card in the sidebar\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The library, with a Kobo connected</figcaption>\n</figure>\n<p>You can keep more than one library. Each has its own data folder, and the library switcher at the top of the sidebar lists them all with their book counts.</p>\n<h2 id=\"older-formats-get-an-epub-3-copy\">Older formats get an EPUB 3 copy</h2>\n<p>Readwyn accepts EPUB, MOBI (with AZW3, combo files and PalmDOC), PDF, OEB 1.x packages, and the comic archives CBZ, CB7 and CBT. A book in an older format gets an EPUB 3 copy when it comes in. The original file stays untouched beside it, and the book’s Details tab lists both. Comics aren’t converted: they open in a page reader of their own.</p>\n<p>A PDF is reflowed into paragraphs, headings and images, without the running headers, footers and page numbers. The conversion was tested on 50 Project Gutenberg PDFs, each checked against Gutenberg’s own EPUB 3 of the same book. Readwyn’s word error rate came to 8.68 %, against 10.37 % for calibre’s <code>ebook-convert --enable-heuristics</code>. (One dictionary is left out of that comparison, because calibre didn’t finish it within five minutes.) Readwyn was also faster on all 50 (<a href=\"/blog/convert-pdf-to-epub-without-calibre/\">the full table</a>). A scan with no text layer won’t convert, and nor will an encrypted PDF.</p>\n<h2 id=\"a-reader-of-its-own\">A reader of its own</h2>\n<p>“Read” opens the EPUB 3 in Readwyn’s own reader: two pages side by side in a wide window, highlights in four colours with notes, dictionary lookup, and find in the book. The Notebook gathers every book’s highlights, chapter by chapter. Because the reader records how long you read, the book screen can show the time left at your own reading speed.</p>\n<h2 id=\"kobo-sync\">Kobo sync</h2>\n<p>Plug a Kobo in over USB. The first time, the Mac App Store version asks you to <a href=\"/blog/kobo-access-mac-app-store/\">give Readwyn access to it</a>, and you select the device in Finder. After that the Sync screen shows two directions, “Device → Readwyn” and “Readwyn → Device”, each with a count of its changes. You review every change before anything is written, and Readwyn copies the device database to a backup before each write. A sync carries your highlights and notes, status, progress, rating and time read, your shelves, and the words you saved in My Words. “Send to device” sends the EPUB 3 as a KEPUB, unless you turn off “Convert EPUB to KEPUB for Kobo e-readers” in Settings › E-readers.</p>\n<h2 id=\"other-computers\">Other computers</h2>\n<p>A second Mac can use the first Mac’s libraries. You pair it in Settings › Other devices, and each paired device gets a level: Guest, Read or Admin. Readwyn also runs on Linux, both <a href=\"/blog/ebook-manager-for-linux/\">as a desktop app</a> and <a href=\"/blog/headless-ebook-server-on-linux/\">as a headless server</a>.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/calibre-published-dates-time-zones/","url":"https://readwyn.com/blog/calibre-published-dates-time-zones/","title":"How Readwyn reads calibre published dates in any time zone","summary":"How the Readwyn calibre import turns each stored published date into the day you entered, and why that day stays the same in every time zone.","content_html":"<p>When Readwyn <a href=\"/blog/import-calibre-library/\">imports a calibre library</a>, each book’s published date comes over as the date you entered in calibre. Readwyn stores it as a calendar day with no time attached, so a Mac in Tokyo and a Mac in New York show the same day.</p>\n<h2 id=\"why-a-date-can-move-by-a-day\">Why a date can move by a day</h2>\n<p>A date kept as a moment in time, at midnight, belongs to one time zone. To the west of that zone, the same moment is still the evening before. Read it in the wrong zone and 29 April turns into 28 April.</p>\n<p>calibre keeps a published date as a moment at midnight, and which midnight depends on where the date came from: a date read from a book file is local midnight on the computer that read it, and a date set with <code>calibredb</code> is midnight UTC.</p>\n<h2 id=\"what-the-import-does\">What the import does</h2>\n<p>The import checks which midnight each date is stored at:</p>\n<table><thead><tr><th>Stored time</th><th>Day that Readwyn keeps</th></tr></thead><tbody>\n<tr><td>Local midnight of this Mac</td><td>The local date</td></tr>\n<tr><td>Midnight UTC</td><td>The UTC date</td></tr>\n<tr><td>Any other time</td><td>The local date</td></tr>\n</tbody></table>\n<p>Say a book’s published date is 29 April 2003, set with <code>calibredb</code>, so it’s stored as midnight UTC. On a Mac in New York that moment is 8 pm on 28 April, but Readwyn sees that it’s midnight UTC and keeps 29 April.</p>\n<h2 id=\"what-the-library-stores\">What the library stores</h2>\n<p>Once imported, the published date has no time at all. It can be a year, a month and a year, or a full date, stored as <code>YYYY</code>, <code>YYYY-MM</code> or <code>YYYY-MM-DD</code>, and nothing about it depends on the clock of the Mac that shows it.</p>\n<p>On the book screen the year comes first in the line under the author: “1871 · ~880 pages · English · EPUB, 1.8 MB”.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book.webp?h=6c9b53f864b64d2872b9\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book.webp?h=1970bca0dae546ce847c\" alt=\"The book screen of Middlemarch: under the author, 1871, about 880 pages, English and EPUB\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The published year, 1871, first in the line under the author</figcaption>\n</figure>\n<p>Reader apps that browse your library over OPDS see that same day as the book’s publication date.</p>\n<h2 id=\"correcting-a-date-by-hand\">Correcting a date by hand</h2>\n<p>If a date was already wrong in calibre, fix it in Readwyn:</p>\n<ol>\n<li>Open the book and choose “Edit details” (or press ⌘E).</li>\n<li>Type the date into the published date field. It accepts a year (“1871”), a month and a year (“Apr 2003”), or a full date (“29 Apr 2003”).</li>\n<li>Choose “Save 1 change”, or press Return. Until you do, the field shows “Was” with the old date struck through, and “Undo”.</li>\n</ol>\n<h2 id=\"dates-of-finished-reads\">Dates of finished reads</h2>\n<p>Finished reads follow a similar rule. calibre, Goodreads and a Kobo record when you finished a book as a date, not a time. Reading stats take a read that finished at exactly 00:00:00 UTC to be one of those dates, and count it on its UTC date wherever you are. Without that rule, a book finished on 1 January would land in the previous year for anyone west of UTC. A read with a real time counts on its local day.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/comic-with-corrupt-page/","url":"https://readwyn.com/blog/comic-with-corrupt-page/","title":"What happens to a comic with a corrupt page","summary":"A comic with a broken page still opens in Readwyn: a placeholder names the bad file, and the other pages stay where they were.","content_html":"<p>Very little. The comic still imports and opens in Readwyn, and every other page shows as usual. Where the broken one should be, you get a box with a broken-image icon, the words “This page can’t be shown”, and the page's file name inside the archive, so you know which file to replace.</p>\n<h2 id=\"what-counts-as-corrupt\">What counts as corrupt</h2>\n<p>Readwyn reads each page's size from the header of its image (JPEG, PNG, GIF, or WebP). A page is corrupt if those bytes give no size, or if its entry in the archive can't be read. It keeps its place in the reading order and still counts as a page. Readwyn never repairs or rewrites the archive, so the broken file is still there for you to replace.</p>\n<p>A page can also fail later, when the reader asks for its image and extracting it goes wrong. That page gets the placeholder too, and Readwyn tries it again if you turn to another spread and come back.</p>\n<h2 id=\"what-the-placeholder-looks-like\">What the placeholder looks like</h2>\n<p>It's a plain box with a thin outline, standing where the page would be, so the other page of a pair stays where it was. A corrupt page has no size of its own, so the box borrows the shape of the nearest page that has one, or 2:3 if none does. The footer counts it as usual (“p. 7 of 24”), and progress moves past it like any other page.</p>\n<h2 id=\"oversized-pages-and-archives\">Oversized pages and archives</h2>\n<p>A small archive can declare entries that unpack to far more data, by accident or on purpose. Readwyn has limits for that:</p>\n<ul>\n<li>A page over 100 megapixels, going by its header, shows the placeholder.</li>\n<li>A page larger than 256 MiB isn't read.</li>\n<li>An archive whose pages unpack to more than 16 GiB in total isn't scanned or copied.</li>\n<li>An entry that unpacks to more than the archive says it holds is malformed, and reading stops one byte past the declared size.</li>\n</ul>\n<h2 id=\"files-that-aren-t-pages\">Files that aren't pages</h2>\n<p>Files that aren't images never turn up as broken pages. <code>ComicInfo.xml</code>, text files, and <code>Thumbs.db</code> are skipped, and so is anything with a part of its path starting with a dot, such as <code>._page1.jpg</code>, or inside the <code>__MACOSX</code> folder.</p>\n<p>If that leaves an archive with no page images at all, it isn't added. The toast after “Add books” names the file and says “it has no page images.” The other books of the same import still arrive.</p>\n<p>A broken <a href=\"/blog/comicinfo-xml/\"><code>ComicInfo.xml</code></a> doesn't stop the import. The comic comes in under its file name, as one without a <code>ComicInfo.xml</code> does.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/comicinfo-xml/","url":"https://readwyn.com/blog/comicinfo-xml/","title":"What ComicInfo.xml does for your comic library","summary":"Each field of ComicInfo.xml that Readwyn reads when you add a comic: the title, the authors, the series, the date, the cover, spreads, deleted pages, and manga.","content_html":"<p><code>ComicInfo.xml</code> is a small metadata file that sits inside a comic archive. When you add a CBZ, CB7, or CBT, Readwyn reads it to fill in the title, authors, series and number, publisher, date, language, summary, genres, cover, and reading direction. A comic without one still comes in, with its file name as the title.</p>\n<h2 id=\"which-file-readwyn-reads\">Which file Readwyn reads</h2>\n<p>Readwyn looks for an entry called <code>ComicInfo.xml</code>, ignoring upper and lower case. If there's more than one, the one at the root of the archive wins, otherwise the one in the shallowest folder. Copies inside <code>__MACOSX</code> or a hidden folder don't count. If the file is malformed, Readwyn ignores it and imports the comic anyway.</p>\n<h2 id=\"the-fields\">The fields</h2>\n<table><thead><tr><th>In <code>ComicInfo.xml</code></th><th>In Readwyn</th></tr></thead><tbody>\n<tr><td><code>Title</code></td><td>The title. If there's no <code>Title</code>, the series and number (“The Lantern Keeper #1”), then the series alone, then the file name.</td></tr>\n<tr><td><code>Writer</code>, then <code>Penciller</code></td><td>The authors, each name once, in that order. If neither is set, <code>CoverArtist</code>.</td></tr>\n<tr><td><code>Series</code>, <code>Number</code></td><td>The series and the number in it. A <code>Number</code> that isn't numeric becomes 1.</td></tr>\n<tr><td><code>Volume</code></td><td>Kept with the book's metadata. The book screen doesn't show it.</td></tr>\n<tr><td><code>Summary</code></td><td>The description.</td></tr>\n<tr><td><code>Publisher</code></td><td>The publisher.</td></tr>\n<tr><td><code>Year</code>, <code>Month</code>, <code>Day</code></td><td>The published date, as exact as the fields allow: a year, a year and a month, or a full date.</td></tr>\n<tr><td><code>LanguageISO</code></td><td>The language.</td></tr>\n<tr><td><code>Genre</code></td><td>The genres.</td></tr>\n<tr><td><code>Tags</code></td><td>The subjects of the file, not the tags of your library.</td></tr>\n<tr><td><code>GTIN</code></td><td>The ISBN if it's a valid ISBN, else a GTIN identifier.</td></tr>\n<tr><td><code>Manga</code></td><td><code>YesAndRightToLeft</code> makes the comic <a href=\"/blog/read-manga-right-to-left-on-mac/\">read from right to left</a>.</td></tr>\n</tbody></table>\n<p><code>Writer</code>, <code>Penciller</code>, and <code>Genre</code> take lists separated by commas, so a <code>Writer</code> of “Mira Okafor, Tomas Lindqvist” gives two authors.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-comic-book.webp?h=3c139e32e63e50aaf717\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-comic-book.webp?h=22ee8ea9dd9bc59a3077\" alt=\"The book screen of the comic The Lantern Keeper: the title, the authors Mira Okafor and Tomas Lindqvist, 2024, 10 pages, English, and CBZ\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A comic’s book screen, with its title, authors, year, and language</figcaption>\n</figure>\n<h2 id=\"the-pages\">The pages</h2>\n<p>The <code>Pages</code> block describes individual pages. Each <code>Page</code> points at an image by its position in the reading order, starting at 0.</p>\n<ul>\n<li><code>Type=\"FrontCover\"</code> marks the cover. The first page marked this way becomes the book's cover; with no mark, the first page does.</li>\n<li><code>Type=\"Deleted\"</code> takes a page out of the reading order. If every page is marked deleted, Readwyn assumes the marks are a mistake and keeps them all.</li>\n<li><code>DoublePage=\"true\"</code> marks a <a href=\"/blog/double-page-spreads-in-comics/\">double-page spread</a>. The page reader shows it alone across the window, as it does any page that's wider than it is high.</li>\n</ul>\n<h2 id=\"the-page-count-and-the-order\">The page count and the order</h2>\n<p>The page count comes from the archive: it's the number of page images, less any that <code>ComicInfo.xml</code> marks as deleted, and the Details tab of the book screen shows it as “24 · from the archive”. The order of the pages comes from their file names, never from the metadata.</p>\n<h2 id=\"comics-from-calibre\">Comics from calibre</h2>\n<p>A calibre import keeps the metadata you set in calibre. From the archive it adds only the page count, if calibre has none, and the reading direction from <code>ComicInfo.xml</code>. <a href=\"/blog/import-calibre-library/\">Importing your calibre library</a> covers the rest of the import.</p>\n<h2 id=\"change-the-details-later\">Change the details later</h2>\n<p>Readwyn takes the details from the file once, when you add the comic. After that they're Readwyn's own, and you change them with “Edit details” on the book screen, as for an ebook.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/comics-and-ebooks-in-one-library/","url":"https://readwyn.com/blog/comics-and-ebooks-in-one-library/","title":"Keep comics and ebooks in one library","summary":"Readwyn keeps comics, manga, and graphic novels in the same library as your ebooks: how to add them, filter by kind, send them to a Kobo, and read them on another Mac.","content_html":"<p>Comics, manga, and graphic novels go into the same Readwyn library as your ebooks, and the Filter menu splits them apart when you want only one kind. A CBZ, CB7, or CBT shows up among your books with its cover, and “Read” opens it in the page reader instead of the ebook reader.</p>\n<h2 id=\"adding-comics-next-to-your-books\">Adding comics next to your books</h2>\n<p>Comics go in the same way ebooks do, calibre import included, and Readwyn tells the formats apart by what's inside each file. An ebook in an older format then gets an EPUB 3 copy, but a comic is never converted: its archive stays the only file it has.</p>\n<p>If an archive has a <a href=\"/blog/comicinfo-xml/\"><code>ComicInfo.xml</code></a>, the title, authors, series, and cover come from it.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-comic-book.webp?h=3c139e32e63e50aaf717\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-comic-book.webp?h=22ee8ea9dd9bc59a3077\" alt=\"The book screen of a comic in the Home library of 412 books: the cover, the title, the authors, 10 pages, CBZ, Continue reading, the progress, the time read, and the time left\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A comic in the same library as the ebooks</figcaption>\n</figure>\n<h2 id=\"the-book-screen\">The book screen</h2>\n<p>A comic gets the same book screen as an ebook, with its status, shelves, rating, review, and reading history. A few things differ:</p>\n<ul>\n<li>The page count is exact, from the archive (“10 pages”, “p. 5 of 10”), where an ebook shows an estimate from its word count, such as “~p. 196 of 880”.</li>\n<li>The Details tab says which way the pages turn, <a href=\"/blog/read-manga-right-to-left-on-mac/\">left to right or right to left</a>, where an ebook's gives its word count.</li>\n<li>There's no Highlights tab.</li>\n</ul>\n<h2 id=\"filter-by-kind\">Filter by kind</h2>\n<p>When a library holds both, the Filter menu gets a “Format” section with “Comics” and “Ebooks”, each with its count. A library with only one kind of book doesn't show the section.</p>\n<h2 id=\"progress-and-stats\">Progress and stats</h2>\n<p>Progress, status, time read, and reading history work exactly as they do for an ebook, and the time you spend on a comic <a href=\"/blog/reading-stats-for-comics/\">counts in Stats</a>. Comics have a reading speed of their own, in pages per minute, so they don't move your words-per-minute figure.</p>\n<h2 id=\"other-devices\">Other devices</h2>\n<p>“Send to device” puts every comic on a Kobo as a CBZ. A CB7 or CBT goes as its <a href=\"/blog/open-cb7-and-cbt-comics/\">reading copy</a>, a new container around the same page images.</p>\n<p>When another Mac opens this Mac's libraries through Readwyn server, its page reader opens the comics too. They aren't available offline there, though, because the offline copy of a book is an EPUB.</p>\n<p>A phone or tablet gets them through a reader app such as <a href=\"/blog/read-comics-in-panels/\">Panels</a> or <a href=\"/blog/read-comics-in-chunky/\">Chunky</a>. With Readwyn server on and “Share with reader apps (OPDS)” switched on in Settings › Other devices › Reader apps, the catalogue offers each comic as its original archive, with the media type of its format, for download over your home network.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/convert-mobi-to-epub-on-mac/","url":"https://readwyn.com/blog/convert-mobi-to-epub-on-mac/","title":"How to convert MOBI to EPUB on a Mac","summary":"Readwyn turns a MOBI into an EPUB 3 as you add it and keeps the original. The EPUB keeps every character, link, image and font of the MOBI.","content_html":"<p>Drop a MOBI file on Readwyn’s window and it comes in as a book with an EPUB 3 made from it. The MOBI itself isn’t changed: it stays next to the EPUB, and the reader opens the EPUB.</p>\n<p>While you drag files over the window, an overlay lists them and says how many books are going into which library; anything Readwyn can’t take is marked “Skipped”. “Add books” does the same from a file dialog. Once they’re in, a message says what happened, for example “Added 2 books to Home library. 1 file skipped.” MOBIs in a calibre library convert the same way during the import (<a href=\"/blog/import-calibre-library/\">How to import your calibre library into Readwyn</a>).</p>\n<p>Conversion covers every version from MOBI 1 to MOBI 6, AZW3 (KF8), and combo files, which carry a MOBI 6 and a KF8 copy of the book in one file (<a href=\"/blog/mobi-kf8-combo-files/\">What is a MOBI and KF8 combo file?</a>).</p>\n<h2 id=\"where-the-epub-ends-up\">Where the EPUB ends up</h2>\n<p>Open the book and go to the Details tab. Under “Files”, “Original” is the MOBI, with its file name and size. “For reading” is the EPUB 3, marked “Converted from the MOBI”, with its size and the date it was made. While the book’s Kobo is connected, a third line, “On the device”, names the e-reader and the day the copy was sent.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book-details.webp?h=3f5c56cb7a1e9927112f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book-details.webp?h=647f1b629b6b477420d1\" alt=\"The Details tab of Middlemarch: the original MOBI file of 2.1 MB, the EPUB 3 converted from the MOBI, and the KEPUB on a Kobo\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The original MOBI and the EPUB 3 made from it</figcaption>\n</figure>\n<p>The first two lines each have “Show in Finder”. If you want the EPUB in another app, copy it from that folder; the library keeps both files either way.</p>\n<h2 id=\"what-the-epub-keeps\">What the EPUB keeps</h2>\n<p>Every converted EPUB has to pass these checks against the MOBI it came from:</p>\n<ul>\n<li>The text is identical, down to the character.</li>\n<li>Every internal link and every entry in the table of contents points to the same place in the text.</li>\n<li>Every image and every font is byte for byte the one in the MOBI.</li>\n<li>The cover is marked as the cover.</li>\n<li>The metadata Readwyn reads from the MOBI is carried into the EPUB.</li>\n<li>There’s an NCX as well, the older form of the table of contents.</li>\n<li><code>epubcheck</code> reports no errors.</li>\n</ul>\n<p>These are checked on real books from Project Gutenberg, MobileRead and similar sites: 10 MOBI 4 files, 10 MOBI 5, 10 standalone MOBI 6, 10 AZW3 and 10 combo files. Ten more EPUBs are turned into MOBI 6, AZW3 and combo files by three other tools and brought back through Readwyn. Those round trips allow for what the writing tool changed, such as images it re-encoded. The converter reads MOBI 1, 2 and 3 too, though the search for test books didn’t turn up a single public-domain book in those versions. Run the same MOBI through twice and you get the same EPUB, byte for byte.</p>\n<h2 id=\"files-that-won-t-convert\">Files that won’t convert</h2>\n<p>A MOBI with DRM doesn’t convert. A damaged file either converts or fails cleanly; it can’t crash the import. If it fails, the book keeps just the MOBI, Readwyn notes why, and a later version of the converter has another go.</p>\n<p>Without an EPUB, the book screen shows “Open” where “Read” would be. It opens the file in its default app, and the “…” menu names that app.</p>\n<h2 id=\"on-a-kobo\">On a Kobo</h2>\n<p>“Send to device” sends the EPUB, not the MOBI. With “Convert EPUB to KEPUB for Kobo e-readers” on (it is by default), it arrives as a KEPUB.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/convert-old-ebook-formats-keep-original/","url":"https://readwyn.com/blog/convert-old-ebook-formats-keep-original/","title":"Convert old ebook formats and keep the original file","summary":"Your MOBI, AZW3, PalmDOC, OEB, EPUB 2 or PDF file stays as it was. The reader and “Send to device” use an EPUB 3 made from it.","content_html":"<p>When you add a MOBI, an AZW3, a PDF or another older format, Readwyn makes an EPUB 3 from it and leaves your file exactly as it was, so you still have it for another app or device. The book then has two files: your original, and the EPUB 3 that the reader and “Send to device” use.</p>\n<h2 id=\"the-formats\">The formats</h2>\n<p>These are converted when you add a book, and when you import a calibre library:</p>\n<ul>\n<li><a href=\"/blog/read-palmdoc-ebooks/\">PalmDOC</a></li>\n<li>MOBI 1 to 6 (see <a href=\"/blog/convert-mobi-to-epub-on-mac/\">MOBI to EPUB on a Mac</a>)</li>\n<li><a href=\"/blog/open-azw3-files-on-mac/\">AZW3 (KF8)</a></li>\n<li><a href=\"/blog/mobi-kf8-combo-files/\">combo files</a>, with MOBI 6 and KF8 in one file</li>\n<li><a href=\"/blog/oeb-package/\">OEB 1.x packages</a></li>\n<li><a href=\"/blog/epub-2-or-epub-3/\">EPUB 2</a></li>\n<li><a href=\"/blog/read-pdf-as-reflowable-text-on-mac/\">PDF</a></li>\n</ul>\n<p>An EPUB 3 needs no conversion. Comic archives (CBZ, CB7 and CBT) aren't converted either: the archive is the comic's only file. Books that were in a library before Readwyn could convert them get their EPUB 3 the next time that library opens.</p>\n<h2 id=\"both-files-in-details\">Both files in Details</h2>\n<p>Open the book and choose the Details tab. Under “Files”, the “Original” line has the format badge, the file name and the size. The “For reading” line has the EPUB 3 badge, the format it was converted from, its size and the date of the conversion.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book-details.webp?h=3f5c56cb7a1e9927112f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book-details.webp?h=647f1b629b6b477420d1\" alt=\"The Details tab of Middlemarch: the original MOBI file, the EPUB 3 converted from the MOBI on 3 Oct 2026, and the KEPUB on a Kobo Clara 2E\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The original MOBI and the EPUB 3 made from it</figcaption>\n</figure>\n<p>Each line has its own “Show in Finder”. While a Kobo with the book is connected, a third line, “On the device”, shows the copy on it.</p>\n<p>Remove the book with “Remove from library…” and both files go with it. The EPUB 3 never becomes a book of its own. Converting the same original again always gives an identical EPUB 3, and if you later add a copy of that EPUB 3, Readwyn treats it as a duplicate of the book it came from.</p>\n<h2 id=\"how-close-the-epub-3-is-to-the-original\">How close the EPUB 3 is to the original</h2>\n<p>For every format except PDF, the EPUB 3 has the same text as the original, the same table of contents and links, and the same images and fonts down to the byte. EPUB 2 books and OEB packages keep every file path as well, so a highlight anchored in one file is valid in the other.</p>\n<p>A PDF only places lines of text on a page, so there's no exact text to hold its EPUB 3 to. It's measured against a reference edition of the same book instead, and <a href=\"/blog/convert-pdf-to-epub-without-calibre/\">Convert PDF to EPUB without calibre</a> gives the results.</p>\n<h2 id=\"files-that-won-t-convert\">Files that won't convert</h2>\n<p>DRM stops a conversion, and so does a <a href=\"/blog/read-pdf-as-reflowable-text-on-mac/\">scanned or encrypted PDF</a>. A corrupt file never crashes the import: either it converts, or it fails and the book keeps just its original. Readwyn records which version of the converter gave up, so a newer converter in a later version of the app tries the file again.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/convert-pdf-to-epub-without-calibre/","url":"https://readwyn.com/blog/convert-pdf-to-epub-without-calibre/","title":"Convert PDF to EPUB without calibre","summary":"Readwyn converts a PDF with a text layer to EPUB 3 with its own converter, no calibre needed. On 50 test PDFs its word error rate matched or beat ebook-convert’s on 47.","content_html":"<p>On 50 Project Gutenberg PDFs, Readwyn’s own converter matched or beat calibre’s <code>ebook-convert --enable-heuristics</code> on word error rate for 47 of them, and it was faster on all 50. It’s part of the app, so you don’t need calibre, poppler or anything else installed. Add a PDF and you get an EPUB 3 of it in the library, with the PDF kept beside it.</p>\n<h2 id=\"adding-pdfs\">Adding PDFs</h2>\n<p>Drop the PDFs on Readwyn’s window, or pick them with “Add books”. PDFs that already sit in a calibre library are converted during the import (<a href=\"/blog/import-calibre-library/\">How to import your calibre library into Readwyn</a>).</p>\n<p>“Read” then opens the EPUB in Readwyn’s reader; <a href=\"/blog/read-pdf-as-reflowable-text-on-mac/\">Read a PDF book as reflowable text on a Mac</a> shows what that looks like. If you’d rather take the EPUB to another app, the book’s Details tab has “Show in Finder” for it.</p>\n<h2 id=\"what-the-conversion-does\">What the conversion does</h2>\n<p>A PDF puts each line at a fixed spot on a page. To get a book that reflows, the converter:</p>\n<ul>\n<li>joins lines into paragraphs across pages and columns, reading a page with columns one column at a time;</li>\n<li>drops running headers, running footers and page numbers;</li>\n<li>rejoins words that were hyphenated at the end of a line;</li>\n<li>keeps italic, bold, superscripts and subscripts;</li>\n<li>finds headings from their size, their weight and the PDF’s outline;</li>\n<li>builds the table of contents from that outline, or from the headings if there’s no outline;</li>\n<li>turns link annotations into links;</li>\n<li>keeps every JPEG byte for byte and every other image with the same pixels;</li>\n<li>marks an image that covers at least 80 % of page 1 as the cover.</li>\n</ul>\n<h2 id=\"what-it-can-t-convert\">What it can’t convert</h2>\n<p>A scanned PDF with no text layer won’t convert, because there’s no OCR. Nor will an encrypted PDF. Either way you keep the PDF alone, and “Open” passes it to your default PDF app.</p>\n<p>JPEG 2000 and JBIG2 images (<code>JPXDecode</code> and <code>JBIG2Decode</code>) are left out, but the rest of the book still converts.</p>\n<h2 id=\"how-the-conversion-was-measured\">How the conversion was measured</h2>\n<p>A PDF has no exact text to compare with, so each conversion is scored against a reference: the EPUB 3 of the same Project Gutenberg ebook. The 50 PDFs are in English, German, French, Italian, Greek, Spanish, Latin and other languages, and run from 20 to 2,164 pages.</p>\n<p>The baseline is calibre 9.15.0’s <code>ebook-convert --enable-heuristics</code>, run once with <code>--pdf-engine calibre</code> and once with <code>--pdf-engine pdftohtml</code>. For each book, whichever run had the lower word error rate counts. Each run has a time limit of 5 minutes.</p>\n<p>Both outputs are scored against the reference in five ways:</p>\n<ul>\n<li><strong>Word error rate:</strong> wrong, missing and extra words, divided by the number of words in the reference.</li>\n<li><strong>Paragraph-break F1:</strong> how closely the paragraph breaks match the reference’s, from 0 to 1.</li>\n<li><strong>Heading recall:</strong> the share of the reference’s headings that come out as headings.</li>\n<li><strong>Table-of-contents coverage:</strong> the share of the PDF outline’s entries that end up in the table of contents, pointing to the right place.</li>\n<li><strong>Image coverage:</strong> the share of the images drawn on the pages that make it into the output.</li>\n</ul>\n<p>Each mean leaves out the books where its measure doesn’t apply: a reference with no headings, a PDF with no outline, or pages with no images.</p>\n<p>One book skews the total: pg31552, a Portuguese dictionary of 2,164 pages. Both calibre runs went over the limit on it, so its baseline is an EPUB with no text. The second line of the table leaves it out.</p>\n<table><thead><tr><th>Measure</th><th>Readwyn</th><th>Baseline</th></tr></thead><tbody>\n<tr><td>Total word error rate</td><td>9.15 %</td><td>36.81 %</td></tr>\n<tr><td>Total word error rate without pg31552</td><td>8.68 %</td><td>10.37 %</td></tr>\n<tr><td>Mean paragraph-break F1 (50 books)</td><td>0.753</td><td>0.587</td></tr>\n<tr><td>Mean heading recall (42 books)</td><td>0.912</td><td>0.145</td></tr>\n<tr><td>Mean table-of-contents coverage (28 books)</td><td>0.990</td><td>0.727</td></tr>\n<tr><td>Mean image coverage (25 books)</td><td>1.000</td><td>0.859</td></tr>\n</tbody></table>\n<p>On the three books where calibre did better, Readwyn’s word error rate was higher by 0.14 percentage points at most. <code>epubcheck</code> reports no errors for any of the 50 EPUB 3 files.</p>\n<p>Both were timed on the same Mac, one conversion at a time, with Readwyn’s release build. Readwyn’s slowest conversion was that dictionary, at 0.45 s. Mother’s Remedies, 1,150 pages, took 0.06 s against 115.6 s for the baseline.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/count-words-in-ebook/","url":"https://readwyn.com/blog/count-words-in-ebook/","title":"How to count the words in an ebook","summary":"Open a book's Details tab to see its word count, made on your Mac when you added the book. The count drives page estimates, time left and stats.","content_html":"<p>Open the book, choose the Details tab, and look at the “Words” row under “About this edition”. It gives the count and when it was made, for example “counted 3 Oct 2026”. Click “Count again” next to it to count the book once more. A book with no count shows “–”.</p>\n<p>You don't have to start the count yourself. Readwyn counts the words when you add a book, on your Mac, with no web service. Books imported from calibre aren't counted on the way in. If your calibre library has a custom column with the lookup name #word_count, its value comes along; if not, the row shows “–” until you click “Count again”.</p>\n<p>The text is split at spaces. Chinese and Japanese don't put spaces between words, so there each ideograph counts as one word. For a MOBI, a PDF or another older format, Readwyn counts the <a href=\"/blog/convert-old-ebook-formats-keep-original/\">EPUB 3 it converted the book to</a>. A file with no text gets no count. Comics have pages rather than words, and in a comic's Details tab a “Reading” row takes the place of “Words”, giving the reading direction, such as “Left to right”.</p>\n<p>Divided by 250 and rounded up, the count becomes the <a href=\"/blog/ebook-page-count-from-word-count/\">page estimate</a> on the book screen. The words you haven't read yet, at your own reading speed, give the <a href=\"/blog/how-long-to-read-a-book/\">time left in the book</a>. Stats adds up the word counts of the books you've finished and shows the total under “Pages read”.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-stats.webp?h=6be17290a8c05ed4461f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-stats.webp?h=4b431a86ee04dd47470d\" alt=\"Reading stats for October 2026: 21 h 46 min read, 4 books finished, 1,286 pages read with 391,000 words in 3 of 4 books, and a reading speed of 284 words per minute\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Pages read, with the words of the finished books</figcaption>\n</figure>\n<p>A finished book with no word count can't add to that total, so the line says how many books it covers: “391,000 words in 3 of 4 books”. When every finished book has a count, you just get the words, for example “93,000 words”. To close the gap, open the uncounted book's Details tab and click “Count again”. A recount also puts the page estimate back over any page count you typed.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/dark-mode-for-ebooks-on-mac/","url":"https://readwyn.com/blog/dark-mode-for-ebooks-on-mac/","title":"Dark mode for reading ebooks on a Mac","summary":"Read ebooks in dark mode in Readwyn: set the app to Dark or Match the Mac, or give only the reader a dark page while the rest of the app stays light.","content_html":"<p>Readwyn has two switches for dark mode. Settings › Appearance sets the whole app to Light, Dark, or “Match the Mac”. The Text panel in the reader sets the page on its own: App, Light, or Dark. So you can read on a dark page while the library stays light, or the other way round.</p>\n<h2 id=\"the-whole-app\">The whole app</h2>\n<p>Open Settings (⌘, or Settings in the sidebar) and choose Appearance. The first row, “Theme”, offers Light, Dark, and “Match the Mac”, which follows macOS and switches with it while Readwyn runs. The change is instant, and Readwyn remembers it. Under the theme you can also pick one of five accent colours.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-settings.webp?h=a1d1630c4977afb8a021\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-settings.webp?h=ffd7c6ebffe9fd175e49\" alt=\"Settings › Appearance: the theme Light, Dark, or Match the Mac, the five accent colours with Cobalt chosen, and the library options\" width=\"780\" height=\"581\" loading=\"lazy\">\n</picture>\n<figcaption>Settings › Appearance: the theme and the accent colour</figcaption>\n</figure>\n<p>On Linux, the third choice is called “Match the system”.</p>\n<h2 id=\"just-the-reader\">Just the reader</h2>\n<p>In the reader, click “Text” in the top bar. The last row of the panel is “Theme”, with App, Light, and Dark. App follows the app, so with “Match the Mac” the page darkens when macOS does. Light and Dark hold the reader in that scheme whatever the app is set to, and they cover everything in the reader window: the background, the page, the top and bottom bars, and the popovers. Your accent colour stays the same.</p>\n<p>A note at the foot of the panel reads “App follows Settings › Appearance. These apply to every book on this Mac.” The same panel also sets the font, size, line spacing and margins (<a href=\"/blog/read-epub-on-mac/\">more on those</a>).</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader.webp?h=8f10e162ef8d8cef7c82\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader.webp?h=9141239fb26850877292\" alt=\"A two-page spread of Middlemarch in the reader, with two highlights and the popover of a note\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Put your system in dark mode and this capture turns dark too</figcaption>\n</figure>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/dictionary-order/","url":"https://readwyn.com/blog/dictionary-order/","title":"Choose the order of your dictionaries","summary":"Set which dictionary Readwyn asks first when you look up a word: the order in Settings, how a lookup picks its entry, and how to see another dictionary once.","content_html":"<p>Your dictionaries are listed in Settings › Reading › Look up, under the note “First one that has the word”. Each time you <a href=\"/blog/look-up-words-in-ebooks-on-mac/\">hold a word in the reader</a>, Readwyn asks them in that order and opens the popover on the first entry it finds. Drag a row up or down to change the order.</p>\n<p>A Mac has two rows: the macOS Dictionary, which is always “Ready”, and Wiktionary, which Readwyn <a href=\"/blog/offline-dictionary-wiktionary/\">downloads for offline use</a>. Before its download, Wiktionary still sits in the list, but lookups skip it until it’s ready. On Linux there’s only Wiktionary.</p>\n<p>The “macOS Dictionary” row stands for every dictionary switched on in the Dictionary app’s settings, and Readwyn asks them together, at that row’s place in your order. To change which of those Readwyn uses, switch them on or off there. If none of your ready dictionaries has the word, the popover says “No entry in the dictionaries on this Mac.”</p>\n<p>The order is kept per Mac. If you pair a second Mac with your library, it keeps its own order and looks words up in its own dictionaries.</p>\n<h2 id=\"another-dictionary-for-one-word\">Another dictionary for one word</h2>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-lookup.webp?h=100a081b3204e86878eb\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-lookup.webp?h=048cd0b33c717f3ab5f2\" alt=\"The lookup popover over a page of Middlemarch, with the entry for forlorn and the name macOS Dictionary at the bottom\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The bottom of the popover names the dictionary the entry came from</figcaption>\n</figure>\n<p>With more than one dictionary ready, the dictionary’s name at the bottom of the popover opens a menu. It lists the dictionary you’re looking at first, then the rest in your order. Picking one shows its entry for this lookup only, and the next word goes back to your order.</p>\n<h2 id=\"my-words-follows-the-same-order\">My Words follows the same order</h2>\n<p>Words you save to <a href=\"/blog/save-looked-up-words/\">My Words</a> are stored without their definitions. Each time the My Words view shows a word, it looks it up afresh, in the same order. Change the order, or let another dictionary finish downloading, and the view looks its words up again, so a word that had no entry before gets one as soon as a dictionary that knows it is ready.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/double-page-spreads-in-comics/","url":"https://readwyn.com/blog/double-page-spreads-in-comics/","title":"How Readwyn shows double-page spreads in digital comics","summary":"How Readwyn finds a double-page spread in a comic archive and shows it alone across the window, and how it pairs the other pages around it.","content_html":"<p>Readwyn's page reader gives a double-page spread the window to itself, across the full width, as one picture. A page counts as a spread if its image is wider than it is high, or if the archive's <code>ComicInfo.xml</code> marks it with <code>DoublePage=\"true\"</code>, even when the image is taller than it is wide. The pages before it pair up so that a spread never has to share the window with another page.</p>\n<h2 id=\"where-the-page-size-comes-from\">Where the page size comes from</h2>\n<p>Readwyn gets each page's width and height from the header of its image file. JPEG, PNG, GIF, and WebP all keep them there, so it never has to decode the picture, and it reads no more than the first 256 KB of a page to find out.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-comic-spread.webp?h=1ebbec3459ebd7ae3959\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-comic-spread.webp?h=24835cf8d48886d7498f\" alt=\"The page reader with one wide page across the window, Fit page and Fit width at the top right, and p. 6 of 10, 50%, and 3 min left in the footer\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A double-page spread, alone across the reader</figcaption>\n</figure>\n<h2 id=\"how-the-pages-pair-up\">How the pages pair up</h2>\n<p>In a window at least 1000 px wide you get two pages side by side, paired like this:</p>\n<ul>\n<li>The cover, the first page, shows alone.</li>\n<li>A double-page spread shows alone, across the full width.</li>\n<li>Every other page shares the window with its neighbour.</li>\n<li>A page left without a partner, just before a spread or at the end of the comic, shows alone.</li>\n</ul>\n<p>The two pages of a pair are scaled to one height and meet at the spine. A soft shade on either side of the middle and one shadow under both make them look like an open book. A page whose size can't be read, such as a <a href=\"/blog/comic-with-corrupt-page/\">corrupt page</a>, takes its proportions from a page nearby. In a <a href=\"/blog/read-manga-right-to-left-on-mac/\">manga</a>, the earlier page of a pair sits on the right.</p>\n<h2 id=\"one-page-at-a-time\">One page at a time</h2>\n<p>“Always one page” in Settings › Reading turns the pairing off; the default is “Two pages when the window is wide”. Below 1000 px you get one page whatever the setting.</p>\n<p>“Fit width” in the top bar also shows a single page, as wide as the window. A tall or zoomed page scrolls with the trackpad, a drag, or the up and down arrow keys.</p>\n<h2 id=\"zoom\">Zoom</h2>\n<p>Fitted to the window, a spread can leave the panels small. Each press of ⌘= zooms in by a factor of 1.25, up to 400%, and a pinch on the trackpad works too, with the middle of the view staying put. Thin scroll bars appear on the right, and along the bottom when the page is wider than the window. ⌘0, or the zoom button in the top bar (“150%”, say), takes you back to the fit.</p>\n<h2 id=\"page-numbers\">Page numbers</h2>\n<p>A spread is a single image in the archive, so it counts as one page. The footer shows a pair as “pp. 12–13 of 24” and a spread on its own as one page, such as “p. 6 of 10”.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/download-books-to-ereader-over-wifi/","url":"https://readwyn.com/blog/download-books-to-ereader-over-wifi/","title":"Download books to an e-reader over Wi-Fi from your own library","summary":"Download books from your Readwyn library to an e-reader over Wi-Fi with an OPDS app such as KOReader, with one sign-in per device and a password that’s easy to type.","content_html":"<p>Readwyn gives each e-reader a password of its own, something like <code>k7mq-2xfp-9rtw-hb3n</code>, plain enough to type on the e-reader itself. With that password, a user name, and an address, an OPDS reader app such as KOReader downloads books from the library on your Mac over your home Wi-Fi.</p>\n<p>On the Mac, it takes three settings in Settings › Other devices: “Readwyn server” on, the e-reader’s network in “Listen on”, and “Share with reader apps (OPDS)” on. <a href=\"/blog/browse-ebook-library-from-phone/\">Browse your ebook library from your phone on the home network</a> explains “Listen on” and the warnings that come with it.</p>\n<h2 id=\"one-sign-in-for-each-e-reader\">One sign-in for each e-reader</h2>\n<p>Click “Add a reader app…” and type a name for this e-reader and its app. The sheet’s hint is “The name helps you tell the reader apps apart.” Readwyn builds the user name from it: the ASCII letters and digits in lower case, words joined by hyphens, 24 characters at most. “Kindle Scribe KOReader” becomes <code>kindle-scribe-koreader</code>. If another reader app already has that user name, the new one gets <code>-2</code> on the end, then <code>-3</code>.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-apps.webp?h=e0024127c108d4d70c9a\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-apps.webp?h=b3168403e87d03df89f1\" alt=\"Settings › Other devices › Reader apps: the address http://192.168.1.20:7314/opds with a warning about the Wi-Fi, and three reader apps: Kindle Scribe KOReader, last connected today, Sid’s iPad · Panels, last connected yesterday, and Boox Note Air, never connected\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>One reader app for each e-reader</figcaption>\n</figure>\n<h2 id=\"a-password-you-can-type-on-an-e-reader\">A password you can type on an e-reader</h2>\n<p>The next sheet takes the name you typed as its title, with “In the reader app, add an OPDS catalogue and type:” in smaller text below. It shows the address, user name, and password, each with a Copy button, though Copy won’t help much if you’re typing on the e-reader.</p>\n<p>That’s why the password has 16 characters in four groups of four, all lowercase letters and digits, with no i, l, o, 0, or 1, so there’s little to mistype. It’s random, with about 79 bits of entropy.</p>\n<p>You see it once: “Shows only now. If you lose it, remove the reader app and add it again.” Readwyn keeps only a salted hash, never the password itself.</p>\n<h2 id=\"download-a-book\">Download a book</h2>\n<p>On the e-reader you browse your libraries by author, series, or shelf, or search them. Each book offers one download per format, the EPUB first (<a href=\"/blog/download-books-to-moon-reader-from-mac/\">which one to take</a>). In Readwyn’s test with KOReader, the EPUB 3 of “Crime and Punishment” downloaded and opened (<a href=\"/blog/opds-catalogue-for-koreader/\">the full test</a>).</p>\n<p>To see the books you’re reading on the Mac, tick “Currently reading” next to the e-reader in the list. The catalogue then lists them, without your progress.</p>\n<h2 id=\"if-you-lose-an-e-reader\">If you lose an e-reader</h2>\n<p>Click “Remove…” next to it in the list of reader apps. The same goes for one you give away. Readwyn asks first, then shuts it out at once: its next request is refused with HTTP 401, even on a connection that’s already open. Your other e-readers keep their passwords.</p>\n<h2 id=\"a-kobo-over-usb\">A Kobo over USB</h2>\n<p>A Kobo can also go by cable. Plugged into the Mac, it takes books over USB with “Send to device” (<a href=\"/blog/send-books-to-kobo-from-mac/\">how to send books to a Kobo</a>). A sync also brings your progress, highlights, and time read back to the Mac, which the read-only catalogue can’t do.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/download-books-to-moon-reader-from-mac/","url":"https://readwyn.com/blog/download-books-to-moon-reader-from-mac/","title":"Download books to Moon+ Reader from your Mac","summary":"Download books from the Readwyn library on your Mac to Moon+ Reader over Wi-Fi, and pick between the EPUB 3 copy and the original file.","content_html":"<p>In Readwyn’s OPDS catalogue, a book that Readwyn converted gives Moon+ Reader a choice of two files: the EPUB 3 copy that Readwyn made, and the original you added, such as a MOBI, AZW3, PDF, or EPUB 2. Moon+ Reader downloads whichever you pick straight from the Mac over your home Wi-Fi.</p>\n<h2 id=\"what-the-mac-has-to-do\">What the Mac has to do</h2>\n<p>Moon+ Reader downloads from the Mac itself, so the Mac has to be on, with you logged in. Readwyn doesn’t have to be open. The line under “Readwyn server” in Settings › Other devices says so: “Other devices can open the libraries of this Mac. It keeps running when you quit Readwyn, and starts at login.”</p>\n<p>With that setting on, pick your Wi-Fi in “Listen on”, turn on the catalogue under “Reader apps”, and add a reader app called “Moon+ Reader”. The sheet that follows gives the address, user name, and password to type into the app. Moon+ Reader itself hasn’t been tried against the catalogue yet. KOReader has, and <a href=\"/blog/opds-catalogue-for-koreader/\">its post</a> gives the steps and what happened.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-settings-devices.webp?h=5e2bf654f800618406e3\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-settings-devices.webp?h=1263ffab1becc7426e6c\" alt=\"Settings › Other devices: Readwyn server is on and running with 2 devices connected, Listen on shows Wi-Fi · 192.168.1.20 with port 7313 and a warning, the paired devices, and the reader apps with the address http://192.168.1.20:7314/opds\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>Readwyn server on, listening on the Wi-Fi</figcaption>\n</figure>\n<h2 id=\"which-file-to-download\">Which file to download</h2>\n<p>Each book lists one download per format, with its file type and size, always in this order:</p>\n<ol>\n<li>The EPUB that Readwyn’s own reader opens: the converted EPUB 3, labelled “EPUB 3”, or the original EPUB if there’s no converted copy.</li>\n<li>The original file the EPUB 3 was made from, such as “MOBI (original)” or “EPUB (original)”.</li>\n</ol>\n<p>Take “EPUB 3” if you want what Readwyn itself reads, or the original if you want the file exactly as it came; Readwyn never changes it. A book that Readwyn couldn’t convert, such as a scanned PDF, has only its original, and a comic has only its archive (CBZ, CB7, or CBT). The <a href=\"/#formats\">formats section</a> on the home page covers the conversion.</p>\n<p>Nothing you do in Moon+ Reader reaches back to the Mac. The catalogue is read only, so what you read there doesn’t move your place in Readwyn.</p>\n<h2 id=\"file-names\">File names</h2>\n<p>Readwyn suggests a name made of the title and the first author, for example <code>Klara and the Sun - Kazuo Ishiguro.epub</code>. It sends that name twice, once in plain ASCII and once in UTF-8, so an app that reads the UTF-8 version keeps a title in Tamil or any other script intact.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ebook-library-in-thorium-reader/","url":"https://readwyn.com/blog/ebook-library-in-thorium-reader/","title":"Browse your ebook library in Thorium Reader","summary":"Add your Readwyn library to Thorium Reader as an OPDS catalogue, then browse it by author, series, or shelf and download books.","content_html":"<p>Thorium can open your Readwyn library as an OPDS catalogue. Give it the address, user name, and password that Readwyn makes for it, and you can browse your books by author, series, or shelf and download the ones you want to read in Thorium.</p>\n<p>Thorium is one of the three apps that Readwyn names in its own settings. Under “Reader apps” in Settings › Other devices, the line below the switch reads “Reader apps such as KOReader, Thorium or Moon+ Reader can browse the libraries of this Mac and download books. They change nothing.” With “Readwyn server” on, switch the catalogue on, add a reader app called “Thorium” so its row is easy to spot later, and copy what the next sheet shows into Thorium. The <a href=\"/blog/opds-catalogue-for-koreader/\">KOReader post</a> goes through every screen.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-apps.webp?h=e0024127c108d4d70c9a\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-apps.webp?h=b3168403e87d03df89f1\" alt=\"Settings › Other devices › Reader apps: the switch Share with reader apps (OPDS) is on, the address http://192.168.1.20:7314/opds with Copy, and the list of reader apps with Add a reader app\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>The address that Thorium needs</figcaption>\n</figure>\n<p>If Thorium runs on this same Mac, leave “Listen on” at “This Mac only”. The address is then <code>http://127.0.0.1:7314/opds</code>, and the note under it says “Only apps on this Mac can reach the catalogue.” Nothing goes over the Wi-Fi, the password included. Thorium on another computer needs your network picked in “Listen on” first.</p>\n<p>Only KOReader has been tried against the catalogue so far, but every feed passes the Atom and OPDS 1.2 checks in Readwyn’s tests.</p>\n<h2 id=\"what-you-ll-find\">What you’ll find</h2>\n<p>If you keep one library, Thorium opens straight into it. Otherwise the first list is your libraries, each with its book count. Inside a library there’s Recently added (newest first), By author, By series, By shelf, and All books. An author or a shelf lists its books by title, and a series lists them in series order.</p>\n<p>Your Readwyn shelves come across as they are, so a shelf you made for your book club becomes a list of its own. The catalogue’s search matches titles, authors, and series, exactly as the search on Readwyn’s library screen does, so a search from Thorium should turn up the same books. Long lists arrive 50 books at a time.</p>\n<h2 id=\"what-a-book-entry-carries\">What a book entry carries</h2>\n<p>Each entry has the cover, title, authors, languages, publisher, publication date, ISBN, and genres. OPDS 1.2 has no field for a series, so Readwyn puts it at the top of the book’s text, for example “The Expanse, book 2”, with the description below. The description keeps its paragraphs, line breaks, bold, and italics; any other formatting is stripped.</p>\n<p>Under all that are the downloads, each with its file type and size. A converted book offers its EPUB 3 and the original (<a href=\"/blog/download-books-to-moon-reader-from-mac/\">which one to take</a>). A comic offers only its CBZ, CB7, or CBT archive.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ebook-manager-arm64-linux/","url":"https://readwyn.com/blog/ebook-manager-arm64-linux/","title":"An ebook manager on arm64 Linux","summary":"Readwyn's Linux app (deb and rpm) and its headless server (deb, rpm and tar) have arm64 builds with the same features as x86_64; the app was tested on Ubuntu, Debian and Fedora.","content_html":"<p>Every Readwyn package for Linux also comes in an arm64 build, and nothing is left out of it: the 3D library, Kobo sync and conversion to EPUB 3 work just as they do on x86_64. The headless server has arm64 builds too, so a home server or NAS with an arm64 processor can hold your libraries.</p>\n<h2 id=\"the-arm64-files\">The arm64 files</h2>\n<table><thead><tr><th>Package</th><th>arm64 file</th></tr></thead><tbody>\n<tr><td>The app, Ubuntu and Debian</td><td><code>readwyn_&lt;version&gt;-1_arm64.deb</code></td></tr>\n<tr><td>The app, Fedora</td><td><code>readwyn-&lt;version&gt;-1.aarch64.rpm</code></td></tr>\n<tr><td>The server, Ubuntu and Debian</td><td><code>readwyn-server_&lt;version&gt;-1_arm64.deb</code></td></tr>\n<tr><td>The server, Fedora</td><td><code>readwyn-server-&lt;version&gt;-1.aarch64.rpm</code></td></tr>\n<tr><td>The server, other systems</td><td><code>readwyn-server-&lt;version&gt;-aarch64.tar.gz</code></td></tr>\n</tbody></table>\n<p>The deb files install on Ubuntu 22.04 LTS and later and on Debian 12 and later, the rpm files on any Fedora release that is still supported. Underneath, they need glibc 2.35 (the C library of Ubuntu 22.04 LTS) or newer. You install them the same way as the x86_64 packages on <a href=\"/blog/install-readwyn-on-ubuntu/\">Ubuntu</a> or <a href=\"/blog/install-readwyn-on-fedora/\">Fedora</a>; only the file name changes.</p>\n<p>Those two formats install with no extra setup on about 76% of Linux desktops, and on about 93% of arm64 ones (figures from October 2026). There are no Flatpak, Snap or AppImage builds.</p>\n<h2 id=\"don-t-take-the-x86-64-file-by-mistake\">Don't take the x86_64 file by mistake</h2>\n<p>If your arm64 Ubuntu also has apt's <code>amd64</code> architecture enabled, App Center will happily install the x86_64 package, and Readwyn then runs under x86_64 emulation. Take the arm64 file instead. Fedora doesn't allow the mix at all: it won't install a package for a second architecture beside its own.</p>\n<h2 id=\"where-it-was-tested\">Where it was tested</h2>\n<p>Readwyn's install tests run on arm64 virtual machines. On 24 October 2026 the arm64 app package passed them on:</p>\n<ul>\n<li>Ubuntu 24.04 with GNOME, on X11 and Wayland</li>\n<li>Ubuntu 22.04 with GNOME, on Wayland</li>\n<li>Debian 12 with GNOME, on Wayland</li>\n<li>Fedora 42 with KDE Plasma 6.6, on Wayland</li>\n</ul>\n<p>Each time it installed (with a double-click, or with <code>apt</code> on Debian), appeared in the app menu with its icon, started from there, showed the library and opened a book. Removing it didn't touch the libraries.</p>\n<p>The x86_64 packages went through the same test on the Ubuntu and Debian machines, under emulation. On Fedora the x86_64 rpm got a separate test with <code>dnf</code> alone.</p>\n<h2 id=\"the-server\">The server</h2>\n<p><code>ryn-server</code> is the version of Readwyn for a machine without a display. Its arm64 deb was tested alongside the x86_64 one on 22 October 2026, on Ubuntu 24.04 server. Setup is in <a href=\"/blog/headless-ebook-server-on-linux/\">Run a headless ebook server on Linux</a>.</p>\n<h2 id=\"the-3d-library-without-a-gpu\">The 3D library without a GPU</h2>\n<p>The 3D library draws with Vulkan, so it needs a Vulkan driver, and the package recommends Mesa's. On a machine with no GPU, Mesa draws it on the processor instead. It works, slowly.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ebook-manager-for-linux/","url":"https://readwyn.com/blog/ebook-manager-for-linux/","title":"An ebook manager for Linux","summary":"Readwyn runs on Linux with the library, reader, Kobo sync and 3D library of the Mac app, from deb and rpm packages for x86_64 and arm64.","content_html":"<p>Readwyn is an ebook library and reader, and its Linux version has every feature of the Mac one: EPUB books and comics (CBZ, CB7 or CBT) in as many libraries as you like, conversion of older formats to EPUB 3, Kobo sync, a one-off calibre import, reading stats and the 3D library. It comes as a deb package for Ubuntu and Debian and as an rpm for Fedora, each built for x86_64 and arm64.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/linux/dark-library.webp?h=02876295485e7080879c\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/linux/light-library.webp?h=24919ab2ffd9d07527ff\" alt=\"The Readwyn library on Linux: the sidebar with Home library, Library, Notebook, Stats, Sync, and the Kobo Clara 2E card; a grid of covers with titles and authors; the search with Ctrl+K and the window controls at the top right\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The library on Linux</figcaption>\n</figure>\n<h2 id=\"where-it-finds-and-keeps-things\">Where it finds and keeps things</h2>\n<p>You can <a href=\"/blog/import-calibre-library/\">import a calibre library</a> once, and on Linux Readwyn finds calibre's default library through <code>~/.config/calibre/global.py.json</code>. A Kobo turns up wherever udisks mounts it: <code>/media/&lt;user&gt;</code> on Ubuntu and Debian, <code>/run/media/&lt;user&gt;</code> on Fedora.</p>\n<p>Your Goodreads sign-in token goes into your desktop's keyring (gnome-keyring or KWallet, through the Secret Service). Libraries and settings live in <code>~/.local/share/readwyn</code>, caches in <code>~/.cache/readwyn</code>.</p>\n<h2 id=\"what-changes-from-the-mac\">What changes from the Mac</h2>\n<ul>\n<li>Wherever the Mac app says “Mac” for a computer, the Linux app says “computer”: “this computer”, “No computers connected”. It can't tell whether another machine is a Mac.</li>\n<li>Shortcuts swap ⌘ for Ctrl, so Ctrl+, opens Settings and Ctrl+Shift+L enters the 3D library. Super stays with the desktop. Back is Alt+←, as in Linux browsers and file managers, and the Mac's Emacs keys in text fields are gone: Ctrl+A selects all.</li>\n<li>Settings › Appearance offers “Match the system” instead of “Match the Mac”. There's no Readwyn menu at the top of the screen as on a Mac, so “About Readwyn” sits at the foot of that tab.</li>\n<li>Linux has no dictionary of its own, so Wiktionary is the only one. Until you download it, a lookup offers it there and then: English entries, about 900 MB, once. After that, lookups work offline.</li>\n<li>The speak button in a lookup or in My Words goes through speech-dispatcher, the system's speech service.</li>\n<li>GNOME on Wayland leaves the title bar to the app, so Readwyn draws its own minimize, maximize and close at the right end of the top bar. There's more on the windows in the post on <a href=\"/blog/read-ebooks-on-linux-wayland-x11/\">Wayland and X11</a>.</li>\n<li>With Readwyn server on in Settings › Other devices, the background server starts at login from an entry in <code>~/.config/autostart</code>. Its icon sits in the system tray on KDE, and on GNOME when the AppIndicator extension is on (Ubuntu turns it on by default).</li>\n</ul>\n<h2 id=\"other-computers-and-reader-apps\">Other computers and reader apps</h2>\n<p>Another computer running Readwyn can pair with this one and open its libraries, and reader apps can browse them over OPDS. For a machine with no display, such as a NAS, there's <code>ryn-server</code> instead (<a href=\"/blog/headless-ebook-server-on-linux/\">Run a headless ebook server on Linux</a>). It's narrower than the app: it serves the Readwyn apps only.</p>\n<h2 id=\"which-distributions\">Which distributions</h2>\n<p>The packages need glibc 2.35 or later, which puts the floor at Ubuntu 22.04 LTS and Debian 12. On Fedora, any release that's still supported will do. They were tested on Ubuntu, Debian and Fedora, with GNOME and KDE Plasma (the <a href=\"/blog/ebook-manager-arm64-linux/\">arm64 post</a> lists each system).</p>\n<p>You'll find the steps in <a href=\"/blog/install-readwyn-on-ubuntu/\">Install Readwyn on Ubuntu from a deb package</a> and <a href=\"/blog/install-readwyn-on-fedora/\">Install Readwyn on Fedora from an rpm package</a>, and the arm64 files in <a href=\"/blog/ebook-manager-arm64-linux/\">An ebook manager on arm64 Linux</a>. When a new version is out, a <a href=\"/blog/update-readwyn-on-linux/\">card in the sidebar</a> tells you.</p>\n<p>Readwyn sticks to the freedesktop standards rather than one distribution or one desktop. If a desktop service is missing, only the feature that needs it stops working. Without speech-dispatcher, for example, the speak button does nothing, and the rest of the app carries on.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ebook-page-count-from-word-count/","url":"https://readwyn.com/blog/ebook-page-count-from-word-count/","title":"How many pages is this ebook? Estimating pages from the word count","summary":"An ebook's page count is estimated from its word count, one page for every 250 words, and carries a tilde until you type a real page count.","content_html":"<p>An ebook has no fixed pages: how many you get depends on the screen, the font and the size. So Readwyn uses a page count that stays put. It counts the words and allows one page for every 250, rounded up, which makes a book of 220,000 words “~880 pages”. The tilde marks the number as an estimate.</p>\n<p>Rounding up means a single extra word can add a page: 80,000 words is “~320 pages”, but 80,001 is “~321 pages”.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book.webp?h=6c9b53f864b64d2872b9\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book.webp?h=1970bca0dae546ce847c\" alt=\"The book screen of Middlemarch: 1871, about 880 pages, English, EPUB; the reading well shows 22 percent read, about page 196 of 880, 7 h 40 min read, and about 27 h left\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The estimate on the book screen: “~880 pages”, and your page as “~p. 196”</figcaption>\n</figure>\n<p>You'll find the estimate in four places:</p>\n<ul>\n<li>Under the author's name: “1871 · ~880 pages · English · EPUB, 1.8 MB”.</li>\n<li>In the reading well, as your page, worked out from the percent you've read: “~p. 196”, with “of 880” under it.</li>\n<li>In the Details tab, where the “Pages” row reads “~880” with “from the word count”, and “Your reading” gives “~p. 196 of 880” beside your progress.</li>\n<li>In the edit details sheet, where “Pages” shows the estimate with the note “From the word count”.</li>\n</ul>\n<p>The word count behind it is in the Details tab too, in the “Words” row. <a href=\"/blog/count-words-in-ebook/\">How to count the words in an ebook</a> has the details.</p>\n<h2 id=\"typing-a-printed-page-count\">Typing a printed page count</h2>\n<p>If you know the page count of a printed edition, you can store it:</p>\n<ol>\n<li>Click “Edit details” on the book screen, or press ⌘E.</li>\n<li>Type the page count in “Pages”.</li>\n<li>Click “Save 1 change”.</li>\n</ol>\n<p>The number then shows without the tilde, unless it happens to match the estimate exactly. “Fetch metadata” in the sheet can also fill in a missing page count, or replace the estimate, from Open Library or Google Books; the field names its source, and nothing is stored until you save.</p>\n<p>“Count again” in the Details tab puts the estimate back, so after a recount you'll need to type the page count again.</p>\n<h2 id=\"pages-in-the-reader\">Pages in the reader</h2>\n<p>The reader numbers the pages of its own layout. Its footer shows your page in that layout, and the numbers move when you change the font or resize the window. The estimate on the book screen doesn't, because it depends on the word count alone.</p>\n<p>A comic's page count isn't an estimate at all. It's the number of pages in its archive, and the Details tab says “from the archive”.</p>\n<h2 id=\"stats-and-the-3d-library\">Stats and the 3D library</h2>\n<p>In Stats, “Pages read” adds up the page counts of the books you finished in the range, and the year in review has a card for your longest book: the finished book with the most pages.</p>\n<p>How thick a book looks on the shelf of the 3D library comes from its page count, typed or estimated. A book with neither is drawn as if it had 300 pages.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ebook-reader-keyboard-shortcuts-mac/","url":"https://readwyn.com/blog/ebook-reader-keyboard-shortcuts-mac/","title":"Keyboard shortcuts for reading ebooks on a Mac","summary":"In the Readwyn reader on a Mac, ⌘F and ⌘G find text, ⌘[ takes you back to your page, ⌥⌘R opens a book privately, and ⌘= zooms a comic.","content_html":"<p>In the Readwyn reader, the arrow keys, Space, Page Up and Page Down turn the pages. ⌘F finds text in the book, and ⌘[ takes you back to where you were reading after a search.</p>\n<h2 id=\"finding-text\">Finding text</h2>\n<table><thead><tr><th>Key</th><th>What it does</th></tr></thead><tbody>\n<tr><td>⌘F</td><td>Opens the find bar</td></tr>\n<tr><td>Return</td><td>Goes to the first match at or after your place, then to the next one</td></tr>\n<tr><td>⌘G</td><td>Goes to the next match, and from the last back to the first</td></tr>\n<tr><td>⇧⌘G</td><td>Goes to the previous match</td></tr>\n<tr><td>Escape</td><td>Closes the list of matches; a second Escape closes the find bar</td></tr>\n<tr><td>⌘[</td><td>Takes you back to the page where you were reading before the search</td></tr>\n</tbody></table>\n<p><a href=\"/blog/find-text-in-ebook-on-mac/\">Finding text in an ebook</a> explains what counts as a match and what the footer’s “Stay here” does.</p>\n<h2 id=\"selecting-copying-and-notes\">Selecting, copying and notes</h2>\n<table><thead><tr><th>Key</th><th>What it does</th></tr></thead><tbody>\n<tr><td>⌘C</td><td>Copies the selected text, with a blank line between paragraphs</td></tr>\n<tr><td>⌃-click</td><td>Opens the menu of a word, a selection, or a highlight, as a right-click does</td></tr>\n<tr><td>Return</td><td>In the note editor: saves the note</td></tr>\n<tr><td>⇧Return</td><td>In the note editor: starts a new line</td></tr>\n<tr><td>Escape</td><td>Closes a popover, or ends a selection without a highlight</td></tr>\n</tbody></table>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-select.webp?h=f7d09bb574dbc078aa37\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-select.webp?h=30358a89252416889c61\" alt=\"A selection in the reader, with the popover above it: four colour dots, Add note, Copy, and Find in the book\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The popover’s Copy does what ⌘C does</figcaption>\n</figure>\n<p>Note fields, like every text field in Readwyn, accept the usual macOS editing keys. ⌥← and ⌥→ move by word, ⌘← and ⌘→ jump to the start and end of the line, ⌥⌫ deletes the word before the cursor, and holding ⇧ with a movement key selects. Highlights and notes have <a href=\"/blog/highlight-and-take-notes-in-epub-on-mac/\">a post of their own</a>.</p>\n<h2 id=\"the-book-screen-and-the-app\">The book screen and the app</h2>\n<table><thead><tr><th>Key</th><th>What it does</th></tr></thead><tbody>\n<tr><td>⌥⌘R</td><td>Opens the book in private reading (“Read privately”)</td></tr>\n<tr><td>⌘E</td><td>Opens “Edit details…” for the book</td></tr>\n<tr><td>⌥-click on a star</td><td>Sets a half star in the Review tab</td></tr>\n<tr><td>⌘,</td><td>Opens Settings, or brings it to the front</td></tr>\n</tbody></table>\n<p><a href=\"/blog/read-a-book-privately/\">Private reading</a> saves no progress, dates or time read.</p>\n<h2 id=\"comics\">Comics</h2>\n<p>The page reader for comics has keys of its own:</p>\n<table><thead><tr><th>Key</th><th>What it does</th></tr></thead><tbody>\n<tr><td>Space, ⇧Space, Page Up, Page Down, arrow keys</td><td>Turn the page (on a zoomed or tall page, ↑ and ↓ scroll it instead)</td></tr>\n<tr><td>← in a right-to-left comic</td><td>Goes to the next page</td></tr>\n<tr><td>⌘= and ⌘-</td><td>Zoom in and out, in steps of 1.25 times, up to 400 %</td></tr>\n<tr><td>⌘0</td><td>Back to Fit page or Fit width, whichever is on</td></tr>\n<tr><td>Escape</td><td>Closes the comic</td></tr>\n</tbody></table>\n<h2 id=\"on-linux\">On Linux</h2>\n<p>Ctrl takes the place of ⌘ in every table above, so ⌘F is Ctrl+F and ⌘= is Ctrl+=. Back to your page is Alt+← rather than Ctrl+[. Text fields follow Linux habits too: Ctrl+← and Ctrl+→ move by word, Home and End go to the ends of the line, and Ctrl+Backspace deletes a word.</p>\n<h2 id=\"without-a-key\">Without a key</h2>\n<p>Some things have no key. Selecting text is a drag, and making a highlight starts with one too. You <a href=\"/blog/look-up-words-in-ebooks-on-mac/\">look a word up</a> by holding it for 0.3 s, or from the menu of a right-click or Control-click. If you turn on “Turn the page with a click on its left or right side” in Settings › Reading, a click on either side of the pages turns them as well.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/edit-ebook-metadata-on-mac/","url":"https://readwyn.com/blog/edit-ebook-metadata-on-mac/","title":"How to edit ebook metadata on a Mac","summary":"Change the title, authors, series, published date, language, cover and other details of an ebook in Readwyn, and see or undo each change before you save.","content_html":"<p>You edit a book's title, authors, series, cover and the rest of its details in one sheet. Open the book and click “Edit details”, or press ⌘E. The sheet marks each field you change and stores nothing until you click “Save”. Even then your EPUB, MOBI or PDF stays byte for byte as it was: Readwyn keeps the details in its library, not in the file.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book.webp?h=6c9b53f864b64d2872b9\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book.webp?h=1970bca0dae546ce847c\" alt=\"The book screen of Middlemarch: the cover, the title and author, and the buttons Continue reading, Status: Reading, and Edit details\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>“Edit details” is next to the status</figcaption>\n</figure>\n<h2 id=\"the-sheet\">The sheet</h2>\n<p>“Edit details…” in the book's “…” menu opens it too. The cover, “Change cover” and the identifiers sit on the left. On the right is a single column: title, authors, series and number, publisher, published date, language, voice, pages, genres, tags, the section in the 3D library, and the description.</p>\n<p>A changed field fills with colour, gets a dot before its label, and shows “Was” with the stored value struck through, next to “Undo”. If the field was empty before, it says “Was Not set”. The footer counts your changes, and so does the button: “Save 2 changes”.</p>\n<p>Return in a single-line field saves. In the description, Return starts a new line and ⌘Return saves. Escape or “Cancel” closes the sheet, but if anything changed you're asked first: “Discard 2 changes?”, with the fields that would go back, “Keep editing” and “Discard changes”.</p>\n<h2 id=\"fields-with-their-own-rules\">Fields with their own rules</h2>\n<p>Typing in Authors opens a menu of matching authors from your library, each with its number of books, and “Add “Kal” as a new author” at the end. Up, Down and Return pick an item.</p>\n<p>A title is kept exactly as you type it, in any script (a Tamil title such as “பொன்னியின் செல்வன்” comes back unchanged), but it can't be blank. The published date can be a year (“1871”), a month and year (“Apr 2003”) or a full date (“29 Apr 2003”), and Readwyn stores only what you typed, so a year stays a year.</p>\n<p>If the book has no page count, Pages shows an estimate from the word count (<a href=\"/blog/metadata-sources-for-each-field/\">how it's worked out</a>), and typing a number replaces it. The language you pick from the menu becomes the book's first language, while any others stay. Changing it also sets the Voice back to Automatic.</p>\n<p>“Change cover” offers “Choose an image…” (⌘O), “Paste image” (⌘V), “Use the cover in the book file”, a fetch “From Open Library” or “From Google Books”, and “Remove cover”. A JPEG, PNG or GIF cover is stored with its bytes unchanged; a TIFF, WebP or BMP image is turned into a PNG.</p>\n<h2 id=\"fetching-instead-of-typing\">Fetching instead of typing</h2>\n<p>“Fetch metadata” in the sheet asks Open Library and Google Books for the cover, description, page count and genres. The post on <a href=\"/blog/fetch-covers-and-descriptions/\">fetching covers and descriptions</a> says what it fills and what it leaves alone.</p>\n<h2 id=\"after-you-save\">After you save</h2>\n<p>Only the fields you changed are saved; the rest stay exactly as they were. If the book is on your connected Kobo, the footer warns that “your Kobo keeps the old details until you send the book again”.</p>\n<p>The series, authors or tags of a whole set of books change in one <a href=\"/blog/edit-metadata-of-many-books/\">sheet for many books</a>: select them in the library and press ⌘E there.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/edit-metadata-of-many-books/","url":"https://readwyn.com/blog/edit-metadata-of-many-books/","title":"Edit the metadata of many books at once","summary":"Select books in Readwyn and change their authors, series, language, genres or tags in one save, with Add and Remove modes and series numbers in library order.","content_html":"<p>One sheet changes the authors, series, publisher, published date, language, genres and tags of a whole selection of books, and saves them together. Select the books in the library grid or list, then press ⌘E or click “Edit details…” in the selection bar. Any field you leave alone stays as it was in each book.</p>\n<h2 id=\"selecting\">Selecting</h2>\n<p>A plain click opens a book and clears the selection. ⌘-click adds a book or takes it out again, and Shift-click selects everything from the last selected book to the one you clicked, in list order. ⌘A selects every book that the current filters and search show. Escape closes a menu that's open in the selection bar or, if none is, clears the selection.</p>\n<p>A selected cover sits on the selection colour with a check at its corner. In the list, a selected row gets the same fill and a check at the end of the title.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library-list.webp?h=8ef4e13c111a8ae0fdcd\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library-list.webp?h=0051b07c7f0ee42ce4d5\" alt=\"The library in the list view: each row has a cover, the title and author, the status, the progress, the rating, the last read date, and the date added\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The list view: ⌘-click or Shift-click rows here as in the grid</figcaption>\n</figure>\n<p>Once two or more books are selected, a bar appears near the bottom of the library: “12 books selected”, “Edit details…”, “Add to shelf”, “Status”, “Remove…” and “Clear”.</p>\n<h2 id=\"the-sheet\">The sheet</h2>\n<p>It's titled “Edit details of 12 books”, with a line such as “12 books · George Eliot, Gabrielle Zevin and 9 other authors” under it. On the side are the first three covers as a stack and the first three titles.</p>\n<p>The sheet also has “Number in series” and the section in the 3D library. A field with the same value in every book shows it; otherwise it shows “Mixed values”. Change one and it's marked just as in the <a href=\"/blog/edit-ebook-metadata-on-mac/\">sheet for a single book</a>: the old value (or “Mixed values”) struck through, with “Undo”. The save button counts the books: “Save 12 books”.</p>\n<p>Authors, genres and tags each have three modes. “Replace with” starts from the names every book shares, and each book ends up with exactly what's in the field. “Add” starts empty and leaves the rest alone (“Each book keeps its other tags.”). “Remove” also starts empty: “Each book that has it loses it.” Names match regardless of case.</p>\n<p>To number a series, choose “Number in order” with “from 1”. It numbers the books 1, 2, 3 and so on, in whatever order the library is sorted, and the sheet tells you which order that is: “1 to 12, in the order of the library: Title, A to Z”. So sort first, then number. A book without a series gets no number.</p>\n<h2 id=\"when-you-save\">When you save</h2>\n<p>The save is all or nothing: if one book fails, none of them change, and a toast shows the error. The selection stays afterwards, minus any books the list no longer shows.</p>\n<p>On a debug build with small books, saving a tag and the authors for 1,000 books took about 0.15 s. A new language for the same 1,000 took about 0.5 s, because Readwyn reads each book's text again to pick the voice for words you look up in it.</p>\n<p>Books already on your Kobo don't change there. The footer warns you, for example “3 of these books are on your Kobo. They keep the old details until you send them again.”</p>\n<h2 id=\"status-shelves-and-removing\">Status, shelves and removing</h2>\n<p>“Status” and “Add to shelf” in the bar set a <a href=\"/blog/reading-status-and-shelves/\">status or a shelf</a> for every selected book. “Remove…” asks first: it names the books, counts the highlights that go with them, says the copies on your Kobo stay, and warns that it can't be undone.</p>\n<p>The sheet for many books has no fetch, no cover change and no “Send to device”. Those work on one book at a time, from that book's own screen.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/epub-2-or-epub-3/","url":"https://readwyn.com/blog/epub-2-or-epub-3/","title":"EPUB 2 or EPUB 3: what changes when your books are converted","summary":"Readwyn keeps your EPUB 2 file and reads an EPUB 3 copy with the same text and file paths, so highlights on the Mac or a Kobo point to the same words.","content_html":"<p>Readwyn’s reader opens EPUB 3, so when you add an EPUB 2 book (2.0 or 2.0.1) it leaves your file alone and makes an EPUB 3 copy beside it. The package and the chapter markup are brought up to EPUB 3; the text, every file path, the images and the fonts stay as they were.</p>\n<h2 id=\"inside-the-copy\">Inside the copy</h2>\n<p>Each chapter becomes valid EPUB 3 XHTML, and the book gains an EPUB 3 navigation document plus a new NCX (the EPUB 2 form of the table of contents) built from the old one. Images, fonts and style sheets keep their bytes. So does the book’s unique identifier, because the obfuscation of embedded fonts depends on it.</p>\n<p>Five real EPUB 2.0 books and five real EPUB 2.0.1 books make up the test set. In every copy the file paths and the text match the original, and <code>epubcheck</code> finds no errors. OEB packages, the format that came before EPUB, go through the same conversion and keep every file path too (<a href=\"/blog/oeb-package/\">What is an OEB package, and how to open one</a>).</p>\n<p>Your EPUB 2 stays the “Original” in the Details tab; it’s the copy that the reader opens and “Send to device” sends. An EPUB 3 that you add gets no copy: Readwyn checks it once, notes that it needs no conversion, and reads it as it is.</p>\n<h2 id=\"highlights-on-the-mac-and-on-a-kobo\">Highlights, on the Mac and on a Kobo</h2>\n<p>Because the two files have the same chapters, with the same text in each, a highlight works in either of them. Highlight a passage on the Mac, in the EPUB 3, and it still points to the same words in your EPUB 2.</p>\n<p>That matters when a Kobo holds a KEPUB made from the EPUB 2. When Readwyn syncs with a Kobo, it looks at the version in each KEPUB’s package, <code>2.0</code> or <code>3.0</code>, and maps the highlight positions through the file of that version in your library:</p>\n<ul>\n<li><strong>The library has that version.</strong> Positions go through that file. A highlight made on the Mac carries over to the EPUB 2 original, and so to the KEPUB made from it.</li>\n<li><strong>The library doesn’t have that version.</strong> Readwyn matches the chapters by their text, and positions only carry over through the text the two chapters share at their start or their end.</li>\n</ul>\n<p>Readwyn never rewrites a KEPUB on the device to make the positions line up. It doesn’t need to: a book that’s already on the Kobo still maps through the EPUB 2 in your library.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/export-kobo-highlights-on-mac/","url":"https://readwyn.com/blog/export-kobo-highlights-on-mac/","title":"How to export Kobo highlights on a Mac","summary":"Bring the highlights and notes of a Kobo into Readwyn with Device → Readwyn, then export each book’s highlights as a Markdown or a plain-text file.","content_html":"<p>Readwyn exports highlights from its Notebook, so they have to come off the Kobo first. A Device → Readwyn sync copies them into your library. Then you pick the book in the Notebook and click “Export”, which saves its highlights and notes as a Markdown or plain-text file with chapter titles and a page for each highlight.</p>\n<h2 id=\"getting-the-highlights-off-the-kobo\">Getting the highlights off the Kobo</h2>\n<p>With the Kobo plugged in, open “Sync” in the sidebar and click the review button on the “Device → Readwyn” card. The <a href=\"/blog/kobo-sync-preview/\">review sheet</a> lists each new highlight with its text, grouped by book, and you can untick any you don’t want before you apply them.</p>\n<p>A Kobo book that isn’t in your library is skipped, and so are dog-ears and stylus drawings. Highlights travel the other way too: highlight a book on the Mac and the next Readwyn → Device puts those highlights on the Kobo, as long as the book is there.</p>\n<p>On the way in, Readwyn also looks for each highlight in your copy of the book and saves its page, so the export keeps its page numbers after the Kobo is unplugged.</p>\n<h2 id=\"the-notebook\">The Notebook</h2>\n<p>“Notebook” in the sidebar collects the highlights of your whole library. On the left is every book you’ve highlighted, with counts such as “14 highlights · 3 notes”. Click one and its highlights appear on the right in reading order, under their chapter titles, each with a coloured dot and a page number.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-notebook.webp?h=c5d907fb6a66e4603293\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-notebook.webp?h=758885220be4b17df8b7\" alt=\"The Notebook: the books with highlights at the left, and the highlights of Middlemarch at the right, grouped by chapter, with their colours, pages, and a note\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The Notebook, with Copy all and Export over the highlights of a book</figcaption>\n</figure>\n<p>Chips above the list filter it by <a href=\"/blog/kobo-highlight-colours/\">highlight colour</a>, and “Search highlights and notes” looks through every highlight and note in the library.</p>\n<h2 id=\"export\">Export</h2>\n<p>Above the book’s highlights are three buttons: “Copy all”, “Export”, and “Open book”. “Export” offers “Markdown…” and “Plain text…”. Either one opens the save panel in Documents with a name already filled in, such as “Middlemarch - Highlights.md”. The file contains the book’s title and authors, each chapter title, and every highlight with its page and its <a href=\"/blog/kobo-notes-in-your-library/\">note</a>.</p>\n<p>Which page you get depends on the book. If it has a page list, it’s the printed page. If not, Readwyn estimates the page from the book’s page count, and a book without a page count gets a percentage instead.</p>\n<p>The Markdown file suits a notes app that works on plain files. Obsidian’s help says it keeps notes as <a rel=\"external\" href=\"https://obsidian.md/help/data-storage\">Markdown-formatted plain text files</a> in a vault, which is a folder on your disk. Choose a folder inside your vault in the save panel, and the book’s highlights turn up there as a note. The post on <a href=\"/blog/kindle-highlights-in-ebook-library/\">Kindle highlights</a> shows how such a file begins, and how Notion imports one.</p>\n<p>If you only want the text somewhere else, “Copy all” puts every annotation of the book on the clipboard. Hover over a single highlight and “Copy” takes just that one, page included, or you can select part of a highlight and copy it. The Highlights tab on the book screen has “Copy all” too, next to “Open in Notebook”.</p>\n<h2 id=\"highlights-from-another-edition\">Highlights from another edition</h2>\n<p>Sometimes a highlight was made in a different edition from the one in your library. If Readwyn can’t find its place, page, or context in your copy, it goes to the end of the list under “Not found in this copy”, with a count and the name of the device it came from. It isn’t thrown away.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/fetch-covers-and-descriptions/","url":"https://readwyn.com/blog/fetch-covers-and-descriptions/","title":"Fetch covers and descriptions for your ebooks","summary":"Fetch a cover, a description, a page count and genres for a book from Open Library and Google Books in Readwyn, and check each value before anything is saved.","content_html":"<p>A fetch asks Open Library and Google Books for a book's cover, description, page count and genres, and puts what it finds into the edit details sheet. Nothing is stored until you click “Save”, so you see every value, and where it came from, before it touches your library.</p>\n<h2 id=\"starting-a-fetch\">Starting a fetch</h2>\n<p>On the book screen, open the “…” menu and choose “Fetch metadata”. The Details tab has the same button: “Fetch metadata”, or “Fetch again” once the book has been fetched. If the edit details sheet is already open, its own “Fetch metadata” does the same job.</p>\n<p>A fetch always works on one book. The <a href=\"/blog/edit-metadata-of-many-books/\">sheet for many books</a> has no fetch button.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book.webp?h=6c9b53f864b64d2872b9\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book.webp?h=1970bca0dae546ce847c\" alt=\"The book screen of Middlemarch: the cover, the title, the reading progress, and the description on the Overview tab\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The description shows on the Overview tab</figcaption>\n</figure>\n<h2 id=\"what-it-fills\">What it fills</h2>\n<p>A fetch brings in the description and the cover, any identifiers the book is missing, and a page count if the book has none or only the estimate from its word count. It only adds genres that the book lacks, and the sheet says which, for example “Added Victorian and Romance”. It never adds tags: fetched genres go in the genre field, and your tags remain your own.</p>\n<h2 id=\"checking-the-result\">Checking the result</h2>\n<p>Each fetched field shows its source on the right, for example “Open Library”, and a fetched cover is marked “New cover · Open Library”. The sheet sums the fetch up as “Fetched from Open Library. 4 fields changed.”, and the save button reads “Save 4 changes”.</p>\n<p>If you don't want one of the values, click “Undo” on its field, as with any other edit in the sheet. “Cancel” or Escape closes the sheet without saving anything, after an alert if fields changed.</p>\n<p>Once you save, the labels go, and the Details tab's <a href=\"/blog/metadata-sources-for-each-field/\">“Details from” row</a> keeps the source and the date of the fetch.</p>\n<h2 id=\"changing-only-the-cover\">Changing only the cover</h2>\n<p>Open the edit details sheet (⌘E) and click “Change cover”. Under “Fetch a cover” are “From Open Library” and “From Google Books”, and each asks that source only. As with a full fetch, the new cover waits in the sheet until you save. The other items of that menu, from “Choose an image…” to “Remove cover”, are in <a href=\"/blog/edit-ebook-metadata-on-mac/\">the post on editing a book's details</a>.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/filter-library-by-language-series-author-format/","url":"https://readwyn.com/blog/filter-library-by-language-series-author-format/","title":"Filter your library by language, series, author, or format","summary":"How Readwyn's Filter menu narrows a library by shelf, genre, tag, author, series, language, or comics versus ebooks, and how the filters combine with search.","content_html":"<p>The “Filter” menu above your books narrows the library by shelf, genre, tag, author, series, language, or format (“Comics” or “Ebooks”), and it has a switch for the books on your connected Kobo. Each filter you turn on shows up as a chip above the books (the Kobo switch is the exception), and the badge on “Filter” says how many are on. The line before the chips counts what's left: “Showing 4 of 412 books”.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library-filter.webp?h=94c5a584efe6ac8eca30\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library-filter.webp?h=c3782eccd71eb19fe9d1\" alt=\"The library with the Filter menu open: the shelf Victorian and the author George Eliot are ticked, the languages English, Tamil, French, and German are listed with their counts, and the switch Only books on Kobo Clara 2E is off\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Two filters on: a shelf and an author</figcaption>\n</figure>\n<h2 id=\"what-s-in-the-menu\">What's in the menu</h2>\n<ul>\n<li><strong>Shelf.</strong> Two shelves with their counts (more, if you've ticked more). A link such as “Show all 9 shelves” opens the rest.</li>\n<li><strong>Genre.</strong> Up to six chips, each with its count: the genres with the most books first and “Other” last. A link such as “All 14” shows the rest. One genre is on at a time.</li>\n<li><strong>Tag.</strong> The tags of your books, as Edit details shows them: three rows with their counts, and “Find a tag” for the others, which shows up to eight matches. Tick more than one and you see the books with any of them.</li>\n<li><strong>Author.</strong> The three authors with the most books, each with its count. Type a name into “Find an author” and up to eight matches appear.</li>\n<li><strong>Series.</strong> A drop-down of your series. Unlike the shelf, tag, and author lists, it takes one series at a time; “Any series” clears it.</li>\n<li><strong>Language.</strong> Each language is named in itself: “English”, “தமிழ்”, “Français”, “Deutsch”. The chips use the same names.</li>\n<li><strong>Format.</strong> “Comics” and “Ebooks” with their counts, shown only when the library has <a href=\"/blog/comics-and-ebooks-in-one-library/\">both kinds</a>.</li>\n<li><strong>The connected device.</strong> “Only books on Kobo Clara 2E”, with how many of your books are on it.</li>\n</ul>\n<p>“Clear all” turns everything off at once.</p>\n<h2 id=\"how-filters-combine\">How filters combine</h2>\n<p>Tick two shelves and you see books on either shelf. Add an author as well and only that author's books on those shelves remain. Each choice shows as its own chip, such as “Shelf: Victorian” or “Author: George Eliot”.</p>\n<p>The status tabs follow along. With “Author: George Eliot” on, “All 4” and “Read 1” count only her books. The numbers inside the Filter menu don't change, though: they always count the whole library.</p>\n<h2 id=\"search-and-sort\">Search and sort</h2>\n<p>Search looks at titles, authors and series, and ignores case. As you type, a list drops down with the matching books first, then matching authors and series with their books. Press Return on an author or a series in the results and it becomes a filter chip; the search box clears.</p>\n<p>“Sort” then orders whatever is showing by date added, rating, last read, percent read, or title. Sort by “Last read” with the latest first and the books you're in the middle of come to the top.</p>\n<h2 id=\"genre-from-the-menu-or-from-stats\">Genre, from the menu or from Stats</h2>\n<p>Every book has one primary genre, and that's the genre the menu filters by. The chips use the names of Readwyn's own list of genres, such as “Literary fiction”. If every book in the library is “Other”, the menu leaves the Genre section out.</p>\n<p>The Stats screen sets the same filter. Click a genre among the top genres there and the library opens with just that filter, shown as a chip such as “Genre: Literary fiction”. Authors and series work the same way from Stats.</p>\n<h2 id=\"speed-with-10-000-books\">Speed with 10,000 books</h2>\n<p>Even with 10,000 books, turning the shelf, author, series or language filter on or off took 15 to 67 ms to show the result, measured in a virtual Mac on a <a href=\"/blog/large-ebook-library-on-mac/\">test library of that size</a>. Searches, sorts and status tabs took up to 71 ms. Each figure is the slowest of three runs, against a target of 100 ms.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/find-text-in-ebook-on-mac/","url":"https://readwyn.com/blog/find-text-in-ebook-on-mac/","title":"How to find text in an ebook on a Mac","summary":"Press ⌘F in the Readwyn reader to find text in the open book: how a query matches, how to step through the matches, and how to get back to your page.","content_html":"<p>Press ⌘F in the Readwyn reader, or click “Find” in the top bar, and start typing. Results come in as you type: a list under the field shows every match with its chapter, its page, and the words around it. Click a match to go there.</p>\n<h2 id=\"the-find-bar\">The find bar</h2>\n<p>The bar opens in the middle of the top bar. It has the field, the position (“62 of 284”), up and down buttons (“Previous match” and “Next match”), and “Done”.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-find.webp?h=66e4acb1e2683d5c64ed\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-find.webp?h=d61537c5bc78f92487bd\" alt=\"The find bar with the query heart, 1 of 14, and the list of matches under it, each with Book II · Chapter XX, its page, and its context\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Click a row to jump to that match</figcaption>\n</figure>\n<p>The list opens with a count, for example “41 matches · in any case, also inside words”. Each row gives the chapter, the page, and the match with about 40 characters either side. If the reader hasn’t laid that page out yet, the row shows a percentage instead.</p>\n<p>On the page itself every match is tinted, and the current one is also underlined in your accent colour.</p>\n<h2 id=\"moving-between-matches\">Moving between matches</h2>\n<ul>\n<li>Click a row, or press Return in the field. The first Return goes to the first match at or after your place, and each one after that moves on.</li>\n<li>⌘G, or the down button in the bar, goes to the next match. After the last one, it wraps round to the first.</li>\n<li>⇧⌘G, or the up button, goes back one.</li>\n<li>Escape closes the list, and a second Escape (or “Done”) closes the bar.</li>\n</ul>\n<h2 id=\"what-counts-as-a-match\">What counts as a match</h2>\n<p>A query matches anywhere in the text, inside words too, and ignores case. “heart” finds “hearth” and “Sweetheart”, “Heart” finds “heart”, and “sss” finds “potsss”.</p>\n<p>Accents, spaces and punctuation have to match as you typed them. “ville” won’t find “villé”, and “gave up” won’t find “gave—up”.</p>\n<p>Matches never overlap, so “ss” occurs twice in “ssss”, and none runs across a chapter break. If there’s nothing to find, the field says “No matches”.</p>\n<h2 id=\"getting-back-to-your-page\">Getting back to your page</h2>\n<p>Searching never loses your place. As soon as you jump to a match, the footer offers “Back to p. 196” (with your own page number) and “Stay here”, and a mark on the progress track shows where you were. The first button, or ⌘[, takes you back. “Stay here” makes the current page your new place. Turning the pages back to your place settles it too.</p>\n<p>Until you decide, Readwyn keeps your saved place where it was, so if you close the book mid-search, it reopens where you were reading. Both buttons stay in the footer after you close the find bar.</p>\n<h2 id=\"from-a-selection\">From a selection</h2>\n<p>Select some text and the selection popover has a Find in the book button, which opens the bar with your selection already filled in. Right-clicking a selection or a highlight gives the same option, as Find “…” in the book with your text in the quotes.</p>\n<h2 id=\"how-fast-it-is\">How fast it is</h2>\n<p>Readwyn’s target is 100 ms or less from the search to the first frame that shows the result. In the latest measurement across four test libraries of 10 to 10,000 books, the slowest of three runs of each English query took between 11 and 33 ms. One query missed the target: a Tamil word took 269 ms in the 10-book library, against 45 to 46 ms in the others.</p>\n<p>The scan of the text on its own is quicker still. The largest book in the Readwyn test corpus is a Portuguese dictionary of 11.85 million characters, and a search for “e” in it finds 941,208 matches in 16 ms. In Ulysses, every query takes 2 ms or less. The search reads the text that the reader already holds, and it keeps no copy of the book.</p>\n<h2 id=\"one-book-at-a-time\">One book at a time</h2>\n<p>Find only searches the open book; Readwyn doesn’t search the text of every book in a library at once. To find a book by its title, author or series, use the library search. Your <a href=\"/blog/all-highlights-in-one-notebook/\">highlights and notes</a> from every book have a search of their own in the Notebook, “Search highlights and notes”.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/guest-access-to-ebook-library/","url":"https://readwyn.com/blog/guest-access-to-ebook-library/","title":"Give another Mac guest access to your library","summary":"Pair a Mac at the Guest level: it can open every book in your libraries, but only in private reading, so nothing it reads is saved.","content_html":"<p>A Mac paired at the Guest level can open every book in your libraries, but each one opens in private reading: it saves no progress, no time read, no highlights, and no status. It can't add, remove, or edit books either. Your reading data stays exactly as it was, and you can <a href=\"/blog/read-guest-admin-access-levels/\">raise or lower the level</a> whenever you like.</p>\n<h2 id=\"pairing-as-a-guest\">Pairing as a guest</h2>\n<p>First, “Readwyn server” has to be on, with your network chosen in “Listen on”. The <a href=\"/blog/pair-second-mac/\">steps for pairing a second Mac</a> go through that, and through the errors you might see. Then:</p>\n<ol>\n<li>Choose “Pair a device…”. The sheet shows your Mac's address, a 6-digit code, and “It can” with a level. Under the code it says “Works once, for the next 10 minutes.”</li>\n<li>Set “It can” to Guest. The menu describes it as “Read books. Nothing is saved; every book opens privately.”</li>\n<li>On the other Mac, open Settings › Other devices › Another device, type the address, choose Connect, and enter the code.</li>\n</ol>\n<p>If the other Mac sends a request instead of entering a code, the request appears in “Paired devices” with a level menu, Deny, and Allow. Set the menu to Guest before you allow it. Allowing from the notification gives Read instead, though you can lower it to Guest in the list afterwards.</p>\n<h2 id=\"what-the-guest-mac-shows\">What the Guest Mac shows</h2>\n<p>Under “Your access” in Settings › Other devices, the Guest Mac says “Read every book privately. Nothing is saved.” It doesn't show what it can't use:</p>\n<ul>\n<li>The book screen's main button is “Read privately”, with no status, no “Edit details”, and no “…” menu, so there's no “Available offline” either.</li>\n<li>The reader's top bar shows the pill “Private reading · nothing is saved”.</li>\n<li>Selecting text offers no highlight colours; the popover says “No highlights in private reading”.</li>\n<li>There's no Stats in the sidebar.</li>\n<li>The 3D library shows the shelves without the status and progress of each book.</li>\n</ul>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-private.webp?h=eb31eb5ff26f565de2ff\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-private.webp?h=ef3b87892f49a7311e3e\" alt=\"A two-page spread in the reader with the pill Private reading, nothing is saved, in the top bar, and a selection popover that says No highlights in private reading\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Each book on a Guest Mac opens in private reading</figcaption>\n</figure>\n<p>Looking up a word, hearing it spoken, copying text, and finding text in the book all still work, with the Guest Mac's own dictionaries and voices. It's the same private reading that <a href=\"/blog/read-a-book-privately/\">“Read privately”</a> gives you for one book on your own Mac.</p>\n<h2 id=\"changing-the-level-or-removing-the-mac\">Changing the level or removing the Mac</h2>\n<p>In “Paired devices” on your Mac, each device has a level menu and “Remove…”. The other Mac picks up a new level within about 30 seconds while its window is open, without having to reconnect. “Remove…” asks first, and a removed Mac can't connect again until it pairs with a new code.</p>\n<h2 id=\"when-the-guest-mac-asks-for-more\">When the Guest Mac asks for more</h2>\n<p>The Guest Mac has “Ask for Read access…” under “Your access”. Your Mac then posts a notification such as “Living room iMac asks for Read access.”, with “Allow gives it the Read level. You can change it in Settings.” You can answer there or in “Paired devices”. If you deny it, the Guest Mac keeps its level and shows a line such as “Sid’s iMac did not allow Read access.”</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/headless-ebook-server-on-linux/","url":"https://readwyn.com/blog/headless-ebook-server-on-linux/","title":"Run a headless ebook server on Linux","summary":"Readwyn's headless server, ryn-server, keeps your libraries on a Linux NAS or home server with no display. It runs under systemd and pairs with the app by a six-digit code.","content_html":"<p><code>ryn-server</code> is Readwyn's server without the app. It holds your libraries on a Linux machine with no display, such as a NAS or a home server, and the Readwyn apps on a Mac or a Linux computer connect to it just as they would to a Mac, with the same Guest, Read and Admin levels. It works on the home network only.</p>\n<p>Reader apps can't use it yet, though. No <code>ryn-server</code> command turns on the OPDS catalogue or adds a reader app. If you want your phone's reader app to browse the libraries, keep them in the Readwyn app instead (<a href=\"/blog/home-ebook-server-on-mac/\">Run a home ebook server on a Mac</a>).</p>\n<h2 id=\"choose-the-package\">Choose the package</h2>\n<p>The server comes as a deb, an rpm and a tar file, for x86_64 and arm64:</p>\n<table><thead><tr><th>System</th><th>x86_64</th><th>arm64</th></tr></thead><tbody>\n<tr><td>Ubuntu and Debian</td><td><code>readwyn-server_&lt;version&gt;-1_amd64.deb</code></td><td><code>readwyn-server_&lt;version&gt;-1_arm64.deb</code></td></tr>\n<tr><td>Fedora</td><td><code>readwyn-server-&lt;version&gt;-1.x86_64.rpm</code></td><td><code>readwyn-server-&lt;version&gt;-1.aarch64.rpm</code></td></tr>\n<tr><td>Other Linux systems</td><td><code>readwyn-server-&lt;version&gt;-x86_64.tar.gz</code></td><td><code>readwyn-server-&lt;version&gt;-aarch64.tar.gz</code></td></tr>\n</tbody></table>\n<p>The tar file holds <code>ryn-server</code>, the systemd unit <code>readwyn-server@.service</code> and a <code>README</code>. The <a href=\"/download/#server\">download page</a> lists every file.</p>\n<h2 id=\"start-it\">Start it</h2>\n<p>The deb and rpm packages install the systemd unit <code>readwyn-server@.service</code>. Enable it for your user (put your user name in place of USER):</p>\n<pre><code>sudo systemctl enable --now readwyn-server@USER\n</code></pre>\n<p>From then on the server starts with the machine, even when nobody logs in, and systemd restarts it after a crash.</p>\n<p>On its first start it listens on every network, on port 7313, because its whole job is serving other devices. <code>ryn-server listen local</code> limits it to the machine itself, and <code>ryn-server listen all</code> opens it up again. Put a port number after either to change the port.</p>\n<h2 id=\"pair-the-readwyn-app\">Pair the Readwyn app</h2>\n<p>As the server's user, make a pairing code:</p>\n<pre><code>ryn-server code\n</code></pre>\n<p>The code has six digits and works once, for 10 minutes. It gives the Read level unless you ask for another with <code>ryn-server code guest</code> or <code>ryn-server code admin</code>. In Readwyn, open Settings › Other devices › Another device, type the server's address and click Connect. Six boxes then appear for the code; enter it and click Pair, and the app shows the server's libraries.</p>\n<p>A device can also ask to pair without a code. <code>ryn-server devices</code> lists the paired devices and any requests waiting, and <code>ryn-server allow DEVICE</code> or <code>ryn-server deny DEVICE</code> answers one.</p>\n<h2 id=\"the-other-commands\">The other commands</h2>\n<table><thead><tr><th>Command</th><th>What it does</th></tr></thead><tbody>\n<tr><td><code>ryn-server status</code></td><td>Shows how many devices are connected and the addresses the server listens on</td></tr>\n<tr><td><code>ryn-server level DEVICE LEVEL</code></td><td>Sets the level of a device: <code>guest</code>, <code>reading</code> (the Read level) or <code>admin</code></td></tr>\n<tr><td><code>ryn-server remove DEVICE</code></td><td>Removes a paired device</td></tr>\n<tr><td><code>ryn-server version</code></td><td>Prints the version</td></tr>\n</tbody></table>\n<p>Each command talks to the running server over its socket, so run it as the server's user. If no server is running, the command says so and exits with status 1.</p>\n<h2 id=\"where-your-libraries-live\">Where your libraries live</h2>\n<p>Libraries and settings go in the user's <code>~/.local/share/readwyn</code>, caches in <code>~/.cache/readwyn</code>. A machine without a display has no Secret Service (the keyring of a Linux desktop), so the server keeps the Goodreads sign-in token in <code>secrets.json</code> in the same folder, where only that user can read it. Removing the package leaves all of this in place.</p>\n<h2 id=\"updates\">Updates</h2>\n<p>When a newer server is out, the Readwyn app on a Linux computer connected to it shows an <a href=\"/blog/update-readwyn-on-linux/\">update card</a> such as “Readwyn server 0.2.0 is out”. The new package goes on over the old one and restarts the running server itself.</p>\n<h2 id=\"a-restart-with-nobody-logged-in\">A restart with nobody logged in</h2>\n<p>On 22 October 2026 the deb of each architecture ran on Ubuntu 24.04 server with no display, holding one library and one book. The Readwyn app on a Mac paired with the server through a code from <code>ryn-server code</code> and opened the library and its book. Then the Linux machine restarted. With nobody logged in, the server answered again, and the app reconnected by itself.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/hear-how-a-word-sounds/","url":"https://readwyn.com/blog/hear-how-a-word-sounds/","title":"Hear how a word sounds while you read","summary":"Readwyn’s lookup can say a word aloud, offline, in a voice that matches the book’s language and its variety of English. You can also pick the voice.","content_html":"<p>Readwyn’s lookup popover has a speaker button right after the pronunciation. Click it and you hear the word, in a voice that matches the language of its book. Readwyn uses the voices already on your computer, so this works without a network connection.</p>\n<h2 id=\"speak-a-word\">Speak a word</h2>\n<p>Hold a word on the page for a moment to <a href=\"/blog/look-up-words-in-ebooks-on-mac/\">look it up</a>. The speaker button’s tooltip says “Speak”. While the word plays, the speaker fills in, the tooltip changes to “Stop”, and a line beside the button names the voice, for example “Speaking, English (UK)”. Click again to stop it.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-lookup.webp?h=100a081b3204e86878eb\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-lookup.webp?h=048cd0b33c717f3ab5f2\" alt=\"The lookup popover for forlornness, with the pronunciation and a speaker button after it\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The speak button follows the pronunciation</figcaption>\n</figure>\n<p>You hear the word as the page prints it, “forlornness” rather than the headword “forlorn”. When the popover shows two readings, the phrase and its single word, you hear the one that is open.</p>\n<p>If the entry has no pronunciation, the button appears on its own under the headword. A word with no dictionary entry gets no button at all.</p>\n<p>Speech stops when you press Escape, click outside the popover, look up another word, start a drag, or switch between the two readings of a phrase. It works in <a href=\"/blog/read-a-book-privately/\">private reading</a> too.</p>\n<h2 id=\"which-voice-you-hear\">Which voice you hear</h2>\n<p>For a book in English, Readwyn also picks a variety: UK, US, Australia, Canada, or India. Books in other languages don’t get one.</p>\n<p>The variety is settled when you add the book, from the file and its text. Spelling counts (“-or” against “-our”), and so do local words such as “arvo” (Australian) or “lakh” (Indian). The ISBN, the publisher, the language tag and any “first published in” line count as well. If none of these decides it, the default voice for that language speaks the word.</p>\n<p>A regional voice wins over your system default. On a Mac whose default voice is English (India), a word from an Australian book is spoken in an Australian voice. If macOS has no voice for exactly that region (ask it for Canadian English and it offers a US one), your default voice speaks instead.</p>\n<p>On Linux, the voices come from speech-dispatcher, the system’s speech service. The app package recommends it but doesn’t bundle it. Without it the button does nothing, and the rest of the app carries on as usual.</p>\n<h2 id=\"pick-the-voice-yourself\">Pick the voice yourself</h2>\n<p>Open the book screen and choose “Edit details”. The Voice field is under Language and Pages. It offers “Automatic” with the detected voice and where it came from (for example “Automatic · from the setting”, which means the local words of the story decided it), every voice your Mac has for the book’s language, and “System voice”.</p>\n<p>Pick a voice and the field reads “Chosen by you”. Pick “Automatic” to go back to the detected one. If you change the book’s language, the voice resets to Automatic.</p>\n<h2 id=\"words-you-ve-saved\">Words you’ve saved</h2>\n<p><a href=\"/blog/save-looked-up-words/\">My Words</a> in the Notebook has the same button. Click a word group to open its full entry, then click the speaker after the pronunciation. This time you hear the headword, in the voice of the book you saved the newest term from. If that term has lost its book (you deleted the book, say), your default voice says it.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/highlight-and-take-notes-in-epub-on-mac/","url":"https://readwyn.com/blog/highlight-and-take-notes-in-epub-on-mac/","title":"How to highlight and take notes in an EPUB on a Mac","summary":"Highlight a passage in the Readwyn reader in one of four colours, add a note to it, and change or delete it later, with the exact buttons and keys.","content_html":"<p>A highlight in Readwyn starts with a drag. Drag over the passage in the reader and let go of the mouse button: four colour dots appear above the selection, and clicking one highlights the passage in that colour. For a note, click “Add note” in the same popover instead.</p>\n<h2 id=\"selecting-text\">Selecting text</h2>\n<p>The drag always takes whole words. While you drag, the words take a pale tint of the accent colour, so a selection can’t be mistaken for a highlight. Nothing is saved yet.</p>\n<p>When you let go, the selection popover opens above the passage (or below it, if the top bar is in the way). It has:</p>\n<ul>\n<li>the four colour dots (yellow, pink, blue and green), with a ring round the current colour,</li>\n<li>“Add note”,</li>\n<li>Copy (⌘C), which copies the text with a blank line between paragraphs,</li>\n<li>Find in the book, which opens the find bar on the selected text, with a list of every place it appears.</li>\n</ul>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-select.webp?h=f7d09bb574dbc078aa37\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-select.webp?h=30358a89252416889c61\" alt=\"A selection in the reader, with the popover above it: four colour dots, Add note, Copy, and Find in the book\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Click a colour to highlight the passage, or Add note to write a note</figcaption>\n</figure>\n<p>Escape, or a click outside the popover, drops the selection without highlighting anything.</p>\n<p>Once you’ve picked a colour, the popover of the new highlight takes over, so you can still switch colour or add a note straight away. That colour also becomes the current one: “Add note” and the right-click “Highlight” use it until you pick another.</p>\n<h2 id=\"from-the-right-click-menu\">From the right-click menu</h2>\n<p>If you prefer menus, right-click the selection (or Control-click it). The menu starts with a “Look up “…”” item for each possible match of the word under the pointer, followed by “Highlight” with a dot in the current colour, “Add note”, “Copy” and “Find “…” in the book”. Right-clicking a single word selects it first.</p>\n<h2 id=\"adding-a-note\">Adding a note</h2>\n<p>“Add note” highlights the selection in the current colour and opens the note editor. Type your note, then:</p>\n<ul>\n<li>“Save note” or Return saves it,</li>\n<li>Shift+Return starts a new line (the hint reads “⇧Return for a new line”),</li>\n<li>Cancel takes you back to the popover.</li>\n</ul>\n<p>A highlight with a note gets a small raised mark after its last word. Click the highlight and its popover shows the note above the colours; click the note to edit it.</p>\n<h2 id=\"changing-or-deleting-a-highlight\">Changing or deleting a highlight</h2>\n<p>A highlight’s popover has the four colour dots, “Add note” (“Edit note” once there’s a note), Copy, and a trash button whose tooltip is “Delete highlight”. Pick another colour to recolour the highlight, or the trash button to delete it. Escape or a click outside closes the popover and leaves the highlight alone.</p>\n<p>Right-clicking a highlight gives a menu with the “Look up “…”” items, “Add note” (or “Edit note”), “Copy”, “Find “…” in the book”, and “Delete highlight” at the bottom. From this menu, Copy and Find take the whole highlight.</p>\n<h2 id=\"where-your-highlights-go\">Where your highlights go</h2>\n<p>Each highlight is saved with its colour, its note and an anchor in the text, so it stays with the words you marked. Besides the page itself, it shows up in the Highlights tab of the book screen and in the <a href=\"/blog/all-highlights-in-one-notebook/\">Notebook</a>, which gathers the highlights of every book in the library and can export a book’s highlights as Markdown or plain text. If you also read on a Kobo, a sync carries highlights and notes between the Kobo and your library, in whichever direction you choose (<a href=\"/blog/export-kobo-highlights-on-mac/\">Kobo highlights on a Mac</a>).</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/home-ebook-server-on-mac/","url":"https://readwyn.com/blog/home-ebook-server-on-mac/","title":"Run a home ebook server on a Mac","summary":"Turn a Mac into a home ebook server with the Readwyn server setting: it keeps running after you quit the app, starts at login, and serves other Macs and reader apps.","content_html":"<p>A Mac with Readwyn is already a server for its own libraries. The setting “Readwyn server” opens that server to the other devices in your home: Macs that you pair, and reader apps on phones, tablets, and e-readers. With it on, the server keeps running after you quit Readwyn and starts when you log in, so your libraries stay available while the Mac is on.</p>\n<h2 id=\"turn-on-readwyn-server\">Turn on Readwyn server</h2>\n<p>Open Settings › Other devices and turn on “Readwyn server”. The line under it says “Other devices can open the libraries of this Mac. It keeps running when you quit Readwyn, and starts at login.” macOS then shows its own notice that Readwyn added a background item.</p>\n<p>The state line changes to something like “Running · 2 devices connected”, and the line below names this Mac: “This Mac: Sid’s iMac”. From now on the server runs in the background, and the Readwyn window talks to it over a local socket on the same Mac. Each macOS user on the Mac has a server and libraries of their own.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-settings-devices.webp?h=5e2bf654f800618406e3\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-settings-devices.webp?h=1263ffab1becc7426e6c\" alt=\"Settings › Other devices: Readwyn server is on, Running with 2 devices connected; This Mac: Sid’s iMac at 192.168.1.20:7313; Show libraries from This Mac; Listen on Wi-Fi · 192.168.1.20, port 7313; the paired devices with their levels; and the reader apps\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>Readwyn server on, with paired devices and reader apps</figcaption>\n</figure>\n<h2 id=\"choose-the-network\">Choose the network</h2>\n<p>“Listen on” starts as “This Mac only”. For other devices to reach the Mac, choose your home network, for example “Wi-Fi · 192.168.1.20”, and keep port 7313 or type another. The Mac’s address then joins the line under the switch, “This Mac: Sid’s iMac · 192.168.1.20:7313”, and Readwyn warns: “Anything on your Wi-Fi can reach this Mac. Only paired devices can open the libraries.” macOS may ask whether Readwyn can accept incoming connections; allow it.</p>\n<p>The server is meant for your home network only. Nothing in Readwyn opens it to the internet.</p>\n<h2 id=\"pair-another-mac\">Pair another Mac</h2>\n<p>Click “Pair a device…”. The sheet shows the address to type, with Copy, a 6-digit code, and what the device will be able to do (“It can”). The code works once, for 10 minutes. On the other Mac, open Settings › Other devices › Another device, type the address, click Connect, and enter the code.</p>\n<p>Each paired device has a level:</p>\n<ul>\n<li><strong>Guest</strong>: reads the libraries, and each book opens in private reading.</li>\n<li><strong>Read</strong>, the default: also saves progress, status, highlights, notes, and My Words.</li>\n<li><strong>Admin</strong>: also adds, removes, and edits books, and changes the server.</li>\n</ul>\n<p>A device can also ask to pair without a code. The request shows up in “Paired devices” with Deny and Allow, and as a notification such as “Sid’s MacBook Air asks to connect.” In Settings you can first pick its level from the menu beside the buttons; it starts at Read. Allow in the notification always gives Read. <a href=\"/blog/read-guest-admin-access-levels/\">Read, Guest, and Admin</a> explains what each level can do.</p>\n<h2 id=\"serve-reader-apps\">Serve reader apps</h2>\n<p>Under “Reader apps”, turn on “Share with reader apps (OPDS)”. The Mac then also serves an OPDS catalogue on a second port, 7314, for example <code>http://192.168.1.20:7314/opds</code>. Reader apps such as KOReader, Thorium, or Moon+ Reader can browse it and download books, but they can’t change anything. <a href=\"/blog/opds-catalogue-for-koreader/\">Set up an OPDS catalogue for KOReader</a> gives the steps.</p>\n<h2 id=\"the-menu-bar-icon\">The menu bar icon</h2>\n<p>While the server is on, the Readwyn mark sits in the menu bar. Its menu has “Readwyn server is running”, a status line such as “2 devices connected · 1 asks to connect”, “Open Readwyn”, “Pair a device…”, “Connected devices…”, “Turn off Readwyn server”, and “Don’t show this icon”.</p>\n<h2 id=\"when-the-setting-is-off\">When the setting is off</h2>\n<p>With the setting off, no other device can connect, even while Readwyn is open, and the state line says “Off”. Your paired devices stay paired, and they connect again when you turn it back on.</p>\n<h2 id=\"a-server-without-a-mac\">A server without a Mac</h2>\n<p>For a machine without a display, such as a NAS, Readwyn has a server for Linux: <a href=\"/blog/headless-ebook-server-on-linux/\">Run a headless ebook server on Linux</a>. It serves the Readwyn apps, but not reader apps yet.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/hours-read-this-year/","url":"https://readwyn.com/blog/hours-read-this-year/","title":"How many hours did you read this year?","summary":"Pick Year in Readwyn's Stats to see the hours you've read since 1 January, with a monthly average and a bar for each month.","content_html":"<p>Choose “Year” at the top of Stats. The large figure is your answer: the time you spent in Readwyn's reader between 1 January and 31 December, for example 249 h read. The monthly average sits beside it, and a bar for each month runs underneath. The Previous arrow takes you back a year.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-stats-year.webp?h=c0ea8ac829fdb8e9570c\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-stats-year.webp?h=eb5f65b0c94273d46135\" alt=\"Reading stats for 2026: 249 h read, a monthly average of 24 h 52 min, a bar for each month from January to October, and the figures of the year\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The hours of a year, month by month</figcaption>\n</figure>\n<p>Below the chart come the year's other figures: books finished (with “Up from 31 in 2025” when you've beaten last year), pages read, your streak, reading speed, and average rating. The same hours, books, and pages fill the first card of <a href=\"/blog/year-in-books/\">your year in books</a>.</p>\n<h2 id=\"where-the-hours-come-from\">Where the hours come from</h2>\n<p>Only time recorded in Readwyn's own reader counts, as the line “Time from the Readwyn reader” says. The reader <a href=\"/blog/track-reading-time-automatically/\">records that time from your page turns</a> and leaves out long pauses. Private reading adds nothing, and neither does a book you've removed.</p>\n<p>Reading that runs past midnight on New Year's Eve is split between the two years, at midnight in your own Mac's time zone. Late nights during the rest of the year <a href=\"/blog/busiest-reading-hour-and-weekday/\">split the same way</a>, between two days.</p>\n<h2 id=\"why-kobo-time-isn-t-in-the-year\">Why Kobo time isn't in the year</h2>\n<p>A Kobo sync brings each book's reading time from the device as one total, with no date. Readwyn can't tell which year that time belongs to, so the Year range leaves it out. All time includes it and spells it out under the chart, for example “1,105 h from the Readwyn reader by month, and 64 h from the Kobo without a date. The bars show the time with a date.”</p>\n<h2 id=\"the-monthly-average\">The monthly average</h2>\n<p>For the current year, the average divides by the months so far, not by twelve: in October it's the year's time divided by ten. A past year divides by twelve. If the average is above zero but under a minute, it reads “Monthly average under 1 min”.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-long-calibre-import-takes/","url":"https://readwyn.com/blog/how-long-calibre-import-takes/","title":"How long does it take to import a calibre library?","summary":"Readwyn estimates a thousand 2 MB EPUBs at about a minute to import and the same books as MOBI at about 26 minutes, then corrects the time left as books arrive.","content_html":"<p>Mostly it depends on how many books need converting. By Readwyn’s own estimate, a thousand EPUBs of 2 MB each take about a minute, and the same thousand as MOBI files about 26 minutes. You see the figure for your own library on the Import from calibre sheet as soon as you pick the folder, for example “About 15 minutes for this library.”</p>\n<h2 id=\"where-the-estimate-comes-from\">Where the estimate comes from</h2>\n<p>The sheet allows a twentieth of a second for each book, assumes the files copy at 100 MB a second, and adds a second and a half for every file in a format other than EPUB, which gets an EPUB 3 copy. Comics aren’t converted, so they add no conversion time. For the thousand EPUBs that comes to 70 seconds, which the sheet rounds to “About 1 minute for this library.” The MOBI files cost 1,500 seconds more.</p>\n<p>The line sits at the bottom of the sheet, under “What comes over”. Until you pick a folder it reads “The time shows after you choose the library.” A small library gets “Less than a minute for this library.”</p>\n<h2 id=\"pdfs-convert-faster-than-that\">PDFs convert faster than that</h2>\n<p>A second and a half is generous for a PDF. On the 50 PDFs of Readwyn’s test corpus, converted one at a time on one Mac, the release build took less than a tenth of a second for 49 of them. The slowest, a Portuguese dictionary of 2,164 pages, took 0.45 seconds.</p>\n<h2 id=\"how-the-time-left-updates\">How the time left updates</h2>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-import-progress.webp?h=3b799b037de2cb2e9ef7\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-import-progress.webp?h=e75a05a85d47b44336b1\" alt=\"The library during an import from calibre: Importing 412 of 1,206 books, the current book Middlemarch being converted from MOBI to EPUB 3, a progress bar, and Stop import; the sidebar shows 34 percent and about 10 minutes left\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The time left, in the card at the foot of the sidebar</figcaption>\n</figure>\n<p>The sidebar card counts down (“About 10 min left”), starting from the estimate for the books still to come. As books arrive, Readwyn compares how long they really took with what the estimate allowed them, and scales the rest by the same amount. So on a fast Mac the countdown runs out sooner than the sheet promised, and on a slow one it runs longer.</p>\n<h2 id=\"stopping-an-import\">Stopping an import</h2>\n<p>“Stop import” stops after the current book. The books that already arrived stay, and the message says how many: “Import stopped. 412 of 1,206 books are in Calibre Library.” A stopped import doesn’t pick up where it left off, and since the library now has books, it can’t take another import. Nor can you compact the library while an import runs, because a library runs one job at a time.</p>\n<h2 id=\"importing-into-another-mac\">Importing into another Mac</h2>\n<p>If the app shows the libraries of another Mac, the scan reads the calibre folder on the Mac in front of you. The import sends the whole folder across first, and then the other Mac imports it. The estimate doesn’t include that copy across the network. The other Mac must have given this one Admin access.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-long-to-read-a-book/","url":"https://readwyn.com/blog/how-long-to-read-a-book/","title":"How long will this book take to read?","summary":"The time left in a book and in a chapter comes from the words you haven't read and a reading speed measured as you turn pages.","content_html":"<p>The reading well on the book screen shows how long the rest of the book will take you: “about 27 h” left, for example. Readwyn works this out from the words you haven't read yet and your own reading speed, which it measures in words per minute as you read in its reader. Inside the reader, the footer also shows the time left in the chapter, such as “18 min left in chapter”.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader.webp?h=8f10e162ef8d8cef7c82\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader.webp?h=9141239fb26850877292\" alt=\"Middlemarch in the reader: two pages side by side, with page numbers at the bottom and 18 minutes left in the chapter at the right of the footer\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The footer counts down to the end of the chapter</figcaption>\n</figure>\n<p>Next to the time left, the reading well shows the time you've already read (“7 h 40 min”). A finished book shows no time left. A book with no word count shows your Kobo's own estimate instead, if it has one. The chapter countdown disappears on the last spread of a book. Your speed itself is on the “Reading speed” card in Stats, for example “284” for the range you pick.</p>\n<h2 id=\"the-arithmetic\">The arithmetic</h2>\n<p>A 100,000-word book that you're 40% of the way through has 60,000 words left. At 300 words a minute, that's 200 minutes, which the book screen rounds to “about 3 h”. Under an hour, it gives minutes instead.</p>\n<p>For the chapter, Readwyn counts the words from your place to the end of the chapter, which is the next entry of the table of contents at any depth. In a book without a table of contents, a chapter ends where its chapter file ends. (Where the word count comes from is explained in <a href=\"/blog/count-words-in-ebook/\">How to count the words in an ebook</a>.)</p>\n<h2 id=\"measuring-your-speed\">Measuring your speed</h2>\n<p>Each time you turn to a new spread, Readwyn notes the words and the time since the last one. An interval of more than 5 minutes counts for nothing, since you were away from the book. Paging back adds no words.</p>\n<p>An interval only counts towards your speed if you moved forward through some text, at between 50 and 1,000 words a minute. Anything faster is a skim, a jump through the contents or a find; anything slower is a pause on one page. The intervals that pass are the samples your speed is worked out from.</p>\n<h2 id=\"reading-that-doesn-t-count\">Reading that doesn't count</h2>\n<p>“Read privately” (⌥⌘R) in the “…” menu stores no position, so it records no time and no samples. Comics count pages rather than words, at a speed of their own, and their samples stay out of your speed in words.</p>\n<p>Reading this Mac's copy of a book while the Mac that holds the library is out of reach records no time either. All other time in the reader, comics included, adds to the book's time read. Readwyn saves it when you close the book and every 10 minutes while you read, so a crash loses 10 minutes at most. A sync from a Kobo to the Mac sets the time read to the device's value.</p>\n<h2 id=\"a-new-library-starts-at-250-words-per-minute\">A new library starts at 250 words per minute</h2>\n<p>With no samples yet, Readwyn starts from 250 words per minute and counts that start as 10 minutes of reading. The speed is then <code>(W + 250 × 10) / (T + 10)</code>, where <code>W</code> is the words and <code>T</code> the minutes of every sample in the library.</p>\n<table><thead><tr><th>Your reading</th><th>Speed that Readwyn uses</th></tr></thead><tbody>\n<tr><td>No samples</td><td>250 words per minute</td></tr>\n<tr><td>10 minutes at 350 words per minute</td><td>300 words per minute</td></tr>\n<tr><td>10 hours at 350 words per minute</td><td>348 words per minute</td></tr>\n</tbody></table>\n<p>A short history nudges the estimate and a long one decides it, so one quick page at the start won't throw it off.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/how-readwyn-writes-to-kobo-safely/","url":"https://readwyn.com/blog/how-readwyn-writes-to-kobo-safely/","title":"How Readwyn writes to a Kobo safely","summary":"Before every write to a Kobo’s database, Readwyn backs it up to your Mac, writes one checked transaction and leaves no WAL file behind.","content_html":"<p>Readwyn writes to a Kobo’s database only when you apply a Readwyn → Device sync, and only over USB. Device → Readwyn only reads it. For a write, it backs the database up to your Mac first, puts every change into one transaction that it checks before and after the commit, and restores the backup if the last check fails.</p>\n<p>Plugging in a Kobo, counting the changes and building a <a href=\"/blog/kobo-sync-preview/\">preview</a> never write anything. The database is opened through a connection that can’t write, and nothing changes until you click the apply button of a review, for example “Apply 7 changes”.</p>\n<h2 id=\"the-rules-for-every-write\">The rules for every write</h2>\n<p>These rules cover everything Readwyn writes to the database: highlights, notes, reading values, saved words and collections.</p>\n<ol>\n<li>Before each write, Readwyn copies <code>KoboReader.sqlite</code> into a folder on your Mac named after the time. Nothing extra lands on the Kobo.</li>\n<li>Everything goes into one transaction. Before committing it, Readwyn reads the rows back and checks them.</li>\n<li>The Kobo’s database keeps recent changes in a separate <code>-wal</code> file, and the Kobo expects a clean database when it leaves USB mode. So after the commit, Readwyn folds that file into the database, leaves no <code>-wal</code> or <code>-shm</code> file behind, and checks the rows a second time through a fresh connection.</li>\n<li>If the Kobo has book files that aren’t in its database yet, it’s still adding books, and Readwyn writes nothing until it’s done. This happens straight after a send. Apply a Readwyn → Device sync then and Readwyn stops with “The Kobo is still adding books. Eject it, let it finish, and connect it again.” Device → Readwyn works as usual in the meantime.</li>\n<li>Readwyn reaches the database only through the mounted volume, and the Kobo leaves it alone while it’s in USB mode.</li>\n</ol>\n<p>The backups are in <code>~/Library/Application Support/readwyn/kobo-backups/</code>. Settings › E-readers has a “Backups” section with the note “Before every write, Readwyn saves a copy of the e-reader’s database on this Mac.” Its button, “Show backups in Finder”, opens the folder. After a write, the message at the foot of the Sync screen ends with “The device database is backed up.”</p>\n<h2 id=\"book-files-and-the-device-id\">Book files and the device ID</h2>\n<p>Sending a book doesn’t touch the database at all. “Send to device” writes only the book file, and the Kobo adds the book itself when it leaves USB mode. “Remove from device” deletes nothing but the book file, so it needs no backup either.</p>\n<p>The one other file Readwyn writes is <code>.readwyn/device-id</code>, which tells it which Kobo this is. It’s written at the first applied sync or send: first to a temporary file, then renamed into place and read back to check it.</p>\n<h2 id=\"ejecting-during-a-write\">Ejecting during a write</h2>\n<p>“Eject device” waits for a write to finish, for at most 30 seconds. Click it mid-write and an alert says “Readwyn is still writing to the Kobo”, with “Wait” and “Stop and eject”. “Wait” just closes the alert, while “Stop and eject” ejects as soon as the write is done. If the write ends first, the alert closes by itself.</p>\n<h2 id=\"tested-on-a-copy-of-a-real-kobo\">Tested on a copy of a real Kobo</h2>\n<p>The writes were checked in a virtual machine, on a copy of a Kobo Libra Colour’s storage. A highlight with a note written from the Mac matched the form of the Kobo’s own rows, and the table grew by exactly one row. A Readwyn → Device sync wrote 15 ratings. A My Words sync added two new words and didn’t duplicate one the Kobo already had in another case. Every run of SQLite’s <code>PRAGMA integrity_check</code> gave <code>ok</code>, and no <code>-wal</code> or <code>-shm</code> file was left on the Kobo afterwards.</p>\n<h2 id=\"when-your-library-is-on-another-mac\">When your library is on another Mac</h2>\n<p>If your libraries live on another Mac, the Kobo still works with the Mac it’s plugged into. That Mac copies the files the sync needs, the other Mac runs the sync on the copy, and the changed files come back. Before it writes them, the Mac with the Kobo backs up the database and checks that it hasn’t changed since the copy was taken.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/import-calibre-into-new-or-first-library/","url":"https://readwyn.com/blog/import-calibre-into-new-or-first-library/","title":"Import calibre into a new library or into your first library","summary":"Readwyn imports calibre into a new library named after the calibre folder, or into one that has never held a book, such as “My library”.","content_html":"<p>Readwyn imports calibre into a new library, which you name, or into a library that has never held a book, such as the empty “My library” you get at first launch. Under “Import into”, the “Import from calibre” sheet always offers “A new library”. There’s at most one other choice: the library you have open, if it has never held a book, or else another library that has never held one, if there is such a library. Everything else about the import is the same either way, and <a href=\"/blog/import-calibre-library/\">the calibre import walkthrough</a> goes through it screen by screen.</p>\n<figure class=\"plate mid\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-import.webp?h=a063f7fc58b1b4e4e384\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-import.webp?h=d0691a59955fe72517c9\" alt=\"The sheet Import from calibre: under Import into, A new library with the name Calibre Library, and My library, marked Empty\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>“Import into”: a new library, or the empty “My library”</figcaption>\n</figure>\n<h2 id=\"a-new-library\">A new library</h2>\n<p>The name field of “A new library” starts as “Calibre Library”. Once the scan has read your folder, it takes the folder’s name instead, and you can change it before you start. Leave the field blank and the library takes the folder’s name anyway.</p>\n<p>Two libraries can’t share a name, even in a different case: “Kids” and “kids” can’t both exist. Readwyn won’t start an import into a new library with such a name, and shows an error instead. You can rename the library later with “Rename…” in Settings › Libraries; its data folder stays where it is.</p>\n<p>Once every library has held a book, a new library is the only option. It’s also the one to pick if you want one Readwyn library for each calibre library; a second calibre library can come in later the same way.</p>\n<h2 id=\"your-first-library\">Your first library</h2>\n<p>A fresh install has one library, “My library”, with nothing in it. The library switcher says “No books yet” under its name, and the import sheet marks it “Empty”. Choose it and the calibre books go straight in, so you end up with a single library. If Readwyn is new to you and calibre holds all your books, that’s the simpler choice.</p>\n<h2 id=\"why-a-library-can-be-missing-from-the-list\">Why a library can be missing from the list</h2>\n<p>A library that has no books now isn’t necessarily one that never held a book. Add a book to a library and remove it, and the import no longer offers that library, because the removed book still counts. Compacting doesn’t reset that either. The switcher then says “No books” for that library, not “No books yet”, and the import still won’t take it.</p>\n<h2 id=\"while-it-runs\">While it runs</h2>\n<p>Books appear in the chosen library as they arrive, and the switcher shows “Importing · 412 of 1,206” under the library’s name. Each library runs one job at a time, so you can’t compact it while an import is running.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/import-calibre-library/","url":"https://readwyn.com/blog/import-calibre-library/","title":"How to import your calibre library into Readwyn","summary":"Import a calibre library into Readwyn on a Mac: open the sheet, pick the folder with metadata.db, check the scan, and keep reading while the books arrive.","content_html":"<p>Open the Import from calibre sheet, pick the folder that has <code>metadata.db</code> in it, check what the scan found, and start the import. Readwyn copies the books in the background, one at a time, and you can keep reading while they arrive. It’s a one-time copy: afterwards Readwyn never reads calibre’s <code>metadata.db</code> again, so you can keep calibre installed or remove it.</p>\n<h2 id=\"open-the-sheet\">Open the sheet</h2>\n<p>An empty library shows a welcome page with “Import from calibre”. Otherwise, use “Import from calibre…” in the library switcher at the top of the sidebar, or in Settings › Libraries.</p>\n<h2 id=\"pick-the-calibre-folder\">Pick the calibre folder</h2>\n<p>The sheet asks for the folder first. Click “Choose calibre library…” and Finder opens where calibre keeps its library by default. Select the folder with <code>metadata.db</code> in it. Readwyn on the Mac runs in the App Sandbox, so it can open only a folder you pick (<a href=\"/blog/import-calibre-mac-app-store/\">more on that</a>).</p>\n<p>The scan starts straight away and shows what it found, for example “Found 1,206 books, 3.4 GB”. An estimate of <a href=\"/blog/how-long-calibre-import-takes/\">how long the import will take</a> sits at the foot of the sheet, and the import button now counts the books (“Import 1,206 books”). Pick a folder without <code>metadata.db</code> and the sheet says so in red under the folder: “… is not a calibre library (it has no metadata.db)”. The button stays at “Import” and does nothing, and no library gets created.</p>\n<figure class=\"plate mid\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-import.webp?h=a063f7fc58b1b4e4e384\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-import.webp?h=d0691a59955fe72517c9\" alt=\"The Import from calibre sheet, with the number of books, their size, and the time the import takes\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The scan: the books found and the time the import takes</figcaption>\n</figure>\n<h2 id=\"choose-where-the-books-go\">Choose where the books go</h2>\n<p>“Import into” offers a new library, named after the calibre folder (you can change the name), or a library that has never had a book, marked “Empty”. On a fresh install that’s “My library”. If you can’t decide, <a href=\"/blog/import-calibre-into-new-or-first-library/\">another post compares the two</a>.</p>\n<h2 id=\"reading-while-it-runs\">Reading while it runs</h2>\n<p>Once the import starts, you can read, sync your Kobo or close the window. Each book turns up in the library as soon as it’s stored.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-import-progress.webp?h=3b799b037de2cb2e9ef7\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-import-progress.webp?h=e75a05a85d47b44336b1\" alt=\"The library during an import from calibre, with the books done, the current book, and a progress bar\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Books arrive while the import runs</figcaption>\n</figure>\n<p>The library you’re importing into shows a notice with “Importing 412 of 1,206 books”, the book it’s on (“now: Middlemarch, converting the MOBI to EPUB 3”), a progress bar and “Stop import”. Converting is the step where a MOBI, AZW3, PDF or other older file gets its EPUB 3 copy, and the original stays beside it (<a href=\"/blog/what-the-calibre-import-keeps/\">what else the import keeps</a>). In the sidebar, the library switcher reads “Importing · 412 of 1,206”, and the card at the bottom of the sidebar shows the percent and the time left. If you stop the import, the current book finishes first and the books already in stay put.</p>\n<h2 id=\"when-it-finishes\">When it finishes</h2>\n<p>A message tells you how it went: “Imported 1,206 books into Calibre Library.” If some books had problems, it counts those as well (“2 books had problems.”).</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/import-calibre-mac-app-store/","url":"https://readwyn.com/blog/import-calibre-mac-app-store/","title":"Import a calibre library in the Mac App Store version of Readwyn","summary":"The Mac App Store version of Readwyn reads only folders you pick, so the calibre import starts with choosing your calibre library in Finder.","content_html":"<p>The Mac App Store version runs in the macOS App Sandbox, which means Readwyn can only read a folder outside its own container after you’ve picked that folder in a Finder dialog. So the import starts with one extra step: you point it at your calibre library yourself.</p>\n<h2 id=\"picking-the-calibre-library\">Picking the calibre library</h2>\n<p>Open the “Import from calibre” sheet, from the welcome page of an empty library or from the library switcher at the top of the sidebar. Instead of a folder, it shows “Choose the folder of your calibre library.” with the note “Readwyn can open only a folder that you pick.” and a “Choose calibre library…” button.</p>\n<ol>\n<li>Click “Choose calibre library…”. Finder opens at the usual calibre library location in your home folder.</li>\n<li>Select the folder that has <code>metadata.db</code> in it, and click “Choose”.</li>\n<li>Readwyn scans it; while it does, the sheet says “Reading the library…”.</li>\n</ol>\n<p>Until a scan finds books, the main button just says “Import” and does nothing. Once it has, the sheet shows the folder with a “Choose…” button, what the scan found and how long the import should take, for example “Found 1,206 books, 3.4 GB” and “About 15 minutes for this library.” The button then counts the books: “Import 1,206 books”. Under “Import into” you choose where they go: a new library or one that has never held a book (<a href=\"/blog/import-calibre-into-new-or-first-library/\">which to pick</a>).</p>\n<figure class=\"plate mid\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-import.webp?h=a063f7fc58b1b4e4e384\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-import.webp?h=d0691a59955fe72517c9\" alt=\"The sheet Import from calibre after the scan: 1,206 books and 3.4 GB found in Calibre Library, about 15 minutes to import, then the choices under Import into\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>After the scan: what it found and how long the import should take</figcaption>\n</figure>\n<p>Pick a folder without <code>metadata.db</code> and nothing is created: the sheet goes back to its “Choose calibre library…” button, so you can try another folder.</p>\n<p>Once you start it, the import is the same as in any other version of Readwyn. It runs in the background, so you can keep reading or syncing, or close the window; <a href=\"/blog/import-calibre-library/\">the calibre import walkthrough</a> shows the progress and what “Stop import” does.</p>\n<h2 id=\"a-kobo-needs-the-same-permission\">A Kobo needs the same permission</h2>\n<p>The sandbox applies to a Kobo as well. The first time one connects, the Sync screen says “Give Readwyn access to Kobo Clara 2E” with a “Select Kobo Clara 2E…” button. Finder opens at the device, and “Allow” grants access; <a href=\"/blog/kobo-access-mac-app-store/\">Give Readwyn access to your Kobo in the Mac App Store version</a> has the details. If calibre used to fill that Kobo, finish the import before its first sync, so the books calibre sent there can match your library books by their calibre <code>uuid</code> (<a href=\"/blog/leave-calibre-keep-kobo-reading-history/\">more on moving a Kobo over from calibre</a>).</p>\n<h2 id=\"if-readwyn-server-is-on\">If Readwyn server is on</h2>\n<p>If you’ve turned on Readwyn server, the menu bar agent holds your libraries and runs the import, and that process needs to read your calibre folder too. When you pick the folder, Readwyn saves a security-scoped bookmark of it in <code>access.json</code> in its app folder. The agent is the same app in the same container, so it opens that bookmark, and you don’t have to pick the folder twice.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/install-readwyn-on-fedora/","url":"https://readwyn.com/blog/install-readwyn-on-fedora/","title":"Install Readwyn on Fedora from an rpm package","summary":"Readwyn's rpm installs on any supported Fedora release with dnf or a double-click (Discover on KDE Plasma), and a new version goes straight over the old one.","content_html":"<p>On Fedora with KDE Plasma, double-click the rpm and Discover installs it, along with whatever Readwyn requires that the system doesn't have yet: the libraries the app links against, and the DejaVu Sans Mono font, which the reader uses for code. On a Fedora 42 test machine with Plasma 6.6 on Wayland, Discover added 11 packages. Readwyn then shows up in the app menu with its icon. In a terminal, <code>dnf</code> does the same, and any Fedora release that is still supported will take the package.</p>\n<h2 id=\"which-rpm\">Which rpm</h2>\n<p>The <a href=\"/download/#linux\">download page</a> has one rpm for each processor:</p>\n<table><thead><tr><th>Processor</th><th>File</th></tr></thead><tbody>\n<tr><td>x86_64</td><td><code>readwyn-&lt;version&gt;-1.x86_64.rpm</code></td></tr>\n<tr><td>arm64</td><td><code>readwyn-&lt;version&gt;-1.aarch64.rpm</code></td></tr>\n</tbody></table>\n<p>Fedora won't install packages for a second architecture alongside its own, so an arm64 machine can't take the x86_64 rpm.</p>\n<h2 id=\"with-dnf\">With dnf</h2>\n<p>In the folder you downloaded to:</p>\n<pre><code>sudo dnf install ./readwyn-&lt;version&gt;-1.x86_64.rpm\n</code></pre>\n<p>On arm64 the file name ends in <code>.aarch64.rpm</code> instead. Two packages are only recommended: speech-dispatcher and the Mesa Vulkan drivers. Without speech-dispatcher the speak button says nothing, and the rest of the app works. The 3D library draws with the Mesa Vulkan drivers, and on a machine with no GPU they draw it on the processor, slowly.</p>\n<p>A new library starts empty: add books, or <a href=\"/blog/import-calibre-library/\">import a calibre library</a>. When you plug in a Kobo, Readwyn looks for it under <code>/run/media/&lt;user&gt;</code>, where Fedora mounts it. If you uninstall the rpm later, the program, the menu entry and the icons go, and your libraries in <code>~/.local/share/readwyn</code> stay as they are.</p>\n<h2 id=\"new-versions\">New versions</h2>\n<p>The app checks readwyn.com at most once a day and shows this card when a new version is out:</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/linux/dark-update-app.webp?h=284d320c96e1f3ca7a1b\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/linux/light-update-app.webp?h=f48b251e0f78b4d51a66\" alt=\"Readwyn's sidebar on Linux: above the Kobo Clara 2E card, a card says Readwyn 0.2.0 is out and You have 0.1.0, with the buttons Download and Later\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>When a new version is out, this card appears above the e-reader card</figcaption>\n</figure>\n<p>Install the new rpm straight over the old one, with a double-click or the same <code>dnf install</code> command. On Fedora 42, Discover put the 0.2.0 package over 0.1.0, and afterwards the list of libraries and the library were exactly as before. <a href=\"/blog/update-readwyn-on-linux/\">Updating without a repository</a> has the details of the check.</p>\n<h2 id=\"the-x86-64-rpm\">The x86_64 rpm</h2>\n<p>Discover was only tried with the arm64 rpm. The x86_64 one was checked on Fedora 42 with dnf alone: dnf found everything Readwyn needs in Fedora's x86_64 repositories, and Readwyn opened its library and a book.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/install-readwyn-on-ubuntu/","url":"https://readwyn.com/blog/install-readwyn-on-ubuntu/","title":"Install Readwyn on Ubuntu from a deb package","summary":"Double-click the Readwyn deb and App Center installs it on Ubuntu 22.04 or later. For updates, and for a first install on Debian 12, you need apt.","content_html":"<p>Download the deb for your processor from the <a href=\"/download/#linux\">download page</a> and double-click it in Files. App Center opens it, warns that the package doesn't come from App Center, and installs it once you go ahead. Readwyn then appears in the app menu with its icon. This works on Ubuntu 22.04 LTS and later: in tests on 24 October 2026, the arm64 and the x86_64 package both installed this way on 22.04 and 24.04 (the <a href=\"/blog/ebook-manager-arm64-linux/\">arm64 post</a> lists every system tested). Debian 12 is the exception: there you install it with <code>apt</code>.</p>\n<h2 id=\"which-file\">Which file</h2>\n<table><thead><tr><th>Processor</th><th>File</th></tr></thead><tbody>\n<tr><td>x86_64</td><td><code>readwyn_&lt;version&gt;-1_amd64.deb</code></td></tr>\n<tr><td>arm64</td><td><code>readwyn_&lt;version&gt;-1_arm64.deb</code></td></tr>\n</tbody></table>\n<p>The x86_64 build is the one with <code>amd64</code> in its name, Debian's word for that architecture. The commands on this page use it; on arm64, put the arm64 file name in its place.</p>\n<p>If you'd rather use a terminal than App Center, run this in the folder you downloaded to:</p>\n<pre><code>sudo apt install ./readwyn_&lt;version&gt;-1_amd64.deb\n</code></pre>\n<p>The first start opens the “Welcome to Readwyn” page, where you can add books or <a href=\"/blog/import-calibre-library/\">import a calibre library</a>.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/linux/dark-library.webp?h=02876295485e7080879c\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/linux/light-library.webp?h=24919ab2ffd9d07527ff\" alt=\"Readwyn's library window on Ubuntu with GNOME: the sidebar with Home library, Library, Notebook, Stats, Sync and a Kobo card, a grid of covers, and the search, Add books and the minimize, maximize and close buttons along the top\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The library window on Ubuntu with GNOME</figcaption>\n</figure>\n<h2 id=\"updating\">Updating</h2>\n<p>An <a href=\"/blog/update-readwyn-on-linux/\">update card</a> in the sidebar tells you when a new version is out. App Center won't put it over the one you have: it shows “Installed” and stops there. So download the new deb and run the same <code>apt</code> command with the new file name. It replaces the old version without touching your libraries or settings.</p>\n<h2 id=\"what-the-package-puts-on-your-system\">What the package puts on your system</h2>\n<p>The program itself is <code>/usr/bin/readwyn</code>, and the package adds a menu entry and an icon for it. Most of the download, though, is about 195 MB of pictures and models for the 3D library, in <code>/usr/share/readwyn/library3d/</code>.</p>\n<p>Along with the libraries the app links against, the package depends on DejaVu Sans Mono, which the reader uses for code. Two more packages are recommended rather than required: speech-dispatcher, which speaks a looked-up word aloud, and the Mesa Vulkan drivers, which the 3D library draws with. Without speech-dispatcher, the speak button says nothing and the rest of the app works.</p>\n<p>Removing the package takes away the program, its menu entry and its icons. It leaves <code>~/.local/share/readwyn</code> alone, so your libraries and settings are still there if you install again. The background server's autostart entry in <code>~/.config/autostart</code> also remains, if there is one; without the program, it does nothing.</p>\n<h2 id=\"debian-12\">Debian 12</h2>\n<p>On Debian 12 with GNOME, double-clicking a <code>.deb</code> opens Archive Manager, and GNOME Software won't open a local package at all, not even one of Debian's own. So use the <code>apt</code> command above from the start. Updates work the same way as on Ubuntu.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/keep-book-offline-on-laptop/","url":"https://readwyn.com/blog/keep-book-offline-on-laptop/","title":"Keep a book available offline on your laptop","summary":"How Available offline keeps a book and your place on a laptop paired with another Mac, and what happens to your progress and highlights away from home.","content_html":"<p>“Available offline”, in the “…” menu of the book screen, keeps a copy of a book on a laptop that's <a href=\"/blog/pair-second-mac/\">paired with another Mac</a>. The copy holds the book, its details, and your place. When the other Mac is out of reach (on a train, say), the book opens from that copy, and your changes wait until you're home.</p>\n<h2 id=\"keeping-a-book\">Keeping a book</h2>\n<p>You'll only find “Available offline” while the app shows another Mac's libraries. On the Mac that holds them, every book is already there. Once it's on, the menu item has a check and the size of the copy, for example “Kept on this Mac · 1.8 MB”.</p>\n<p>The copy lives on the laptop, not on the other Mac. Keeping one needs the <a href=\"/blog/read-guest-admin-access-levels/\">Read or Admin level</a>; a Guest Mac can't keep books.</p>\n<p>The copy doesn't go stale. If the book's EPUB changes on the other Mac, the laptop replaces its copy the next time it opens the book while the other Mac can be reached.</p>\n<h2 id=\"away-from-home\">Away from home</h2>\n<p>Without the other Mac, the library says “Can’t reach <name>.” with “Try again”. You still see every book, from the last lists the laptop saved. Kept books carry a download mark. The rest are dimmed, and clicking one shows a short message instead of the book.</p>\n<p>The status tabs, Filter, the sort, and the grid and list switch give way to two choices, “All” and “Available offline”, each with its count. Search still matches titles, authors, and series in the saved lists. Readwyn tries the other Mac every 30 seconds, and “Try again” tries straight away.</p>\n<p>If the connection drops while a book is open, you can keep reading it even if you never kept it: the reader holds its own copy of the open book.</p>\n<h2 id=\"what-waits-until-you-re-back\">What waits until you're back</h2>\n<p>Your progress, reading status (even for several books at once), highlights, and notes wait on the laptop, in order. The sidebar card “Waiting to send” counts them, for example “Progress in 1 book and 3 highlights go to Sid’s iMac when it is back.” Once the other Mac answers, the card changes to “Sending to Sid’s iMac” with a count such as “2 of 5 changes from offline”. A highlight you made on the train arrives exactly once, even if the app quits halfway through sending.</p>\n<p>Ratings, reviews, shelves, a Kobo sync, and imports can't wait. They need the other Mac, and fail with a network error while it's out of reach.</p>\n<p>Reading from a kept copy adds no <a href=\"/blog/reading-time-from-two-macs/\">time read</a>. Time is measured on the other Mac, and the pages you turned offline reach it later in one burst, so they move your place without adding minutes.</p>\n<h2 id=\"freeing-the-space\">Freeing the space</h2>\n<p>Settings › Other devices › “Offline books” shows how many books are kept and how much room they take across every library, for example “12 books kept on this Mac · 1.1 GB”. “Remove all…” asks first, then removes the copies. Changes that are still waiting aren't lost: they go to the other Mac as before. The library switcher also counts the kept books of each library, as in “412 books · 12 offline”.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kepub-or-epub-for-kobo/","url":"https://readwyn.com/blog/kepub-or-epub-for-kobo/","title":"KEPUB or EPUB: which format to put on your Kobo","summary":"Use KEPUB on a Kobo. Readwyn converts EPUBs to it by default without changing a word, and KEPUB highlights sync in the same form the Kobo writes itself.","content_html":"<p>Put KEPUB on your Kobo. It’s what Readwyn sends by default, and the option on the Sync screen sums it up as “Kobo’s own format: better page numbers and reading statistics. The text stays the same.” It matters most for highlights. The Kobo pins a highlight in a KEPUB to the <code>koboSpan</code> around its sentence, and the Kobo that Readwyn was tested against held only KEPUB books. Readwyn can sync highlights for a plain EPUB too, but no device has confirmed that form yet.</p>\n<h2 id=\"what-the-conversion-does-to-a-book\">What the conversion does to a book</h2>\n<p>Not a word of the text changes. On the way to the Kobo, Readwyn wraps every sentence of every chapter in <code>&lt;span class=\"koboSpan\" id=\"kobo.&lt;paragraph&gt;.&lt;sentence&gt;\"&gt;</code> and adds two wrapper elements around each chapter’s content. The copy on the Kobo is <code>Books/&lt;first author&gt; - &lt;title&gt;.kepub.epub</code>, and the EPUB in your library stays exactly as it was. So a highlight points to the same text in the KEPUB on the Kobo as in the EPUB in your library.</p>\n<h2 id=\"how-highlights-are-placed\">How highlights are placed</h2>\n<p>In a KEPUB, the Kobo records a highlight as a sentence span plus an offset into it. Readwyn reads that position when it brings highlights in, and checks that the text at the spot matches the highlight exactly. When it does, the highlight gets its place in your EPUB, and with it a page and the passage around it. Going the other way, Readwyn writes a library highlight as a span and an offset. A highlight written from the Mac into a copy of that Kobo’s database came out in exactly the same form as the rows the Kobo writes itself.</p>\n<p>A plain EPUB is different. There the Kobo stores the position as a <code>point(...)</code> path, the same kind of path that EPUB CFI uses. Readwyn writes that form and reads it back, and round-trip tests pass, but the format comes from a public report of the Kobo’s rows rather than from a device Readwyn has checked.</p>\n<h2 id=\"where-to-switch-it\">Where to switch it</h2>\n<p>There’s one setting for all your Kobos, and one for each Kobo:</p>\n<ul>\n<li>Settings › E-readers › “Sending books”: “Convert EPUB to KEPUB for Kobo e-readers”. It’s on by default and applies to every Kobo that has no setting of its own.</li>\n<li>The Sync screen › “Options for this device”: “Convert EPUB to KEPUB when sending”, for the Kobo that’s plugged in.</li>\n</ul>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-sync.webp?h=3748d9769c62e82b8755\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-sync.webp?h=25b1a7e75f31a1231622\" alt=\"The Sync screen of a Kobo Clara 2E. Under Options for this device, the switch Convert EPUB to KEPUB when sending is on\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The option of one device, on the Sync screen</figcaption>\n</figure>\n<p>With it off, Readwyn copies the EPUB across unchanged, as <code>Books/&lt;first author&gt; - &lt;title&gt;.epub</code> (<a href=\"/blog/send-books-to-kobo-from-mac/\">sending a book</a> covers the rest). The option has no effect on comics, which always go as CBZ, a format the Kobo opens itself.</p>\n<h2 id=\"epub-2-epub-3-and-books-already-on-the-kobo\">EPUB 2, EPUB 3, and books already on the Kobo</h2>\n<p>A book in your library can hold two EPUBs: an EPUB 2 original and the EPUB 3 that Readwyn converted from it. “Send to device” sends the EPUB 3, as a KEPUB if the option is on.</p>\n<p>A KEPUB that’s already on your Kobo may have been made from either. During a sync, Readwyn checks whether each KEPUB on the Kobo was made from an EPUB 2 or an EPUB 3, and maps highlight positions through the library file of that version. It never rewrites a KEPUB on the Kobo to make the positions fit.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-access-mac-app-store/","url":"https://readwyn.com/blog/kobo-access-mac-app-store/","title":"Give Readwyn access to your Kobo in the Mac App Store version","summary":"The Mac App Store version of Readwyn runs in the App Sandbox, so the first time you connect a Kobo, you select it in Finder and click Allow.","content_html":"<p>The first time you connect a Kobo, the Sync screen asks for access, using the name the Kobo shows in Finder, usually “Give Readwyn access to KOBOeReader”. Click “Select KOBOeReader…”, then “Allow” in the Finder window that opens at the Kobo. Readwyn remembers that Kobo, so you won’t be asked again.</p>\n<p>The step exists because the Mac App Store version runs in the App Sandbox, where macOS lets an app read a folder outside its own container only after you’ve selected it in a Finder dialog.</p>\n<h2 id=\"before-you-give-access\">Before you give access</h2>\n<p>Readwyn can tell that a device is plugged in, but it can’t look inside. The sidebar card says “Connected · needs access”, and the Sync screen shows the device with “Readwyn can see the model and your books after you give access.” That’s also why the notice uses the volume name rather than the model: the model is written in a file on the Kobo that Readwyn can’t open yet. In a test of the sandboxed build on a copy of a Kobo’s storage, until the Kobo was selected, macOS wouldn’t even let Readwyn read the Kobo’s version file.</p>\n<p>Next to the button is the hint “A Finder window opens at the device. Click Allow.”</p>\n<p>Once you click “Allow”, Readwyn reads the Kobo: its model, its books, and the changes in each direction. The two direction cards appear, and from then on syncing and sending books work as they do outside the sandbox, with the same <a href=\"/blog/how-readwyn-writes-to-kobo-safely/\">backup before every write</a>.</p>\n<h2 id=\"what-readwyn-keeps\">What Readwyn keeps</h2>\n<p>Selecting the Kobo gives Readwyn a security-scoped bookmark, which is how macOS lets a sandboxed app reopen a folder later. The bookmark lives in the app folder and is opened each time the Kobo is read. The background server runs as the same app, in the same container, with the same bookmarks, so you give access once, not once for the app and again for the server.</p>\n<p>A calibre library needs the <a href=\"/blog/import-calibre-mac-app-store/\">same one-off step</a>. In the sandbox, “Import from calibre…” opens a Finder dialog at the usual place of a calibre library in your home folder, and you select the folder that holds <code>metadata.db</code>.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-books-not-in-library/","url":"https://readwyn.com/blog/kobo-books-not-in-library/","title":"Find the books on your Kobo that are not in your library","summary":"Readwyn’s Sync screen lists the books on your Kobo that aren’t in your library, with format, size and how far you read, and a sync leaves them alone.","content_html":"<p>They’re on the Sync screen. Connect the Kobo, open “Sync”, and below the two direction cards a single line counts the Kobo books that your library doesn’t have, for example “3 books on Kobo Clara 2E are not in this library”. Click “Show” and you get the list: each book with its author, its format, its size, and how far you’ve read it on the Kobo.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-sync.webp?h=3748d9769c62e82b8755\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-sync.webp?h=25b1a7e75f31a1231622\" alt=\"The Sync screen of a Kobo Clara 2E. Under the direction cards, the line 3 books on Kobo Clara 2E are not in this library, with Show\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The books on the Kobo that the library does not have</figcaption>\n</figure>\n<p>Under the count, a second line adds: “Sync skips them. Readwyn never adds or removes books on the device.” (Readwyn puts a book on the Kobo only when you choose “Send to device”.)</p>\n<p>A book you’ve opened on the Kobo shows something like “KEPUB · 1.1 MB · 62% read on the device”. One you never opened says “EPUB · 2.0 MB · not opened”. That makes the list a quick way to find books that reached the Kobo some other way and never came into Readwyn.</p>\n<h2 id=\"how-a-book-counts-as-matched\">How a book counts as matched</h2>\n<p>Readwyn matches each book on the Kobo to the library. A book that Readwyn sent itself is easy, because the library recorded where it put it. A book that calibre sent is listed in <code>metadata.calibre</code>, the file calibre keeps on the Kobo, and Readwyn’s calibre import keeps each book’s calibre identity, so those match too. Any other book, including one whose calibre file is missing or damaged, is matched by its title, and by its first author too when several books in the library share that title. A book that none of these match is counted in that line.</p>\n<h2 id=\"what-a-sync-does-with-them\">What a sync does with them</h2>\n<p>Nothing. A sync covers only books that are on the Kobo and in the library. Highlights, notes and reading progress from the others don’t come into the library, and if one of them has reading data or highlights, the review lists it under “Left out” with the reason “Not in this library.” Words you saved in such a book stay on the Kobo, and they appear under “Left out” too.</p>\n<p>A sync never deletes these books or their highlights from the Kobo, even with the merge option off.</p>\n<p>The reverse case shows up in the review too: a book the library recorded on the Kobo that isn’t there any more, with “Not on the device. Send it from its book page first.” To put it back, <a href=\"/blog/send-books-to-kobo-from-mac/\">send it to the Kobo again</a>.</p>\n<p>The count covers only the library in the window. With <a href=\"/blog/more-than-one-ebook-library/\">several libraries</a>, switch at the top of the sidebar to check another.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-highlight-colours/","url":"https://readwyn.com/blog/kobo-highlight-colours/","title":"What Kobo highlight colours mean and how they sync","summary":"A Kobo stores each highlight colour as a number from 0 to 3, and Readwyn keeps the colour in both directions of a sync.","content_html":"<p>They mean whatever you decide. Readwyn attaches no meaning to a colour, so if yellow means a quote and green a word to look up, that system survives the trip in both directions, and the Notebook lets you filter by it. Readwyn’s Kobo research used a Kobo Libra Colour, which offers yellow, pink, blue, and green. The reader in Readwyn has the same four.</p>\n<h2 id=\"what-the-kobo-stores\">What the Kobo stores</h2>\n<p>The Kobo doesn’t store a colour name, only a number. On that Libra Colour, four highlights made in the order yellow, pink, blue, green were stored as 0, 1, 2, and 3:</p>\n<table><thead><tr><th>Number on the Kobo</th><th>Colour</th></tr></thead><tbody>\n<tr><td>0</td><td>Yellow, also the default</td></tr>\n<tr><td>1</td><td>Pink</td></tr>\n<tr><td>2</td><td>Blue</td></tr>\n<tr><td>3</td><td>Green</td></tr>\n</tbody></table>\n<p>Readwyn reads any number outside the table as yellow too.</p>\n<h2 id=\"how-a-colour-syncs\">How a colour syncs</h2>\n<p>A colour moves with its highlight. For each highlight on the sending side, the sync looks at the receiving side:</p>\n<ul>\n<li>If the highlight isn’t there, it’s added with its colour.</li>\n<li>If it’s there with the same note and colour, nothing happens.</li>\n<li>If it’s there with a different colour or note, the sender’s colour and note overwrite it.</li>\n</ul>\n<p>Change a highlight from yellow to blue on the Kobo, and the next Device → Readwyn lists the change; apply it and the highlight turns blue in your library. Recolour one in Readwyn and Readwyn → Device writes the new colour to the Kobo.</p>\n<p>The Kobo only updates a highlight’s date when its note or colour changes, and that date is what it shows in its list of annotations. Readwyn → Device gives each highlight on the Kobo the date it last changed in the library, so a recolour on the Mac also moves the date on the Kobo.</p>\n<h2 id=\"highlights-without-a-colour\">Highlights without a colour</h2>\n<ul>\n<li><strong>A highlight stored before Readwyn had colours</strong> has no colour in the library. The next Device → Readwyn lists the Kobo’s colour as an update to it.</li>\n<li><strong>A Kobo whose firmware stores no colour</strong> gives Readwyn no colour, so every highlight from it comes in as yellow. Readwyn writes to it without a colour, and the Kobo shows each highlight in yellow.</li>\n</ul>\n<h2 id=\"seeing-the-colours-on-the-mac\">Seeing the colours on the Mac</h2>\n<p>In the reader, selecting text opens a popover with the four colour dots and a ring around the current one. Click a different dot in a highlight’s own popover to recolour it.</p>\n<p>In the Notebook, each highlight has a dot in its colour and its text is marked in that colour. Chips at the top of a book, such as “All 5”, “Yellow 2”, and “Pink 1”, show one colour at a time, and only colours the book actually uses get a chip. The Highlights tab on the book screen has the same chips.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-notebook.webp?h=c5d907fb6a66e4603293\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-notebook.webp?h=758885220be4b17df8b7\" alt=\"The Notebook with the highlights of Middlemarch. The chips All 5, Yellow 2, Pink 1, Blue 1, and Green 1 are over the highlights, and each highlight has a dot in its colour\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>One chip for each colour of the book</figcaption>\n</figure>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-my-words-on-mac/","url":"https://readwyn.com/blog/kobo-my-words-on-mac/","title":"Bring the Kobo My Words list to your Mac","summary":"A Device → Readwyn sync copies the words you saved on your Kobo into My Words on your Mac, each with its book and date. The Notebook shows their dictionary entries.","content_html":"<p>Run a Device → Readwyn sync, and the words in the Kobo’s My Words list are copied into Readwyn’s My Words, each with the book you found it in and the date you saved it. On the Mac you can then read each one with its dictionary entry, or copy the whole list in one go.</p>\n<h2 id=\"what-the-kobo-saves\">What the Kobo saves</h2>\n<p>A Kobo doesn’t record every lookup. A word goes into My Words only when you save it. For each word it keeps exactly what you selected (capitals included, and a whole phrase if you selected one), the book, the dictionary you looked it up in, and the date. It doesn’t keep the sentence or where in the book the word was.</p>\n<p>Readwyn was tested on a copy of a Kobo Libra Colour that had 103 saved words from 7 books. The Device → Readwyn preview listed every one under its book, and after the sync My Words held all 103.</p>\n<p>Older Kobo firmware has no My Words list, so there’s nothing to bring across.</p>\n<h2 id=\"bringing-them-in\">Bringing them in</h2>\n<p>Connect the Kobo over USB and open “Sync”, then open the review on the “Device → Readwyn” card. Each word is a row under its book, marked “Word”. Every row starts ticked, so untick any word you don’t want and apply the rest.</p>\n<p>Only words from books that are both on the Kobo and in your library come in. A word saved in any other book stays on the Kobo, and the review names it under “Left out”.</p>\n<p>Case doesn’t split a word in two. “Perfunctory” on the Kobo and “perfunctory” in Readwyn count as one word, and if both sides have it, the earlier date wins. The review shows that as “Update · Word”, with both dates.</p>\n<h2 id=\"reading-them-in-the-notebook\">Reading them in the Notebook</h2>\n<p>Open “Notebook” in the sidebar and choose “My Words”. Words are grouped by headword, newest first, and each group shows the pronunciation, the first sense, and every word you saved under it with its book and date. Click a group for the full entry.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-notebook-words.webp?h=cd3bb9313f6f7a11a159\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-notebook-words.webp?h=b4ba5a2db81a6383cd4a\" alt=\"My Words in the Notebook: All books with 103 words at the left, and at the right the headword forlorn with its full entry and the saved words forlornness in Middlemarch and forlorn in Persuasion\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>My Words, grouped by headword, with the book of each word</figcaption>\n</figure>\n<p>Readwyn doesn’t store definitions. It looks each word up again when it shows it, in your dictionary order, so a newly downloaded dictionary changes what you see. A word that none of your dictionaries knows gets a group of its own that says “No entry in the dictionaries on this Mac.”</p>\n<p>The column on the left starts with “All books” and then lists each book that has words, with a count. Pick a book to see only its words, or type in “Search my words” to find a word or a headword.</p>\n<p>“Copy all” puts each headword, its first sense and the words you saved under it on the clipboard as text.</p>\n<p>My Words belongs to the app rather than to one library. Delete a book, or a whole library, and its words stay, just without a book.</p>\n<h2 id=\"the-other-direction\">The other direction</h2>\n<p>Words you <a href=\"/blog/save-looked-up-words/\">save in Readwyn’s own reader</a> go to the Kobo with Readwyn → Device, as long as their book is on the Kobo. If the Kobo already has the word in another case, no second entry is added.</p>\n<p>Words follow the review’s <a href=\"/blog/kobo-sync-preview/\">merge option</a>, “Merge highlights, notes and words instead of replacing them”, as highlights and notes do. With it on, a word that only the target has stays and shows as “Keep · Word”. With it off, the row says “Delete · Word” and the word goes.</p>\n<p>Readwyn files each word it writes to the Kobo under a dictionary, as the Kobo does with its own words. For an English book that’s the English dictionary. For a book in French, German or another language, it’s that language’s dictionary to English if the Kobo has one installed, and the English one if not.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-notes-in-your-library/","url":"https://readwyn.com/blog/kobo-notes-in-your-library/","title":"Get your Kobo notes into your ebook library","summary":"Readwyn keeps each Kobo note under its highlight, with the passage around it, and a note you edit on the Mac goes back to the Kobo at the next sync.","content_html":"<p>On a Kobo, a note is always attached to a highlight, and that’s how Readwyn keeps it too. A Device → Readwyn sync brings each note in with its highlight, and from then on it sits right under the highlighted text in the Notebook and on the book screen. Edit it on the Mac and the next Readwyn → Device sync puts your version back on the Kobo.</p>\n<p>In the sync’s review sheet, each new note is a row under its book with the highlighted text, so you can leave out any you don’t want (<a href=\"/blog/kobo-sync-preview/\">how the review works</a>). If the library already has that highlight with a different note, the Kobo’s note replaces yours. The exception is a note you changed on the Mac more recently than the Kobo’s: its row says so (“changed here later, on 9 Oct”) and starts unticked.</p>\n<p>Notes in a Kobo book that isn’t in your library have nowhere to go. The review lists the book under “Left out” with the reason “Not in this library.”, and the Sync screen <a href=\"/blog/kobo-books-not-in-library/\">counts those books</a> in a line under its direction cards.</p>\n<h2 id=\"reading-them-on-the-mac\">Reading them on the Mac</h2>\n<p>In the Notebook, and on the “Highlights” tab of the book screen, each note sits directly below its highlight. Notes aren’t counted on their own: “14 highlights · 3 notes” means three of those fourteen highlights carry a note.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-notebook.webp?h=c5d907fb6a66e4603293\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-notebook.webp?h=758885220be4b17df8b7\" alt=\"The Notebook: the highlights of Middlemarch by chapter. The first highlight has the note The whole book in one sentence\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A note shows under its highlight</figcaption>\n</figure>\n<p>When the text at the Kobo’s position matches the highlight exactly, Readwyn also saves its page and the passage around it: the paragraph before, the highlight’s own paragraph, the paragraph after, and any footnotes they point to. That lets you read a note against the passage it was written about. “Show context” swaps the highlight and its note for a single “In context” card: the highlighted words are marked in their paragraph, your note sits right under them, the rest of the paragraphs follow, and the footnotes come last under a line. “Hide context” puts things back.</p>\n<p>To find a note again, type a few of its words into “Search highlights and notes” at the top of the Notebook. It looks in the notes as well as the highlighted text, and the list on the left shrinks to the books that match.</p>\n<h2 id=\"editing-a-note\">Editing a note</h2>\n<p>Hover over a highlight and you get “Copy”, “Add note” (or “Edit note”), “Show context”, “Show in book”, and a delete button. “Edit note”, or a click on the note itself, opens a small editor in place. Return saves and Shift+Return starts a new line. Clear the text and save, and the note is gone; “Cancel” leaves it as it was.</p>\n<p>The next Readwyn → Device sync lists the changed note as an update, and once you apply it, the Kobo has your new wording.</p>\n<h2 id=\"merge-or-replace\">Merge or replace</h2>\n<p>At the top of the review sheet is “Merge highlights, notes and words instead of replacing them”, on by default. With it on, a highlight that only the receiving side has stays, note and all. Notes on highlights that both sides have get no such protection. If the receiving side’s copy of a highlight has a note and the sender’s has none, the sync removes the note. Untick that row to keep it. Turn the option off and the receiving side ends up with exactly the sender’s highlights and notes for each synced book, and the review lists every deletion before you apply anything.</p>\n<p>Unplug the Kobo, or even delete the book from it, and the notes stay in your library. <a href=\"/blog/export-kobo-highlights-on-mac/\">Export in the Notebook</a> saves them to a Markdown or plain-text file, each with its highlight and page.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-sync-after-factory-reset/","url":"https://readwyn.com/blog/kobo-sync-after-factory-reset/","title":"What happens to your Kobo sync after a factory reset","summary":"After a factory reset, Readwyn sees your Kobo as a new device. Your highlights and reading data stay in the library, and the old entry can be forgotten.","content_html":"<p>Readwyn treats the reset Kobo as a new device. Nothing you synced before is lost: highlights and notes with their context and pages, progress, status, time read, ratings, shelves and the words in My Words all stay on your Mac. Settings › E-readers keeps the old entry until you forget it, and the reset Kobo gets a fresh identity at its next sync or send.</p>\n<h2 id=\"why-the-kobo-looks-new\">Why the Kobo looks new</h2>\n<p>Readwyn recognises a Kobo by a random ID in one small file, <code>.readwyn/device-id</code>, on the Kobo’s main storage. The first applied sync or send writes that file and registers the Kobo in the library. From then on the library knows which of its books are on that Kobo and where, which syncs you applied (they’re what “Recent syncs” shows), and the date of the first sync, shown on the Sync screen as “syncing since”.</p>\n<p>A factory reset wipes that file. Connecting the reset Kobo only reads it, so it has no ID until you apply a sync or send a book. The first sync or send after that writes a new ID and registers it as a new device, with today’s date for “syncing since”. Readwyn keeps each Kobo’s options on your Mac under its ID, so the reset Kobo starts with the defaults from Settings › E-readers. The library keeps its records for the old ID, and those are what the old entry in Settings › E-readers shows.</p>\n<h2 id=\"putting-books-and-highlights-back\">Putting books and highlights back</h2>\n<p>A sync never adds books to a Kobo. If the reset cleared yours, <a href=\"/blog/send-books-to-kobo-from-mac/\">send them again</a> with “Send to device” in the “…” menu of each book. Then eject the Kobo, let it add them, and connect it again: Readwyn <a href=\"/blog/how-readwyn-writes-to-kobo-safely/\">won’t write to its database</a> while it’s still adding books.</p>\n<p>If you run Device → Readwyn first, leave “Merge highlights, notes and words instead of replacing them” on, as it is by default (<a href=\"/blog/kobo-sync-preview/\">what the option does</a>). The reset Kobo has no highlights, so with the option off the review would list every highlight of each synced book for deletion. Books you haven’t opened on the reset Kobo have no reading values to give, so their progress in the library stays as it was.</p>\n<p>Readwyn → Device is the one that refills the Kobo: highlights, notes, progress, ratings, shelves and saved words. The review shows every change first. A highlight that has no place in the book on the Kobo isn’t written, and the review names it with the reason.</p>\n<h2 id=\"forgetting-the-old-entry\">Forgetting the old entry</h2>\n<p>Settings › E-readers lists every device that has synced with a library, even one that isn’t connected. After the first sync or send following the reset, you’ll see two entries, and both can have the same name. The new one shows “Connected” and “syncing since” today. The old one has no “Connected” and shows when it last synced. Choose “Forget…” on the old one, or “Rename…” first if you want to tell them apart.</p>\n<p>Forgetting removes the device from every library: its record, its book paths, its syncs and its options on this Mac. The books and the reading data that came from it stay.</p>\n<h2 id=\"if-you-haven-t-reset-yet\">If you haven’t reset yet</h2>\n<p>Run Device → Readwyn first, so the library has the Kobo’s highlights, notes, words and reading values before they’re wiped.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-sync-preview/","url":"https://readwyn.com/blog/kobo-sync-preview/","title":"Why a Kobo sync shows a preview before it changes anything","summary":"A Kobo sync in Readwyn goes one way, so it can move progress back or replace a note. The review sheet shows every change by book, and only ticked rows apply.","content_html":"<p>Because a Kobo sync in Readwyn goes one way. Wherever the Kobo and the library differ, the side you sync to (the target) takes the value of the side you sync from (the source), even when the source is behind, so a sync can move your progress back or replace a note you’d rather keep. Readwyn shows you every change first, book by book, and applies only the rows you leave ticked. Building that preview writes nothing.</p>\n<h2 id=\"one-direction-no-conflicts\">One direction, no conflicts</h2>\n<p>One side is always Readwyn: “Device → Readwyn” or “Readwyn → Device”. There’s no conflict screen, because the direction decides.</p>\n<p>The Sync screen counts each direction’s changes before you open anything, for example “8 changes in 3 books”, and each card has its own button, such as “Review 8 changes”. The line under “Choose what to sync” ends: “You see every change before anything is written.”</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-sync.webp?h=3748d9769c62e82b8755\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-sync.webp?h=25b1a7e75f31a1231622\" alt=\"The Sync screen: the card Device → Readwyn with 8 changes in 3 books and Review 8 changes, and the card Readwyn → Device with 5 changes in 4 books and Review 5 changes\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Each direction counts its changes before a review</figcaption>\n</figure>\n<h2 id=\"reading-the-review\">Reading the review</h2>\n<p>The sheet groups the changes by book. A highlight or note shows its text and whether it’s added, updated or deleted. A reading value shows the field and both sides, as in “Update · Progress” with “18% → 22%”. Ratings appear as stars, and shelves and saved words get rows of their own (“Add · Shelf”, “Update · Word”). Both directions are tabs at the top of the sheet, each with its count, so you can switch between them without going back.</p>\n<p>Almost every row starts ticked. Two kinds start unticked, and each says why. A row whose target changed more recently than the source shows the date, “changed here later, on 9 Oct”. A Readwyn → Device row that would move the Kobo back spells it out: “The Kobo is ahead. This moves it back to 18%. Run Device → Readwyn first to keep 22%.”</p>\n<p>Every row has a checkbox, and so does each book, with a count such as “2 changes · 1 ticked”. The foot of the sheet keeps the total (“7 of 8 changes ticked”), and the button follows it: “Apply 7 changes”.</p>\n<p>After the books comes “Left out”, the things the sync won’t do, each with its reason:</p>\n<ul>\n<li>a saved word whose book isn’t on both sides</li>\n<li>a book the library recorded on the Kobo that has since gone: “Not on the device. Send it from its book page first.”</li>\n<li>a book on the Kobo with reading data or highlights that the library doesn’t have: “Not in this library.”</li>\n<li>a highlight that has no place in the book on the Kobo</li>\n</ul>\n<h2 id=\"two-options-at-the-top\">Two options at the top</h2>\n<p>“Merge highlights, notes and words instead of replacing them” is on by default. With it on, anything only the target has stays, and the review lists it as kept, so the sync deletes nothing. Turn it off and each synced book’s highlights, notes and words on the target become exactly the source’s, with every deletion listed.</p>\n<p>In Device → Readwyn, “Keep finished books” stops the Kobo from changing the status, progress or time read of a book the library already has at 100%. Both options are set per Kobo, and you can change them for a single review.</p>\n<h2 id=\"after-you-apply\">After you apply</h2>\n<p>A Readwyn → Device sync <a href=\"/blog/how-readwyn-writes-to-kobo-safely/\">backs up the Kobo’s database</a> before it touches anything, then writes the ticked rows in one checked transaction. Device → Readwyn doesn’t write to that database, so it needs no backup. When either sync is done, a message at the foot of the screen gives the number of changes applied and, after a write to the Kobo, confirms the backup. That direction’s card then reads “Up to date” (“Applied 7 changes just now. 1 left unticked.”), and the other card counts its changes again. Each applied sync also goes into “Recent syncs” further down, with its time, its direction and what it left out.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/kobo-time-read/","url":"https://readwyn.com/blog/kobo-time-read/","title":"Bring your Kobo time read into your library","summary":"A Device → Readwyn sync copies the Kobo’s time read for each book into Readwyn, where Reading stats counts it under All time.","content_html":"<p>Your Kobo already counts how long you’ve spent in each book, and a Device → Readwyn sync copies that total into your library. It then shows on the book screen as time read, next to the progress, the page, and the time left. Readwyn → Device sends the library’s figure back to the Kobo. In Reading stats, the Kobo’s time only shows up under “All time”, because the Kobo doesn’t say which days you read.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book.webp?h=6c9b53f864b64d2872b9\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book.webp?h=1970bca0dae546ce847c\" alt=\"The book screen of Middlemarch: 22% read, about page 196 of 880, 7 h 40 min time read, about 27 h left, and a rating of four stars\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Time read on the book screen</figcaption>\n</figure>\n<p>Device → Readwyn lists every book whose time read differs, with both values, and you apply the ones you want. A book on the Kobo that your library doesn’t have is skipped.</p>\n<h2 id=\"two-sources-of-time-read\">Two sources of time read</h2>\n<p>Each session in Readwyn’s own reader adds its time to the book, stamped with the day and hour it started. A sync from the Kobo is different: it sets the book’s time read to the Kobo’s total, which has no dates attached.</p>\n<p>So after Device → Readwyn, a book’s time read is whatever the Kobo says. The sessions from Readwyn’s reader are still on record, though, and your reading speed is worked out from them, so it doesn’t change.</p>\n<h2 id=\"how-it-looks-in-reading-stats\">How it looks in Reading stats</h2>\n<p>“Stats” in the sidebar shows your reading for a week, a month, a year, or all time, as a bar for each day or month. Only time from Readwyn’s reader has a day attached, so that’s what the bars show, and the screen says so once: “Time from the Readwyn reader”.</p>\n<p>“All time” adds the Kobo’s undated time on top, and the line under the chart spells out the split, for example: “1,105 h from the Readwyn reader by month, and 64 h from the Kobo without a date. The bars show the time with a date.” For each book, the undated part is its total time read minus the time Readwyn’s reader logged on actual days.</p>\n<h2 id=\"when-the-two-sides-differ\">When the two sides differ</h2>\n<p>A total can go down. Like <a href=\"/blog/sync-reading-progress-kobo-and-mac/\">progress</a>, time read is copied as it is, in either direction. If you’ve read for an hour in Readwyn since the last sync, Device → Readwyn would set the library back to the Kobo’s smaller figure. The review sheet catches this. It compares the Kobo’s last-read date with the last change to the library’s figure: your last session in Readwyn’s reader, or the last-read date that an earlier sync brought from the Kobo. If the library’s change is newer, the row starts unticked.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/large-ebook-library-on-mac/","url":"https://readwyn.com/blog/large-ebook-library-on-mac/","title":"A Mac ebook library of 10,000 books: what the measurements show","summary":"How fast Readwyn stays with 10 to 10,000 books, from start-up and filters to scrolling, idle CPU and memory, and the two results that miss their target.","content_html":"<p>With 10,000 books, Readwyn took 222 ms from a cold start to the first frame of the library, in a virtual Mac. Every screen switch in the test (the library, a book and its tabs, the reader, Notebook, Sync and Settings) took 71 ms or less, and so did every filter, sort and status tab. Scrolling and page turns never needed more than 16.7 ms for a frame, and an idle screen drew no frames at all. Two results miss their target: switching to the 10,000-book library from another library (131 to 149 ms), and one find in a book in the 10-book library, which took 269 ms in one run of three.</p>\n<h2 id=\"how-it-s-measured\">How it's measured</h2>\n<p>There are four test libraries, of 10, 100, 1,000 and 10,000 books. They're made from copies of real books, each with its own title, authors, language, series, dates, and cover image. Each has reading data as well: statuses, shelves, ratings, progress, last-read dates, and highlights on some of the books.</p>\n<p>The measurements run the release build in a virtual Mac with 4 CPUs and 8 GB of memory, in a window of 1280 × 800 points at 60 Hz. Every timing starts when the app begins the switch, filter or sort, and stops at the end of the first frame that shows the result with its data. Covers decode after that frame and are not part of the time. A frame time counts the app's own work (drawing and presenting the frame), not the time spent waiting for the virtual GPU to hand back a free buffer to draw into. How busy the host Mac is changes the results inside the virtual one, so runs are only compared under the same load. The test libraries were not imported from calibre, and the <a href=\"/blog/import-calibre-library/\">import</a> itself isn't measured.</p>\n<h2 id=\"the-targets\">The targets</h2>\n<ul>\n<li>Every frame takes 16.7 ms or less while you scroll or turn pages.</li>\n<li>Every screen switch, filter, sort, and library switch takes 100 ms or less.</li>\n<li>Every find in the reader takes 100 ms or less.</li>\n<li>Idle CPU is below 1 % on every screen. When nothing changes, the app does not redraw.</li>\n<li>Peak memory with 10,000 books is at most 1.5 times the peak with 100 books.</li>\n</ul>\n<h2 id=\"the-results-with-10-000-books\">The results with 10,000 books</h2>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library.webp?h=b388e029cc1f91357982\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library.webp?h=1ae4e63639f580a708ea\" alt=\"The library in the grid view: the status tabs, the Filter and Sort controls, and rows of covers\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The grid view: with 10,000 books, its slowest frame while scrolling took 8.2 ms</figcaption>\n</figure>\n<table><thead><tr><th>Measurement</th><th style=\"text-align: right\">100 books</th><th style=\"text-align: right\">10,000 books</th></tr></thead><tbody>\n<tr><td>Cold start to the first library frame</td><td style=\"text-align: right\">162 ms</td><td style=\"text-align: right\">222 ms</td></tr>\n<tr><td>Open a book</td><td style=\"text-align: right\">29 ms</td><td style=\"text-align: right\">30 ms</td></tr>\n<tr><td>Open the reader</td><td style=\"text-align: right\">52 ms</td><td style=\"text-align: right\">67 ms</td></tr>\n<tr><td>Slowest filter</td><td style=\"text-align: right\">31 ms</td><td style=\"text-align: right\">67 ms</td></tr>\n<tr><td>Slowest sort</td><td style=\"text-align: right\">31 ms</td><td style=\"text-align: right\">62 ms</td></tr>\n<tr><td>Slowest frame while scrolling the grid</td><td style=\"text-align: right\">6.0 ms</td><td style=\"text-align: right\">8.2 ms</td></tr>\n<tr><td>Slowest frame while scrolling the list</td><td style=\"text-align: right\">4.7 ms</td><td style=\"text-align: right\">6.1 ms</td></tr>\n<tr><td>Library grid, resident memory</td><td style=\"text-align: right\">117 MB</td><td style=\"text-align: right\">146 MB</td></tr>\n<tr><td>Peak resident memory</td><td style=\"text-align: right\">291 MB</td><td style=\"text-align: right\">410 MB</td></tr>\n</tbody></table>\n<p>The screen switches are the slowest of three visits, and the filters and sorts the slowest of three runs.</p>\n<h3 id=\"scrolling-and-page-turns\">Scrolling and page turns</h3>\n<p>To test scrolling, the grid and the list each got one wheel event per frame for 5 seconds. No frame took more than 16.7 ms, in any library. The reader turned 12 pages with each animation (slide, cover, fade, curl, and flip). With 10,000 books, no frame took more than 12 ms.</p>\n<h3 id=\"idle\">Idle</h3>\n<p>After a screen settles, Readwyn draws nothing. On every screen and in every library, an idle period of 5 seconds drew 0 frames, and the CPU stayed at 0.35 % of one core or less.</p>\n<h3 id=\"memory\">Memory</h3>\n<p>A hundred times the books doesn't mean a hundred times the memory: the peak with 10,000 books, 410 MB, is within the limit of 436 MB, which is 1.5 times the peak with 100 books. After going round every screen three times, the app used 1.6 % more memory with 10,000 books than after the first round. The target allows 5 %.</p>\n<h2 id=\"what-misses-its-target\">What misses its target</h2>\n<ul>\n<li><strong>A switch to the library of 10,000 books</strong> took 131 to 149 ms from the other libraries. Most of that is opening the library's store for the first time. Switching to the libraries of 10, 100, and 1,000 books took 26 to 58 ms.</li>\n<li><strong>One find in the reader</strong>, a Tamil word in the library of 10 books, took 269 ms. It was one slow run of three. The same find took 45 to 46 ms in the libraries of 100, 1,000, and 10,000 books.</li>\n</ul>\n<h2 id=\"what-keeps-it-quick\">What keeps it quick</h2>\n<p>Your data lives in an append-only log, and every read, search and sort goes through a SQLite cache built from it. The grid and the lists draw a small cover thumbnail rather than the full cover.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/leave-calibre-keep-kobo-reading-history/","url":"https://readwyn.com/blog/leave-calibre-keep-kobo-reading-history/","title":"Leave calibre and keep your Kobo reading history","summary":"Move from calibre to Readwyn with your Kobo: import the library first, then let the first Device → Readwyn sync bring progress, highlights, notes and words.","content_html":"<p>Import your calibre library into Readwyn first, then plug in the Kobo and run a Device → Readwyn sync. The import brings the Kobo data that calibre already holds. The sync then picks up what’s on the device today: progress, highlights, notes, status, ratings, time read, shelves and the words you saved.</p>\n<h2 id=\"1-import-the-calibre-library-first\">1. Import the calibre library first</h2>\n<p>Open “Import from calibre…” and pick the folder that holds <code>metadata.db</code>. If you haven’t run the import before, <a href=\"/blog/import-calibre-library/\">How to import your calibre library</a> covers the scan and the progress.</p>\n<p>The Kobo and Goodreads data that calibre keeps comes over, along with ratings, reviews and past reads, so finished books going back years show up in Reading stats. So does calibre’s <code>uuid</code> for each book, and that matters for the Kobo. calibre leaves a list of the books it sent in <code>metadata.calibre</code> on the device. Readwyn reads that file and matches each book on the Kobo to its library book by <code>uuid</code>; if the file is missing or damaged, it falls back to the title and first author.</p>\n<p>There’s a second reason to import first. A book on the Kobo that isn’t in the library doesn’t sync: its reading data and highlights stay on the device, and the review lists them under “Left out” with “Not in this library.”</p>\n<h2 id=\"2-connect-the-kobo\">2. Connect the Kobo</h2>\n<p>Plug it in over USB. Connecting only reads; nothing on the device changes until you apply a sync. By default Readwyn opens the Sync screen with the changes already counted.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-sync.webp?h=3748d9769c62e82b8755\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-sync.webp?h=25b1a7e75f31a1231622\" alt=\"The Sync screen: Kobo Clara 2E connected over USB, the cards Device → Readwyn with 8 changes in 3 books and Readwyn → Device with 5 changes in 4 books, and a line that 3 books on the device are not in this library\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Two directions, each with its changes counted</figcaption>\n</figure>\n<p>Under the two cards, a line counts the books on the device that aren’t in the library, for example “3 books on Kobo Clara 2E are not in this library”. “Show” lists them with their format, size and how far you got in each on the device.</p>\n<h2 id=\"3-run-device-readwyn-first\">3. Run Device → Readwyn first</h2>\n<p>Order matters because the target always takes the source’s value where the two differ. Run Readwyn → Device first and a book you’ve read further on the Kobo would be moved back there. The review catches that case and warns you: “The Kobo is ahead. This moves it back to 18%. Run Device → Readwyn first to keep 22%.”</p>\n<p>The “Device → Readwyn” card brings “progress, highlights, notes and status from the Kobo into the library.” Click “Review 8 changes” on it. The review sheet lists each change by book with both values, such as “Update · Progress” with “18% → 22%”, and every row has its own checkbox. Only ticked rows are applied. Two kinds start unticked: a row whose value on the target changed later, and, in Readwyn → Device, a row that would move the Kobo back. The button keeps count, for example “Apply 7 changes”.</p>\n<p>“Merge highlights, notes and words instead of replacing them” is on by default. With it on, the sync deletes nothing, so a highlight that only the library has stays put.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-sync-done.webp?h=9a3e0d60ef8718c47169\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-sync-done.webp?h=4d9602c3976b420b2d79\" alt=\"The Sync screen after Device → Readwyn: the card says Up to date and Applied 7 changes just now\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>After Device → Readwyn: “Up to date”</figcaption>\n</figure>\n<p>Afterwards the card says “Up to date” and what it applied. Each time-read row you applied sets the book’s time read to the figure on the device. That time comes without dates, so Reading stats add it to the All time total but can’t place it on a day; the line under the chart says how much of it is undated.</p>\n<p>Highlights keep their colours, progress keeps its last-read date and a Kobo collection comes in as a shelf of the same name. A sync never adds a book to the Kobo or deletes one from it; to put a book on the device, use “Send to device” on its book page.</p>\n<h2 id=\"tested-on-a-real-kobo-s-data\">Tested on a real Kobo’s data</h2>\n<p>On 11 October 2026 a test in a virtual Mac went through steps 1 and 3 with real data: a calibre library and a copied disk of the Kobo Libra Colour that got its books from it. The Kobo held 103 saved words in 7 books. The Device → Readwyn review listed all 103, each under its book, and left none out, because every one of those books had come over in the import. Opening the review wrote nothing; after the sync, My Words held all 103. The same test then wrote words the other way, Readwyn → Device, and the Kobo’s database still passed SQLite’s <code>PRAGMA integrity_check</code>, with no <code>-wal</code> or <code>-shm</code> file left behind.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/library-on-desktop-mac-read-on-laptop/","url":"https://readwyn.com/blog/library-on-desktop-mac-read-on-laptop/","title":"Keep your library on a desktop Mac and read on a laptop","summary":"Keep your Readwyn library on a desktop Mac that serves it even when the app is closed, and read from a laptop at home and on trips.","content_html":"<p>Your laptop can read straight from the desktop's libraries, and the desktop serves them even when Readwyn isn't open on it. All it needs is “Readwyn server” on the desktop and a pairing you do once. Before a trip, the laptop keeps copies of the books you pick, so they still open away from home.</p>\n<h2 id=\"on-the-desktop\">On the desktop</h2>\n<p>Open Settings › Other devices and turn on “Readwyn server”. The line under the switch spells out what changes: “Other devices can open the libraries of this Mac. It keeps running when you quit Readwyn, and starts at login.” macOS posts its own notice that Readwyn added a background item, and it may ask whether Readwyn can accept incoming connections.</p>\n<p>Then pick your home network in “Listen on”. Until you do, the server answers only the desktop itself. The port is 7313 unless you change it.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-settings-devices.webp?h=5e2bf654f800618406e3\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-settings-devices.webp?h=1263ffab1becc7426e6c\" alt=\"Settings, Other devices: Readwyn server on, the address of this Mac, Listen on Wi-Fi with port 7313, and the paired devices\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>The desktop Mac with Readwyn server on</figcaption>\n</figure>\n<p>With the setting on, an icon sits in the menu bar. Its menu shows “Readwyn server is running” with a green dot and a status line such as “2 devices connected · 1 asks to connect”, along with “Pair a device…”, “Connected devices…”, and “Turn off Readwyn server”. If you'd rather not have the icon, choose “Don’t show this icon”. You can bring it back in Settings › Other devices.</p>\n<h2 id=\"on-the-laptop\">On the laptop</h2>\n<p>In Settings › Other devices › Another device, type the desktop's address and choose Connect. You'll need the 6-digit code from “Pair a device…” on the desktop, or the laptop can <a href=\"/blog/pair-second-mac/\">send a request that you allow there</a>. macOS may ask whether Readwyn can use the local network; allow it.</p>\n<p>On the desktop's pairing sheet, leave “It can” at Read for a laptop you only read on, or pick Admin if you'll also add and edit books from it.</p>\n<p>From then on, the laptop opens with the desktop's libraries but keeps its own theme and reader settings, so it can sit in the dark scheme while the desktop stays light.</p>\n<h2 id=\"leaving-the-house\">Leaving the house</h2>\n<p>The laptop reaches the desktop only on your home network. Before you go, open the “…” menu on the book screen of each book you want and choose “Available offline”. The laptop keeps the book, its details, and your place.</p>\n<p>Away from home, the books you didn't keep are dimmed, and your progress, highlights, and notes <a href=\"/blog/keep-book-offline-on-laptop/\">wait on the laptop</a> until the desktop answers.</p>\n<h2 id=\"after-a-restart\">After a restart</h2>\n<p>The libraries come back when you log in on the desktop. Until then the laptop behaves as it does away from home, and it reconnects by itself once the desktop is back.</p>\n<h2 id=\"a-kobo-on-either-mac\">A Kobo on either Mac</h2>\n<p>A Kobo syncs with whichever Mac it's plugged into. Plug it into the laptop and the sync still runs against the libraries on the desktop. The laptop's level applies: Read and Admin can sync, Guest can't.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/look-up-words-in-ebooks-on-mac/","url":"https://readwyn.com/blog/look-up-words-in-ebooks-on-mac/","title":"Look up a word while you read an ebook on a Mac","summary":"Press and hold a word in the Readwyn reader to see its dictionary entry, with phrases such as “gave up”, base forms and a choice of two dictionaries.","content_html":"<p>Press and hold a word in the Readwyn reader. After 0.3 seconds the lookup popover opens over the page with the dictionary entry, while you’re still holding the mouse button. Right-clicking the word (or Control-clicking it) and choosing “Look up “…”” does the same.</p>\n<h2 id=\"holding-a-word\">Holding a word</h2>\n<p>Keep the pointer still while you hold. If it moves 4 px or more, the hold becomes a drag that selects text, as it does for a highlight. Holding on a margin, on white space or on an image does nothing. Comics are left out altogether: they open in a page reader that has no lookup.</p>\n<p>While the popover is open, the word is marked on the page. That mark isn’t a highlight, and nothing is saved. Escape or a click outside closes the popover, and that click does nothing else, so it won’t turn the page by accident.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-lookup.webp?h=100a081b3204e86878eb\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-lookup.webp?h=048cd0b33c717f3ab5f2\" alt=\"The lookup popover over a page of Middlemarch: the headword forlorn, its pronunciation, two senses with examples, phrases, derivatives, and In My Words\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Looked up “forlornness”: the entry of “forlorn”</figcaption>\n</figure>\n<h2 id=\"in-the-popover\">In the popover</h2>\n<p>The entry is laid out like a printed dictionary’s: headword and pronunciation, then each part of speech with numbered senses, examples and sub-senses, and at the end the derived words and the origin. Entries from the macOS Dictionary also list phrases, and a phrase’s entry shows the entry it belongs to.</p>\n<p>When the entry belongs to another form of the word, a line says so: hold “forlornness” and you get the entry for “forlorn” with “Looked up forlornness” under it.</p>\n<p>Next to the pronunciation is a speak button (<a href=\"/blog/hear-how-a-word-sounds/\">hearing how a word sounds</a>), and at the bottom of the popover is “Save to My Words” (<a href=\"/blog/save-looked-up-words/\">saving words to review later</a>). In <a href=\"/blog/read-a-book-privately/\">private reading</a> the lookup works just the same, minus that save button.</p>\n<h2 id=\"phrases-and-base-forms\">Phrases and base forms</h2>\n<p>Readwyn looks for the longest dictionary term that the word belongs to. If the word is part of a phrase the dictionary knows, the phrase comes first.</p>\n<p>So the right-click menu has one “Look up” item for each possible match, phrase first: “Look up “gave up””, then “Look up “gave””. When both the phrase and the single word have entries, the popover opens with a switch at the top to move between them.</p>\n<p>A phrase of two to four words that has no entry as written gets a second try, with each word in its base form. That’s how “gave up” finds “give up”.</p>\n<p>If nothing matches, the popover says “No entry in the dictionaries on this Mac.” and suggests “Turn on more dictionaries in Dictionary › Settings.”</p>\n<h2 id=\"which-dictionary-answers\">Which dictionary answers</h2>\n<p>A Mac has the macOS Dictionary and, if you <a href=\"/blog/offline-dictionary-wiktionary/\">download it</a>, Wiktionary; the entry comes from the first of them in <a href=\"/blog/dictionary-order/\">your order</a> that has the word. On Linux, Wiktionary is the only one.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/manage-ebooks-on-mac-without-calibre/","url":"https://readwyn.com/blog/manage-ebooks-on-mac-without-calibre/","title":"Manage ebooks on a Mac without calibre installed","summary":"After the one-time import, Readwyn doesn’t need calibre: how you add books, convert them, edit their details and send them to a Kobo with Readwyn alone.","content_html":"<p>You don’t need calibre on the Mac at all. Readwyn reads your calibre library once, during the import, and never reads or writes calibre’s <code>metadata.db</code> after that, so keeping calibre or deleting it is up to you. Adding books, converting them, editing their details and sending them to a Kobo all happen in Readwyn.</p>\n<h2 id=\"bringing-the-old-library-over\">Bringing the old library over</h2>\n<p>The import sheet says it plainly, right under its title: “Readwyn copies your calibre library once. calibre stays exactly as it is.” The books can go into <a href=\"/blog/import-calibre-into-new-or-first-library/\">a library of their own or into the empty one you start with</a>, and in the Mac App Store version you <a href=\"/blog/import-calibre-mac-app-store/\">pick the calibre folder yourself first</a>. <a href=\"/blog/what-the-calibre-import-keeps/\">What the calibre import keeps</a> lists everything that comes across.</p>\n<p>No calibre library? Start with “Add books” instead.</p>\n<h2 id=\"adding-books\">Adding books</h2>\n<p>Click “Add books” at the top of the library and pick the files, or drag them in from Finder. While the files hover over the window, an overlay names the number of books and the library, and marks any file it can’t take as “Skipped”. Afterwards a message says what came in, for example “Added 2 books to Home library. 1 file skipped.”, and each new book gets an accent ring.</p>\n<p>It takes EPUB, MOBI (also AZW3, combo files and PalmDOC), PDF, OEB 1.x packages (a ZIP with a package file), and the comic archives CBZ, CB7 and CBT. Titles come from the metadata in each file and keep their own script, Tamil for example.</p>\n<h2 id=\"converting-to-epub-3-and-keeping-the-original\">Converting to EPUB 3 and keeping the original</h2>\n<p>A book in an older format gets an EPUB 3 copy as it’s added. The conversion is Readwyn’s own, and the original file is never touched: the Details tab lists both, the original and the EPUB 3 “For reading” with the date it was converted. Books with DRM aren’t converted.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book-details.webp?h=3f5c56cb7a1e9927112f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book-details.webp?h=647f1b629b6b477420d1\" alt=\"The Details tab of Middlemarch: the original MOBI file, the EPUB 3 converted from the MOBI for reading and the KEPUB on a Kobo Clara 2E, each with Show in Finder or Remove from device\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The original, the EPUB 3 for reading and the copy on the Kobo</figcaption>\n</figure>\n<p>With a PDF, the conversion reflows the text into paragraphs, headings and images, and drops running headers, footers and page numbers. There’s more on this in <a href=\"/blog/convert-pdf-to-epub-without-calibre/\">Convert PDF to EPUB without calibre</a>.</p>\n<h2 id=\"editing-a-book-s-details\">Editing a book’s details</h2>\n<p>“Edit details” on the book screen, or ⌘E, opens a sheet with the cover and identifiers on the left and the fields on the right, from the title and authors to the published date, genres, tags and description. Nothing changes until you choose “Save”, and each changed field shows “Was” with the old value, plus “Undo”. To edit many books at once, select them with ⌘-click or Shift-click and choose “Edit details…” in the selection bar.</p>\n<p>In the sheet for one book, “Fetch metadata” asks Open Library and Google Books for the description, genres, page count, missing identifiers and cover. Each fetched field names its source, and the fetch never adds tags.</p>\n<h2 id=\"sorting-and-finding\">Sorting and finding</h2>\n<p>Books show as a grid of covers or as a list with status, progress, rating, last read date and date added. You can sort by date added, rating, last read, percent read or title; with “Last read”, the books you’re reading come first.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library-list.webp?h=8ef4e13c111a8ae0fdcd\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library-list.webp?h=0051b07c7f0ee42ce4d5\" alt=\"The library in the list view: Middlemarch, Tomorrow and Tomorrow and Tomorrow, and a Tamil title are reading, with their progress; other books follow with the date added\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The list view, sorted by last read</figcaption>\n</figure>\n<p>The Filter menu narrows by shelf, author, series, language and format, and search matches titles, authors and series as you type.</p>\n<h2 id=\"sending-books-to-a-kobo\">Sending books to a Kobo</h2>\n<p>Plug the Kobo in over USB and choose “Send to device” on the book page. The EPUB goes across as a KEPUB, because “Convert EPUB to KEPUB for Kobo e-readers” in Settings › E-readers is on by default. Reading data comes back through the Sync screen, and if calibre used to manage that Kobo, <a href=\"/blog/leave-calibre-keep-kobo-reading-history/\">run Device → Readwyn first</a> so the progress on the Kobo isn’t moved back.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/metadata-sources-for-each-field/","url":"https://readwyn.com/blog/metadata-sources-for-each-field/","title":"Where a book's metadata came from","summary":"How Readwyn shows where a book's details came from: the last fetch and its date, the calibre book ID, the converted file, the word count and estimated page counts.","content_html":"<p>Readwyn shows where a book's details came from on its Details tab: the source and date of the last metadata fetch, how the book came into the library, and the date of each file and of the word count. Values it works out for itself, such as an estimated page count or an automatic voice, say so in the edit details sheet.</p>\n<h2 id=\"the-rows-of-the-details-tab\">The rows of the Details tab</h2>\n<ul>\n<li>Each file shows its format, size and date. A converted EPUB 3 carries the date of its conversion, so it's easy to tell apart from the original.</li>\n<li>The word count comes with the date it was counted and “Count again”.</li>\n<li>“Details from” names the source of the last fetch and its date, for example “Open Library” and “fetched 3 Oct 2026”, with “Fetch again”. A book that has never been fetched says “Not fetched yet”, and its button reads “Fetch metadata” instead.</li>\n<li>“In the library” gives when the book arrived (“Since 12 Mar 2019”) and how: “added”, or “imported from calibre, book 1207” with the ID it had in calibre.</li>\n</ul>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book-details.webp?h=3f5c56cb7a1e9927112f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book-details.webp?h=647f1b629b6b477420d1\" alt=\"The Details tab of Middlemarch: the progress field, then the original MOBI file, the EPUB 3 converted from it on 3 Oct 2026, and the copy on a Kobo\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The original file, the converted EPUB 3, and the copy on the Kobo</figcaption>\n</figure>\n<h2 id=\"values-from-the-file-and-from-calibre\">Values from the file and from calibre</h2>\n<p>A book you add brings the metadata embedded in its file. A converted EPUB 3 carries the metadata that Readwyn read from the original. A book imported from calibre brings what the calibre library held, including its book ID, identifiers such as ISBN and Goodreads, the cover, the rating and the review. Anything you type in the <a href=\"/blog/edit-ebook-metadata-on-mac/\">edit details sheet</a> replaces the value that came from the file or from calibre.</p>\n<h2 id=\"fetched-values\">Fetched values</h2>\n<p>While the edit details sheet is open, each fetched value carries the name of its source (<a href=\"/blog/fetch-covers-and-descriptions/\">how a fetch works</a>). Those labels don't outlive the sheet. When you save, the book records the sources whose values you kept and the date, but not which field came from which source. If you undo every fetched value, or the fetch changed nothing, nothing is recorded, so “Details from” keeps the date of the last fetch that actually changed the book.</p>\n<h2 id=\"values-readwyn-works-out\">Values Readwyn works out</h2>\n<p>Without a page count, Readwyn divides the word count by 250 and rounds up. The field shows the result as “~880”, with “From the word count”. Typing a number replaces the estimate, and so does a fetched page count.</p>\n<p>The Voice field sets how a word you look up in the book is spoken. For an English book, Readwyn tries to pick a variety (UK, US, Australia, Canada or India) when the book is added; if nothing decides, the system's default voice speaks. The field shows “Automatic” and says how it decided: “from the setting”, “from the spelling”, “from the ISBN”, “from the publisher”, or “from the language tag”. Here the setting means where the story takes place, picked up from dense local words such as “arvo” or “lakh”. Pick a voice yourself and the field reads “Chosen by you”.</p>\n<p>The section in the 3D library is worked out too. The Details tab shows it as “Fiction › Literary fiction · automatic”, and the sheet as “Automatic (Literary fiction)”, until you choose a genre yourself. That automatic genre follows from the book's genres and tags, so a new genre or tag can move the book to another section.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/mobi-kf8-combo-files/","url":"https://readwyn.com/blog/mobi-kf8-combo-files/","title":"What is a MOBI and KF8 combo file?","summary":"A combo file packs a MOBI 6 and a KF8 copy of one book into a single file. Readwyn converts the KF8 part to EPUB 3 and falls back to the MOBI 6 part if it’s corrupt.","content_html":"<p>A combo file is one ebook file with the same book in it twice: once as a MOBI 6 book and once as a KF8 book, with a <code>BOUNDARY</code> record between the two. When you add one to Readwyn, the KF8 part is what gets converted to EPUB 3. If that part is corrupt, Readwyn converts the MOBI 6 part instead.</p>\n<p>A standalone MOBI 6 file has only the old book, and an AZW3 only the KF8 one. You don’t need to know which you have: Readwyn takes <code>.mobi</code>, <code>.azw</code>, <code>.azw3</code> and <code>.prc</code> files the same way and goes by the version field in the MOBI header, not the extension.</p>\n<h2 id=\"what-you-end-up-with\">What you end up with</h2>\n<p>The combo file stays as it was, and the reader opens the EPUB 3. That EPUB passes the same checks as any MOBI conversion (see <a href=\"/blog/convert-mobi-to-epub-on-mac/\">How to convert MOBI to EPUB on a Mac</a>): identical text, every link in place, and images and fonts byte for byte. If neither part converts, the book keeps just the combo file.</p>\n<h2 id=\"how-combo-files-are-tested\">How combo files are tested</h2>\n<p>Ten combo files, downloaded from public ebook sites and left exactly as found, are converted whole. Then each one is cut in two at its <code>BOUNDARY</code> record, and the MOBI 6 half and the KF8 half are converted separately: 20 more runs, and each half has to pass every check by itself.</p>\n<p>On top of that come round trips. Between them, the 10 source EPUBs have footnotes, images, tables, poetry and deep tables of contents, and one is in Tamil. Two other tools turn each into a combo file, and Readwyn converts it back. The text, links, table of contents and images have to match the source EPUB, allowing for text the writing tool changed and images it re-encoded.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/more-than-one-ebook-library/","url":"https://readwyn.com/blog/more-than-one-ebook-library/","title":"Keep more than one ebook library","summary":"Create, rename and switch between Readwyn libraries on a Mac: what each library keeps to itself, and what spans them all, like My Words and combined stats.","content_html":"<p>You can keep several libraries in Readwyn. “Create library…” in Settings › Libraries adds one, and the new library has its own data folder, with its own books and reading data. The switcher at the top of the sidebar moves you between them.</p>\n<h2 id=\"creating-and-naming-a-library\">Creating and naming a library</h2>\n<p>Every install starts with one library, “My library”, and no books. While it's empty it shows the “Welcome to Readwyn” page, with “Add books”, “Import from calibre” and “Use the libraries on another device”.</p>\n<p>A name needs at least one character that isn't a space. Names ignore case, so “Fiction” and “fiction” can't both exist: “Create” stays dimmed until the name is new. “Rename…” changes the name later and leaves the data folder where it is. Renaming a library to its own name in different case is fine.</p>\n<p>You can also let a <a href=\"/blog/import-calibre-into-new-or-first-library/\">calibre import make the library</a>. In the “Import from calibre” sheet, “Import into” offers “A new library”, named after the calibre folder.</p>\n<h2 id=\"switching\">Switching</h2>\n<p>Click the switcher. It lists every library with its book count; an empty one says “No books yet”, or “No books” if its books were removed and compacted. Switching clears any selection you had.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library.webp?h=b388e029cc1f91357982\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library.webp?h=1ae4e63639f580a708ea\" alt=\"The library window. At the top of the sidebar, the library switcher shows Home library with 412 books\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The library switcher names the current library and its books</figcaption>\n</figure>\n<p>Switching to a library of 10, 100 or 1,000 books shows it in 26 to 58 ms (release build, measured in a virtual Mac). The 10,000-book library is slower, at 131 to 149 ms, and misses the 100 ms target, mostly because its store opens for the first time.</p>\n<h2 id=\"housekeeping-in-settings-libraries\">Housekeeping in Settings › Libraries</h2>\n<p>Each row shows the library's book count, its size on disk, how much compaction would free (“140 MB to free” or “Nothing to free”) and when it was last compacted. If the data folder is on another disk, the row also names that disk. “Show in Finder” opens the data folder of any library on this Mac.</p>\n<p>“Compact” deletes a library's removed books for good. It runs in the background with a percentage (“Compacting · 40%”) and a “Stop” button. A library runs one job at a time, so an import and a compaction can't overlap.</p>\n<p>The “3D library” row sets how big each library's <a href=\"/blog/3d-library/\">3D library</a> is: “Automatic”, which grows with the number of books, or a larger room you choose yourself.</p>\n<h2 id=\"what-stays-separate-and-what-doesn-t\">What stays separate, and what doesn't</h2>\n<p>Books, files, covers, highlights, notes, shelves, statuses, ratings and reviews all belong to one library. A Kobo is recorded in each library that it syncs with. Its name in Settings › E-readers applies to every library, though, and “Forget…” removes it from all of them.</p>\n<p>A few things sit above the libraries. My Words belongs to the app, so the Notebook lists your saved words under “All books”, for example “32 words · every library”. Once you have a second library, Stats gets a scope: the current library or “All libraries” (<a href=\"/blog/reading-stats-across-libraries/\">how combined stats work</a>). The OPDS catalogue for reader apps lists the libraries at its root. And because all your libraries live on the Readwyn server of your Mac, another Mac paired with it sees every one of them in its switcher.</p>\n<p>Some settings belong to the Mac, not to any library: the theme, the reader settings, the dictionaries and the voices.</p>\n<p>To back up several libraries, back up the data folder of each one, plus My Words and the <code>owner-id</code> file once. The paths are in <a href=\"/blog/back-up-ebook-library/\">how to back up your ebook library</a>.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/move-large-calibre-library-keep-isbns/","url":"https://readwyn.com/blog/move-large-calibre-library-keep-isbns/","title":"Move a large calibre library without losing your ISBNs","summary":"Readwyn’s calibre import keeps every ISBN and other identifier, shows them in Edit details, gives reader apps the ISBN, and estimates 10,000 EPUBs at about 12 minutes.","content_html":"<p>Your ISBNs come across untouched. Readwyn copies every identifier calibre has for a book (the ISBN, the Goodreads ID and the rest), plus its calibre book ID. The Import from calibre sheet promises as much before you start: “Book IDs and identifiers” is the second item under “What comes over”.</p>\n<h2 id=\"where-the-isbns-show-up-afterwards\">Where the ISBNs show up afterwards</h2>\n<p>Open a book and choose “Edit details”, or press ⌘E. The identifiers sit on the left of the sheet, under the cover. calibre’s <code>uuid</code> is kept too, out of sight (<a href=\"/blog/what-the-calibre-import-keeps/\">what it’s for</a>).</p>\n<p>If a book is missing an identifier, “Fetch metadata” in the same sheet fills the gaps from Open Library and Google Books. The fetched values wait in the sheet, and nothing is stored until you save.</p>\n<p>Reader apps get the ISBN as well. If you turn on “Share with reader apps (OPDS)” in Settings › Other devices › Reader apps (Readwyn server has to be on too), each book that has an ISBN carries it in the catalogue as a <code>urn:isbn:</code> identifier, next to the title, authors, publisher and publication date.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-apps.webp?h=e0024127c108d4d70c9a\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-apps.webp?h=b3168403e87d03df89f1\" alt=\"Settings › Other devices › Reader apps: the switch Share with reader apps (OPDS) is on, with the address of the catalogue and the list of reader apps\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>The catalogue that carries the ISBNs to reader apps</figcaption>\n</figure>\n<h2 id=\"a-large-library\">A large library</h2>\n<p>The import adds one whole book at a time, identifiers included, so if you stop it partway, every book that made it across is complete.</p>\n<p>For 10,000 EPUBs of 2 MB each, the sheet estimates about 12 minutes; as MOBI files, the same books come to about 4 hours 20 minutes, because each MOBI gets an EPUB 3 copy (<a href=\"/blog/how-long-calibre-import-takes/\">how the sheet adds it up</a>). Once they’re in, a library of 10,000 books in Readwyn’s own tests drew its first frame 222 ms after a cold start (<a href=\"/blog/large-ebook-library-on-mac/\">the other timings</a>).</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/non-latin-titles-in-ebook-library/","url":"https://readwyn.com/blog/non-latin-titles-in-ebook-library/","title":"Keep Tamil and other non-Latin titles intact in your ebook library","summary":"Readwyn keeps Tamil and other non-Latin titles exactly as the book file has them or as you type them, and converts MOBI and AZW3 text to EPUB 3 unchanged.","content_html":"<p>A title in Tamil stays in Tamil. Readwyn stores the title and authors exactly as the book file gives them, in any script, and keeps a title you type in the edit details sheet exactly as you typed it. Converting a MOBI, an AZW3 or an EPUB 2 to EPUB 3 doesn't change a single character of its text, and the library shows a Tamil title in full wherever it fits.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library-list.webp?h=8ef4e13c111a8ae0fdcd\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library-list.webp?h=0051b07c7f0ee42ce4d5\" alt=\"The library in the list view: among English titles, the Tamil title பொன்னியின் செல்வன் with its subtitle, cut with an ellipsis, by Kalki Krishnamurthy, with its status and progress\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A Tamil title in the list view</figcaption>\n</figure>\n<h2 id=\"editing-the-details-yourself\">Editing the details yourself</h2>\n<p>“Edit details” on the book screen opens the edit details sheet. Nothing is stored until you save. Until then, each changed field shows “Was” with the old value struck through, so you can check the old title against the new one.</p>\n<p>The Language field sets the book's main language. Only English books are hyphenated in the reader, so in a book set to Tamil the lines break between words.</p>\n<h2 id=\"long-titles\">Long titles</h2>\n<p>In the list view, a Tamil title that fits is shown whole, and a longer one ends in “…”. Readwyn measures the line as it's actually drawn, and it cuts only between whole letters, so a vowel sign never comes away from its consonant.</p>\n<p>When a book has no cover image, Readwyn draws a plain cover with the title on it, set at 17 px. If one word is too wide for the cover, the whole title gets smaller, down to 11 px, until that word fits. A Tamil title then wraps at a space rather than inside a word, unless one word is too long even at 11 px.</p>\n<h2 id=\"conversion-keeps-every-character\">Conversion keeps every character</h2>\n<p>The <a href=\"/blog/convert-old-ebook-formats-keep-original/\">conversion to EPUB 3</a> is tested on a Tamil book, one of 10 in the round-trip tests. It comes from archive.org, because Project Gutenberg has no Tamil books. Other tools write each book out as MOBI 6, as AZW3 and as a combo file, and Readwyn converts those back to EPUB 3. The text has to match the source, apart from a few known changes that the writing tools make, such as an added contents page. One of them touches Tamil: a tool stores some vowel signs in composed form, so ெ followed by ௗ comes back as ௌ.</p>\n<p>Highlights in Tamil books handle both forms too. When Readwyn has to find a highlight again by its text, it looks for the text as the book has it and in composed form.</p>\n<h2 id=\"search-and-word-counts-in-other-scripts\">Search and word counts in other scripts</h2>\n<p>Search ignores case in any script that has it. It compares lowercase forms, so Greek “λόγος” finds “ΛΌΓΟΣ”, and the final sigma “ς” counts as “σ”, in the library search and in Find in the book (⌘F in the reader) alike. In Find in the book, accents have to match: “ville” won't find “villé”.</p>\n<p>A Tamil book is counted at its spaces, like an English one, and in <a href=\"/blog/count-words-in-ebook/\">Chinese or Japanese text</a> each ideograph counts as one word.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/oeb-package/","url":"https://readwyn.com/blog/oeb-package/","title":"What is an OEB package, and how to open one","summary":"An OEB 1.x package is a book in Open eBook, the format before EPUB. Add it to Readwyn as a ZIP file and you get an EPUB 3 of it, with every file path kept and the ZIP left as it was.","content_html":"<p>An OEB package is a book in the Open eBook format, which came before EPUB: the chapters and images of the book, plus a package file that lists them. To open one, add it to Readwyn as a ZIP file (ending in <code>.zip</code> or <code>.oeb</code>) with its package file inside. Readwyn makes an EPUB 3 from it straight away and leaves the ZIP untouched; “Read” opens the EPUB.</p>\n<h2 id=\"versions-and-the-table-of-contents\">Versions and the table of contents</h2>\n<p>Three versions are supported: OEB 1.0, 1.0.1 and 1.2. The version comes from the DOCTYPE of the package file, for example <code>DTD OEB 1.0.1 Package</code>.</p>\n<p>An OEB package has no NCX. Readwyn builds the table of contents (both the EPUB 3 navigation and an NCX) from the contents page that the package’s guide points to.</p>\n<h2 id=\"file-paths-and-highlights\">File paths and highlights</h2>\n<p>Every file keeps its path inside the book, and not a character of the text changes. So a highlight you make in the EPUB 3 holds in the OEB package too: it points to the same file and the same words. EPUB 2 books convert the same way, file paths and all (<a href=\"/blog/epub-2-or-epub-3/\">EPUB 2 or EPUB 3: what changes when your books are converted</a>). A package that won’t convert stays in the library as the ZIP alone.</p>\n<h2 id=\"how-well-it-s-tested\">How well it’s tested</h2>\n<p>The aim was five real public-domain books for each OEB version. The search turned up one, an OEB 1.0.1 book, and it passes every check. The converter handles 1.0 and 1.2 as well, but no real book in either version was found to prove it.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/offline-dictionary-wiktionary/","url":"https://readwyn.com/blog/offline-dictionary-wiktionary/","title":"An offline dictionary for your ebooks with Wiktionary","summary":"Download English Wiktionary once in Readwyn and look words up in your ebooks with no network: where to start the download, how it resumes, and what it costs on disk.","content_html":"<p>Wiktionary can live on your Mac, so the reader can look words up without a network. The Download button is in the Wiktionary row of Settings › Reading › Look up. Once the row says “Ready”, a word that you <a href=\"/blog/look-up-words-in-ebooks-on-mac/\">hold in the reader</a> can bring up its Wiktionary entry.</p>\n<h2 id=\"what-you-download\">What you download</h2>\n<p>You get the English entries of English Wiktionary, in the machine-readable extract that Wiktextract makes for kaikki.org. Readwyn downloads that extract and builds its own index from it on your Mac. Before you download, the row in Settings says “About 525 MB for offline use”: that’s the download. Readwyn’s estimate for the index is about 900 MB, but no one has measured an index built from the real extract yet. Once it’s built, the row shows its actual size on this Mac.</p>\n<p>In October 2026 the extract was 523 MB. It holds only the English entries, and kaikki.org plans to retire it. Once it’s gone, Readwyn falls back to the extract of every language (about 3 GB) and keeps only the English lines.</p>\n<p>The index keeps only what the popover shows: the word, its part of speech, the first pronunciation, the senses with one example each, up to 20 derived terms, and the origin. Translations and audio are left out.</p>\n<p>Looking a word up reads only a few pages of the index, and the file is never loaded into memory. With an index of 1,000,000 synthetic entries, a release build of Readwyn finds a word in 50 to 120 µs.</p>\n<h2 id=\"what-the-row-shows\">What the row shows</h2>\n<p>The Wiktionary row changes as the download goes on:</p>\n<ol>\n<li>Before the download: the download size and Download.</li>\n<li>During the download: the progress, the time left, and Stop.</li>\n<li>Once the file is complete: a progress bar marked “Getting it ready” while Readwyn builds the index. After that it deletes the downloaded file.</li>\n<li>When it’s ready: the size on this Mac, the install date, “Ready”, and “Remove…”.</li>\n</ol>\n<h2 id=\"stopping-and-resuming\">Stopping and resuming</h2>\n<p>Stop keeps the bytes your Mac already has, and so does an error or quitting the app. After an error, the row gives the reason, how much this Mac kept, and “Try again”. Both Download and “Try again” ask the server only for the rest of the file. If the server sends the whole file anyway, the download starts over.</p>\n<p>“Remove…” asks before it does anything, then deletes the index and any kept bytes.</p>\n<h2 id=\"on-linux\">On Linux</h2>\n<p>Linux has no system dictionary, so Wiktionary is the only one Readwyn has there. Until you download it, a lookup in the reader shows the word, “No dictionary is ready”, and this offer:</p>\n<blockquote>\n<p>Linux has no dictionary of its own. Wiktionary works offline, in English: about 900 MB, once.</p>\n</blockquote>\n<p>“Download Wiktionary” starts the same download as Settings, with its progress in the popover. When the dictionary is ready, the popover looks the word up again and shows the entry. Words you saved to My Words before the download get their entries at that point too.</p>\n<h2 id=\"next-to-the-macos-dictionary\">Next to the macOS Dictionary</h2>\n<p>On a Mac, Wiktionary sits beside the macOS Dictionary, and the lookup asks them in <a href=\"/blog/dictionary-order/\">the order you set</a>.</p>\n<p>A Wiktionary entry has no list of phrases and no parent entry, because Wiktionary gives each phrase an entry of its own. A word whose senses all point to one other word opens that word’s entry instead: “gave” brings up “give”, and “colour” brings up “color”.</p>\n<h2 id=\"the-licence\">The licence</h2>\n<p>Wiktionary’s text is under the Creative Commons Attribution-ShareAlike licence, CC BY-SA 4.0. While Wiktionary is ready, its credit appears under the dictionary list in Settings and at the bottom of the popover for every Wiktionary entry: “Definitions from Wiktionary, by its contributors, under CC BY-SA 4.0. Extracted by Wiktextract (kaikki.org). Readwyn shortens and arranges the entries.” The index stays on your Mac, and Readwyn doesn’t distribute it.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/opds-catalogue-for-koreader/","url":"https://readwyn.com/blog/opds-catalogue-for-koreader/","title":"Set up an OPDS catalogue for KOReader","summary":"Add your Readwyn library to KOReader as an OPDS catalogue: the settings on the Mac, the password, and what KOReader did in Readwyn’s own test.","content_html":"<p>KOReader can browse the libraries on your Mac and download books from them, because Readwyn serves those libraries as an OPDS catalogue. You switch the catalogue on in Readwyn, add KOReader as a reader app, and type the address, user name, and password that Readwyn gives you into KOReader. Nothing from Readwyn has to be installed on the e-reader.</p>\n<h2 id=\"turn-on-the-catalogue\">Turn on the catalogue</h2>\n<p>The catalogue is part of Readwyn server, so it runs on the Mac that keeps your libraries. In Settings › Other devices:</p>\n<ol>\n<li>Turn on “Readwyn server”.</li>\n<li>In “Listen on”, pick the network your e-reader uses, for example “Wi-Fi · 192.168.1.20”.</li>\n<li>Under “Reader apps”, turn on “Share with reader apps (OPDS)”. It stays off until you do.</li>\n</ol>\n<p>If Listen on is still “This Mac only”, the Reader apps section says “Only apps on this Mac can reach the catalogue. To reach it from a phone or an e-reader, choose a network in Listen on.” Once you’ve picked a network, the address below the switch uses it, for example <code>http://192.168.1.20:7314/opds</code>, with a Copy button.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-apps.webp?h=e0024127c108d4d70c9a\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-apps.webp?h=b3168403e87d03df89f1\" alt=\"Settings › Other devices › Reader apps: the switch Share with reader apps (OPDS) is on, the address http://192.168.1.20:7314/opds with Copy, a warning about the Wi-Fi, and three reader apps, the first named Kindle Scribe KOReader\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>Settings › Other devices › Reader apps</figcaption>\n</figure>\n<h2 id=\"add-koreader-as-a-reader-app\">Add KOReader as a reader app</h2>\n<p>Every reader app gets a user name and password of its own. Click “Add a reader app…” and name it after the device, for example “Kindle Scribe KOReader”. Readwyn turns that into the user name <code>kindle-scribe-koreader</code>.</p>\n<p>The same sheet has “Show Currently reading”, which is off by default. Tick it and KOReader also lists the books you’re reading now, most recent first. It still never sees your progress, ratings, or reviews.</p>\n<p>The next sheet has “Kindle Scribe KOReader” at the top and, under it, “In the reader app, add an OPDS catalogue and type:”, followed by the address, user name, and password, each with a Copy button. Copy the password before you close it: “Shows only now. If you lose it, remove the reader app and add it again.” It looks like <code>k7mq-2xfp-9rtw-hb3n</code>, and it’s made to be typed on an e-reader (<a href=\"/blog/download-books-to-ereader-over-wifi/\">why it’s built that way</a>).</p>\n<h2 id=\"what-koreader-did-in-readwyn-s-test\">What KOReader did in Readwyn’s test</h2>\n<p>Readwyn tested its catalogue with a macOS nightly build of KOReader from 20 October 2026, pointed at a test library. On 22 October, KOReader:</p>\n<ul>\n<li>added the catalogue with its address, user name, and password,</li>\n<li>listed Recently added, Currently reading, By author, By series, By shelf, and All books,</li>\n<li>opened an author, a series, and a shelf,</li>\n<li>found “kafka” with its search,</li>\n<li>downloaded “Crime and Punishment” as EPUB 3 and opened it.</li>\n</ul>\n<p>That book offered two downloads, “EPUB 3” first and “MOBI (original)” second. The file KOReader got was an EPUB 3 package (<code>version=\"3.0\"</code>), the copy Readwyn had converted from the MOBI. Once the reader app was removed in Readwyn, the same user name and password got HTTP 401.</p>\n<p>KOReader is the only reader app Readwyn has tested so far. On the Mac, the steps are the same for Thorium, Moon+ Reader, or any other app that reads OPDS 1.2.</p>\n<h2 id=\"after-it-s-connected\">After it’s connected</h2>\n<p>KOReader can download books and covers, and that’s all. The catalogue is read only, and a reader app has the Guest level, so the feeds carry no status, progress, rating, or review. <a href=\"/blog/what-is-opds/\">What is OPDS?</a> lists every feed it can open.</p>\n<p>The catalogue is meant for your home network. It uses plain HTTP with Basic authentication, and the warning under the address says so: “Anything on your Wi-Fi can reach the catalogue, and the reader apps send their password in clear. Share only on a network you trust.”</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/open-azw3-files-on-mac/","url":"https://readwyn.com/blog/open-azw3-files-on-mac/","title":"How to open AZW3 files on a Mac","summary":"To open an AZW3 file on a Mac, add it to Readwyn: it makes an EPUB 3 of the book when you add it, keeps the AZW3, and opens it in its own reader.","content_html":"<p>An AZW3 opens in Readwyn’s own reader like any other book. What the reader actually shows is an EPUB 3 that Readwyn made from the AZW3 while it was being added. The AZW3 stays in the library as it was.</p>\n<ol>\n<li>Click “Add books”, or drag the AZW3 files from Finder onto the window.</li>\n<li>Click the book in the library to open its book screen.</li>\n<li>Click “Read”. Once you’ve started the book, the button says “Continue reading”.</li>\n</ol>\n<p>After that it opens where you left off, because the place is saved each time you turn to a new spread.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader.webp?h=8f10e162ef8d8cef7c82\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader.webp?h=9141239fb26850877292\" alt=\"Middlemarch in the reader: two pages side by side with two highlights, the chapter path at the top, page numbers, and 18 minutes left in the chapter\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A converted book in the reader</figcaption>\n</figure>\n<p>In a window at least 1000 px wide you get two pages side by side, unless you’ve turned on “Always one page”. Font, size, margins and theme are in the Text panel, and the footer counts down the chapter: “18 min left in chapter”. <a href=\"/blog/ebook-reader-keyboard-shortcuts-mac/\">Keyboard shortcuts for reading ebooks on a Mac</a> lists the keys.</p>\n<h2 id=\"azw3-kf8-and-combo-files\">AZW3, KF8 and combo files</h2>\n<p>An AZW3 is a KF8 book on its own. A combo file has a KF8 copy and an older MOBI 6 copy of the same book in one file (<a href=\"/blog/mobi-kf8-combo-files/\">What is a MOBI and KF8 combo file?</a>). Readwyn converts both, along with the older MOBI versions. Its test set holds ten real AZW3 files, and their converted text has to match the original to the character. <a href=\"/blog/convert-mobi-to-epub-on-mac/\">How to convert MOBI to EPUB on a Mac</a> lists the other checks. A file with DRM won’t convert, though: that book keeps just the AZW3, and its button says “Open” instead of “Read”.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/open-cb7-and-cbt-comics/","url":"https://readwyn.com/blog/open-cb7-and-cbt-comics/","title":"How to open CB7 and CBT comic archives","summary":"Readwyn opens CB7 and CBT comics on a Mac: how it spots each format, the reading copy it makes when you add one, and what it sends to a Kobo.","content_html":"<p>A CB7 is a comic packed with 7-Zip, and a CBT is one packed with TAR. Add either to Readwyn on a Mac and it opens in the same page reader as a <a href=\"/blog/read-cbz-comics-on-mac/\">CBZ</a>, which keeps your progress and reading time. The archive you added stays untouched as the book's only file.</p>\n<h2 id=\"how-readwyn-tells-them-apart\">How Readwyn tells them apart</h2>\n<p>It goes by the bytes, not the extension. The 7-Zip signature at the very start marks a CB7, and <code>ustar</code> at byte 257 (where a TAR header puts it) marks a CBT. A CB7 opens if its entries use LZMA or LZMA2, or aren't compressed at all. Inside either kind, the pages are JPEG, PNG, GIF, or WebP images.</p>\n<p>CBR, the RAR version, isn't supported.</p>\n<h2 id=\"the-reading-copy\">The reading copy</h2>\n<p>When you add a CB7 or CBT, Readwyn writes a reading copy into the library's cache. It's a ZIP that holds every page under its own name, in reading order, with the image bytes unchanged. The page reader works from that copy, so opening the comic never means unpacking the archive again. A calibre import brings these comics across with the rest of the library and makes their copies in the same job, so the first open doesn't wait either.</p>\n<p>That first pass still has to read the whole archive. An 80 MB solid LZMA2 CB7 took 0.87 s to scan with the decoder's usual number of threads, and 1.9 s on a single thread. For a CB7, the <a href=\"/blog/comic-with-corrupt-page/\">16 GiB cap</a> on how much an archive may unpack to counts every entry, not only the pages, since each pass has to unpack all of them.</p>\n<p>If the cache is cleared, Readwyn makes the copy again the next time it needs it.</p>\n<h2 id=\"on-a-kobo\">On a Kobo</h2>\n<p>“Send to device” sends the reading copy. The Kobo gets a CBZ, which it opens natively, and the page images in it are byte for byte the ones in your CB7 or CBT.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/pair-second-mac/","url":"https://readwyn.com/blog/pair-second-mac/","title":"How to pair a second Mac with your library","summary":"How to pair a second Mac with the Mac that holds your Readwyn libraries, by a 6-digit code or a request you allow, and what each pairing error means.","content_html":"<p>Turn on “Readwyn server” on the Mac that holds your libraries. Then, on the second Mac, type the first Mac's address in Settings › Other devices › Another device and enter the 6-digit code the first Mac shows. If you don't have a code to hand, the second Mac can ask to connect instead, and you choose Allow on the first Mac. Either way you pair only once: after that, the second Mac connects by itself and <a href=\"/blog/share-ebook-library-between-two-macs/\">shows the first Mac's libraries</a>.</p>\n<h2 id=\"on-the-first-mac\">On the first Mac</h2>\n<p>Open Settings › Other devices:</p>\n<ol>\n<li>Turn on “Readwyn server”. While it's off, no other device can connect, and “Pair a device…” isn't there.</li>\n<li>In “Listen on”, choose your network, for example “Wi-Fi · 192.168.1.20”. If you leave the default, “This Mac only”, the pairing sheet says “This Mac listens to itself only. Choose a network in Listen on first.”</li>\n<li>Choose “Pair a device…”. The sheet shows this Mac's address with Copy, the code in two groups of three digits (“482 913”, say), and “It can” with the level the new device will get. That's Read unless you pick Guest or Admin.</li>\n</ol>\n<p>The code “Works once, for the next 10 minutes.” After 5 wrong codes in a row, every open code stops working and you have to make a new one.</p>\n<h2 id=\"pairing-with-the-code\">Pairing with the code</h2>\n<p>On the second Mac, open Settings › Other devices and choose “Another device” under “Show libraries from”. Type the address from the sheet and choose Connect.</p>\n<p>When it finds the first Mac, the second Mac says, in green, “Found Sid’s iMac. It needs to know this Mac first.” Below that are six boxes under “Type the code that Sid’s iMac shows”, with the reminder “On Sid’s iMac: Settings › Other devices › Pair a device…”. Type the six digits and choose Pair. The sheet on the first Mac says when the device has paired, and the second Mac switches to the first Mac's libraries.</p>\n<p>The code itself never crosses the network. Instead, the second Mac makes a key pair for the first Mac and sends the public half along with proof that it knows the code. The private key stays in its macOS Keychain, and the first Mac stores only the public key. Every later connection signs in with that key.</p>\n<h2 id=\"pairing-with-a-request\">Pairing with a request</h2>\n<p>Under the six boxes, “Ask Sid’s iMac to allow this Mac instead” sends a request. The second Mac then shows “Waiting for Sid’s iMac.” and “Choose Allow there, in the notification or in Settings › Other devices.”</p>\n<p>On the first Mac, a notification appears, for example “Sid’s MacBook Air asks to connect.” with “It gets the Read level. You can change it in Settings.” Allow in the notification gives it the Read level. To choose another level, answer in Settings › Other devices instead, where the request sits at the top of “Paired devices” with a level menu, Deny, and Allow.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-settings-devices.webp?h=5e2bf654f800618406e3\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-settings-devices.webp?h=1263ffab1becc7426e6c\" alt=\"Settings, Other devices: under Paired devices, Sid’s MacBook Air asks to connect, with the level Read and the buttons Deny and Allow, above three paired devices\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>A request to connect, with its level, Deny, and Allow</figcaption>\n</figure>\n<p>A request waits for 24 hours. Quitting Readwyn on the second Mac doesn't cancel it: the app checks on it again when it opens.</p>\n<h2 id=\"when-the-pairing-fails\">When the pairing fails</h2>\n<p>The second Mac gives the reason in red under the address:</p>\n<ul>\n<li>“Can’t reach 192.168.1.20:7313. Check that both devices are on the same network and that Readwyn server is on there.”</li>\n<li>“The code is not right. Check it on Sid’s iMac.”</li>\n<li>“Sid’s iMac did not allow this Mac.”</li>\n<li>“Sid’s iMac removed this Mac. Pair again.”</li>\n</ul>\n<p>macOS may also ask for permission, on the first Mac to accept incoming connections and on the second to use the local network. Allow both. Readwyn doesn't look for other Macs by name, so you always type the address.</p>\n<p>Until the pairing succeeds, the second Mac keeps showing its own libraries. If it later loses its key, after a reinstall for instance, Readwyn falls back to its own libraries and opens Settings › Other devices at the six boxes so that you can pair again.</p>\n<h2 id=\"a-server-without-a-display\">A server without a display</h2>\n<p>A Linux machine running the <a href=\"/blog/headless-ebook-server-on-linux/\">headless server</a> pairs the same way, only without a window: <code>ryn-server code</code> makes the code, and the second Mac asks you to “Type the code that <name> gives”.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/ratings-and-reviews-with-spoiler-flag/","url":"https://readwyn.com/blog/ratings-and-reviews-with-spoiler-flag/","title":"Ratings and reviews with a spoiler flag","summary":"Rate a book in half stars with ⌥-click, write a review that saves as you type, and flag its spoilers before Readwyn sends it to Goodreads.","content_html":"<p>Both the rating and the review of a book live on the Review tab of its book screen. Stars go in halves, the review saves itself as you type, and a “Contains spoilers” checkbox flags it. The rating is stored with the read you finished, so a second read gets one of its own; <a href=\"/blog/track-rereads/\">Track rereads</a> explains how.</p>\n<h2 id=\"rating-a-book\">Rating a book</h2>\n<p>On the book screen, your rating sits at the right of the panel with your progress, above the words “your rating”. Click it to open the Review tab, where a click on a star sets full stars and ⌥-click sets a half star. “Clear rating” takes the rating away again.</p>\n<p>The line under the stars spells it out: “3½ of 5 · half stars with ⌥-click”, or “Not rated · half stars with ⌥-click” before you've chosen.</p>\n<h2 id=\"writing-a-review\">Writing a review</h2>\n<p>Type straight into the Review tab. About a second after you stop, Readwyn saves the text, and only if it actually changed. The tab says “Saved as you type” and keeps a word count.</p>\n<p>If the review gives the plot away, click “Contains spoilers”. The flag travels with the review when it goes to Goodreads.</p>\n<h2 id=\"where-ratings-show-up\">Where ratings show up</h2>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library-list.webp?h=8ef4e13c111a8ae0fdcd\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library-list.webp?h=0051b07c7f0ee42ce4d5\" alt=\"The library in the list view: a row for each book with its cover, title, author, status, progress, rating, last read date, and date added\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The Rating column of the list view</figcaption>\n</figure>\n<p>The list view has a Rating column, and “Sort” can order the library by rating. The Stats screen shows your average rating, and year in review picks out your highest rated book of the year.</p>\n<h2 id=\"bringing-ratings-in-and-sending-them-out\">Bringing ratings in, and sending them out</h2>\n<p>The <a href=\"/blog/what-the-calibre-import-keeps/\">calibre import</a> keeps the ratings and reviews you had there.</p>\n<p>A Goodreads sync carries the rating and the review, spoiler flag included. The Review tab says when yours goes next (“Goes to Goodreads with the next Readwyn ⇄ Goodreads sync.”) and when the Goodreads copy last changed: “The Goodreads review was last updated 30 May 2019.”</p>\n<p>A Kobo sync copies the rating from whichever side you sync from. Readwyn's half stars convert to the device's stars, and a half star rounds up.</p>\n<h2 id=\"removing-a-book\">Removing a book</h2>\n<p>“Remove from library…” takes the review with it. The alert warns you first: “Its files, 14 highlights and your review go too.”</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-a-book-privately/","url":"https://readwyn.com/blog/read-a-book-privately/","title":"Read a book privately without saving progress or stats","summary":"Private reading in Readwyn opens a book and records nothing: no progress, dates, time read, status change, or reading-service sync. Lookup, find and Copy still work.","content_html":"<p>Private reading opens a book without Readwyn recording anything about it. Start it from the book screen: open the “…” menu and choose “Read privately”, or press ⌥⌘R.</p>\n<h2 id=\"what-isn-t-saved\">What isn’t saved</h2>\n<p>While you read privately, Readwyn doesn’t record:</p>\n<ul>\n<li>your place in the book or your progress,</li>\n<li>a start date or a last-read date,</li>\n<li>the time read, so your reading stats and your reading speed stay as they were,</li>\n<li>a change of status (a book on “To read” or “Next up” normally moves to “Reading” once the reader records progress, but here it stays where it was),</li>\n<li>a sync to a reading service such as Goodreads.</li>\n</ul>\n<p>Close the book and Readwyn forgets where you got to. Open it next time with “Read” or “Continue reading” and it goes to your saved place, as if the private session never happened.</p>\n<h2 id=\"how-you-can-tell\">How you can tell</h2>\n<p>A pill in the top bar, beside the chapter path, reads “Private reading · nothing is saved”, with an eye-off icon.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-private.webp?h=eb31eb5ff26f565de2ff\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-private.webp?h=ef3b87892f49a7311e3e\" alt=\"The reader in private reading: the pill Private reading · nothing is saved in the top bar, and a selection with the popover No highlights in private reading, Copy, and Find\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>In private reading, the selection popover offers no highlight colours</figcaption>\n</figure>\n<p>The footer changes too. Where it normally shows the time left in the chapter, it says “Your saved place stays at p. 196”, with the page you’ll return to. Until Readwyn knows that page, it says “Your saved place stays where it was”. A book with no saved place yet says “Nothing is saved”, as in the capture.</p>\n<h2 id=\"what-still-works\">What still works</h2>\n<p>Everything that leaves no trace still works. <a href=\"/blog/look-up-words-in-ebooks-on-mac/\">Lookup</a> opens when you hold a word, and its speaker button reads the word aloud. Copy and ⌘C copy a selection, and ⌘F <a href=\"/blog/find-text-in-ebook-on-mac/\">searches the open book</a>.</p>\n<p>The menus lose whatever would save something:</p>\n<ul>\n<li>The selection popover shows “No highlights in private reading” instead of the colour dots and “Add note”.</li>\n<li>Right-clicking a word gives no highlight items and no My Words items.</li>\n<li>The lookup popover has no “Save to My Words”.</li>\n</ul>\n<h2 id=\"on-a-guest-mac\">On a Guest Mac</h2>\n<p>A Mac that you pair at the <a href=\"/blog/guest-access-to-ebook-library/\">Guest level</a> reads every book this way. Its book screen offers “Read privately” as the main button and has no status, no “Edit details”, and no “…” menu, so nothing it reads touches your progress or your time read.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-cbz-comics-on-mac/","url":"https://readwyn.com/blog/read-cbz-comics-on-mac/","title":"How to read CBZ comics on a Mac","summary":"Add CBZ comics to Readwyn on a Mac and read them in its page reader: two pages side by side, fit page or fit width, zoom, and the keys that turn pages.","content_html":"<p>Drop the CBZ on Readwyn's window, or use “Add books”, then click “Read” on the comic's book screen. It opens in the page reader, and nothing gets converted: the CBZ you added is the only file the book has.</p>\n<h2 id=\"adding-comics\">Adding comics</h2>\n<p>Comics from a calibre library come in with the rest of the import. A CBZ is a ZIP archive of page images in JPEG, PNG, GIF, or WebP, and Readwyn will also take a plain <code>.zip</code> as a CBZ if it holds page images and isn't an EPUB or an OEB package.</p>\n<p>The pages are the images in the archive, sorted the way a person would sort the names, so <code>page2.jpg</code> comes before <code>page10.jpg</code>. Anything that isn't an image is left out, and so is the <code>__MACOSX</code> folder. An archive with no page images at all isn't added, and the toast after the import says why. A <a href=\"/blog/comic-with-corrupt-page/\">comic with one broken page</a> still comes in and opens.</p>\n<p>If there's a <code>ComicInfo.xml</code> inside, Readwyn takes the title, the writers and artists, the series and its number, and the cover from it. The <a href=\"/blog/comicinfo-xml/\">ComicInfo.xml post</a> goes through every field.</p>\n<p>In the library a comic looks like any other book, and its book screen gives the exact page count from the archive (“24 pages”) instead of an estimate. Because the archive is never converted, “Send to device” copies that same CBZ to a Kobo, under <code>Books/&lt;first author&gt; - &lt;title&gt;.cbz</code>, and the Kobo opens it as it is.</p>\n<h2 id=\"reading\">Reading</h2>\n<p>Once you've started a comic, the button says “Continue reading” and opens it where you stopped. A wide window shows two pages side by side, with the cover and any <a href=\"/blog/double-page-spreads-in-comics/\">double-page spread</a> on their own. If you'd rather see one page at a time, choose “Always one page” in Settings › Reading.</p>\n<ul>\n<li>“Fit page”, the default, shows the whole spread. “Fit width” makes one page as wide as the window, and you scroll down it.</li>\n<li>⌘= and ⌘- zoom, ⌘0 goes back to the fit, and a pinch works too.</li>\n<li>Space, Page Down, the left and right arrow keys, and the round buttons beside the pages turn the page. Shift+Space and Page Up go back, and Escape closes the comic.</li>\n</ul>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-comic-spread.webp?h=1ebbec3459ebd7ae3959\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-comic-spread.webp?h=24835cf8d48886d7498f\" alt=\"A comic open in Readwyn's page reader, one wide page filling the window, with the fit buttons at the top right and the page number, percentage, and time left in the footer\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A CBZ open in the page reader</figcaption>\n</figure>\n<p>The footer shows the page you're on (“p. 7 of 24”), how far through you are as a percentage, and the minutes left. Press or drag on the track to jump to a page. The time left comes from your <a href=\"/blog/reading-stats-for-comics/\">reading speed in pages per minute</a>, which Readwyn learns as you read comics.</p>\n<p>A comic is pictures, so there's no find, no dictionary lookup, and no highlighting in the page reader, and its book screen has no Highlights tab.</p>\n<h2 id=\"large-archives\">Large archives</h2>\n<p>A 1 GB CBZ of 1,000 uncompressed 1 MB pages, timed on a Mac with the file already in the disk cache, took 5.9 ms to scan and 3.5 ms to open, and one page took 0.17 ms to read. The reader unpacks just the current spread and two pages on either side of it, and decodes only the pages around the one you're on.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-comics-in-chunky/","url":"https://readwyn.com/blog/read-comics-in-chunky/","title":"Browse your comics in Chunky from your home library","summary":"Browse your Readwyn comics in Chunky over OPDS: each series in issue order, shelves of comics only, and the books you’re reading now.","content_html":"<p>What Chunky shows of your comics comes from the details you keep in Readwyn, because it reads them through Readwyn’s OPDS catalogue. Get the series and numbers right and every series lists its issues in order. Shelves do the rest: Chunky sees each one as a list of its own.</p>\n<p>To connect it, switch on the catalogue under “Reader apps” in Settings › Other devices, add a reader app called “Chunky”, and type the address, user name, and password from the next sheet into Chunky. Each comic then downloads as the CBZ, CB7, or CBT archive you added. The Mac side is the same for every app, and <a href=\"/blog/opds-catalogue-for-koreader/\">the KOReader walkthrough</a> shows it screen by screen. KOReader is also the app Readwyn tested with; Chunky hasn’t been tried.</p>\n<h2 id=\"series-in-issue-order\">Series in issue order</h2>\n<p>By series lists your series by name. Open one and its issues come by number, not by title, so issue 2 sits before issue 10.</p>\n<p>Readwyn reads the series and number from <code>ComicInfo.xml</code> when you add a comic, if the archive has one:</p>\n<ul>\n<li><code>Series</code> and <code>Number</code> give the series and the issue number. A <code>Number</code> that isn’t a number becomes 1.</li>\n<li><code>Title</code> gives the title. Without one, Readwyn uses the series and number, such as “The Lantern Keeper #1”, or the series alone. With neither, it falls back to the file name.</li>\n</ul>\n<p>If a comic has no <code>ComicInfo.xml</code>, or the number is wrong, fix it yourself: “Edit details” on the book screen has fields for the series and its number.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-comic-book.webp?h=3c139e32e63e50aaf717\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-comic-book.webp?h=22ee8ea9dd9bc59a3077\" alt=\"The book screen of the comic The Lantern Keeper by Mira Okafor and Tomas Lindqvist: 2024, 10 pages, English, CBZ of 1.2 MB, with Continue reading, 40% read, page 5 of 10, the shelves Favourites and Book club, and the reading history\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A comic in the library, with its shelves</figcaption>\n</figure>\n<h2 id=\"shelves-of-nothing-but-comics\">Shelves of nothing but comics</h2>\n<p>Recently added and All books mix comics and ebooks, because your Readwyn library keeps them together. Shelves are where you separate them. By shelf lists every shelf, and each one opens as a list of its own, sorted by title. Put your comics on a shelf called “Comics”, or make one shelf per publisher, and Chunky gets lists with nothing else in them.</p>\n<h2 id=\"what-you-re-reading-now\">What you’re reading now</h2>\n<p>Each reader app has a “Currently reading” checkbox, off by default. Tick it next to Chunky in the list of reader apps and Chunky gets a Currently reading list: everything with the status Reading, most recently read first. Your progress stays behind on the Mac, and so do your ratings and reviews.</p>\n<p>The list follows the status, not the kind of book, so an ebook you’re halfway through shows up there next to your comics.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-comics-in-panels/","url":"https://readwyn.com/blog/read-comics-in-panels/","title":"Download your comics to Panels from your home library","summary":"Get the comics in your Readwyn library into Panels over OPDS: each one downloads as the CBZ, CB7, or CBT archive you added, with its cover, series, and number.","content_html":"<p>Panels reads OPDS catalogues, and Readwyn serves one from the Mac that keeps your library. Your comics sit in it alongside your ebooks, and each one downloads to Panels as the CBZ, CB7, or CBT archive you added, byte for byte.</p>\n<h2 id=\"add-panels-as-a-reader-app\">Add Panels as a reader app</h2>\n<p>Switch the catalogue on as the <a href=\"/blog/opds-catalogue-for-koreader/\">KOReader post</a> describes, then click “Add a reader app…” and pick a name that says both the device and the app, for example “Sid’s iPad · Panels”. If you read comics on more than one device, add each one separately: every reader app gets its own row, user name, and password.</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-apps.webp?h=e0024127c108d4d70c9a\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-apps.webp?h=b3168403e87d03df89f1\" alt=\"Settings › Other devices › Reader apps: the address http://192.168.1.20:7314/opds, and three reader apps, the second named Sid’s iPad · Panels and last connected yesterday\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>One reader app for each device and app</figcaption>\n</figure>\n<h2 id=\"the-archive-you-added-unchanged\">The archive you added, unchanged</h2>\n<p>Readwyn never converts a comic, so the archive you added is its only file, and that’s what Panels downloads. A Readwyn library takes three kinds of archive: CBZ (a ZIP of page images), CB7 (a 7-Zip archive), and CBT (a TAR archive). Each has its own file type in the catalogue:</p>\n<table><thead><tr><th>Archive</th><th>File type in the catalogue</th></tr></thead><tbody>\n<tr><td>CBZ</td><td><code>application/vnd.comicbook+zip</code></td></tr>\n<tr><td>CB7</td><td><code>application/x-cb7</code></td></tr>\n<tr><td>CBT</td><td><code>application/x-cbt</code></td></tr>\n</tbody></table>\n<p>Those are the types Readwyn sends. Nobody has checked how Panels handles each one yet: so far the catalogue has only been tried with KOReader. The suggested file name is the title and first author, as for any book.</p>\n<h2 id=\"covers-series-and-numbers\">Covers, series, and numbers</h2>\n<p>When you add a comic whose archive has a <code>ComicInfo.xml</code> inside, Readwyn takes the title, authors, series and number, summary, publisher, date, language, and genres from that file. The catalogue serves them as they stand in your library, so a series you correct with “Edit details” is the series Panels gets. <a href=\"/blog/comicinfo-xml/\">What ComicInfo.xml does for your comic library</a> goes through each field.</p>\n<p>The cover is the page that <code>ComicInfo.xml</code> marks as the front cover, or the first page if none is marked. The catalogue serves it in two sizes: the full cover, and a small copy for lists.</p>\n<p>Each entry’s text opens with the series and number, for example “The Lantern Keeper, book 1”, and By series lists the issues of a series in number order.</p>\n<h2 id=\"progress-stays-on-the-mac\">Progress stays on the Mac</h2>\n<p>Readwyn’s page reader on the Mac keeps the progress and time read of each comic, shows double-page spreads, and reads manga right to left (<a href=\"/blog/comics-and-ebooks-in-one-library/\">more on comics in Readwyn</a>). The catalogue is read only, so the pages you read in Panels never count towards that progress.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-ebooks-on-linux-wayland-x11/","url":"https://readwyn.com/blog/read-ebooks-on-linux-wayland-x11/","title":"Read ebooks on Linux with Wayland or X11","summary":"Readwyn runs in both Wayland and X11 sessions. The 3D library locks the mouse differently in each, and the app draws its own window controls unless the desktop does.","content_html":"<p>Both work. Readwyn runs in a Wayland session and in an X11 one, and it relies only on services that every major desktop provides. Tests covered Wayland with GNOME and KDE Plasma, and X11 with GNOME (the <a href=\"/blog/ebook-manager-arm64-linux/\">arm64 post</a> lists each system). The two sessions differ in how the 3D library takes the mouse, and on GNOME under Wayland a window that the tray icon opens doesn't come to the front. Who draws the window frame depends on the desktop rather than the session.</p>\n<h2 id=\"window-controls\">Window controls</h2>\n<p>GNOME on Wayland doesn't draw title bars for apps, so Readwyn draws its own: minimize, maximize and close, three small circles at the right end of each window's top bar. The library, the reader, the comic reader and Settings all have them. GNOME's X11 session shows the same controls; the Linux screenshots on this site were taken there.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/linux/dark-library.webp?h=02876295485e7080879c\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/linux/light-library.webp?h=24919ab2ffd9d07527ff\" alt=\"The Readwyn library on Linux: at the top right, after the search and Add books, three round window controls for minimize, maximize, and close\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The window controls at the right end of the top bar</figcaption>\n</figure>\n<p>The top bar then behaves like a title bar:</p>\n<ul>\n<li>Dragging an empty part of it, or the empty strip at the top of the sidebar, moves the window once the pointer has travelled 4 px.</li>\n<li>Double-clicking there maximizes the window, or restores it.</li>\n<li>Right-clicking there opens the system's window menu.</li>\n<li>A 5 px band along each edge resizes the window, and each corner catches the pointer from 16 px along both of its edges.</li>\n</ul>\n<p>A maximized or tiled window has no border, and its tiled edges don't resize. Closing the library window quits the app, as on the Mac.</p>\n<p>Some systems draw the decorations themselves: an X11 window manager without a compositor, say, or a Wayland compositor that keeps its own title bars. Readwyn then leaves the frame to them and draws no controls, border or resize bands.</p>\n<p>Every window carries the app ID <code>com.readwyn.Readwyn</code>, the same name as the menu entry, so the desktop matches each window to that entry and shows Readwyn's icon for it.</p>\n<h2 id=\"mouse-look-in-the-3d-library\">Mouse look in the 3D library</h2>\n<p>Click the scene in the 3D library and the pointer locks: a crosshair appears and the mouse turns the view. Esc frees it again. Wayland and X11 handle the lock differently:</p>\n<ul>\n<li><strong>Wayland.</strong> The compositor locks the pointer and reports its motion without acceleration. When you press Esc, the cursor comes back in the middle of the window. A compositor that lacks <code>zwp_pointer_constraints_v1</code> and <code>zwp_relative_pointer_v1</code> leaves the pointer free: the crosshair shows, but the mouse doesn't turn the view.</li>\n<li><strong>X11.</strong> Readwyn keeps the cursor inside the window and turns the view by the raw motion of the mouse or touchpad, skipping the X server's acceleration.</li>\n</ul>\n<p>On 25 October 2026 this was tested on Ubuntu 24.04 with GNOME 46, with a virtual mouse that moved 10 counts at a time, forty times. On both Wayland and X11 the view turned exactly 10 points per move. On Wayland the compositor reported 11.07 points of accelerated motion for each of those moves; the view uses the unaccelerated 10.</p>\n<h2 id=\"reduced-motion\">Reduced motion</h2>\n<p>If you turn animations off in your desktop's settings, Readwyn tones its own motion down too, as it does with Reduce Motion on a Mac. In the 3D library, a glide to a book becomes a short fade, and Settings › 3D library says so: “With animations off in the settings of the desktop, a glide to a book is a short fade.” Readwyn reads the setting through the settings portal, so GNOME's <code>enable-animations</code> set to false and KDE's <code>AnimationDurationFactor</code> at 0 both count.</p>\n<h2 id=\"when-a-window-won-t-come-forward-on-gnome\">When a window won't come forward on GNOME</h2>\n<p>The background server can open a page in the running app, for example “Connected devices…” from its icon in the system tray. On GNOME under Wayland, though, the window doesn't come to the front this way. In a test, the app opened Settings › Other devices as asked, and GNOME showed a notification that “Other devices” is ready instead of raising the window.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-epub-on-mac/","url":"https://readwyn.com/blog/read-epub-on-mac/","title":"How to read EPUB books on a Mac with Readwyn","summary":"Open an EPUB in the Readwyn reader on a Mac: how pages turn, what the Text panel sets, how the contents and footnotes work, and where the book opens again.","content_html":"<p>Any EPUB in your Readwyn library opens in the app’s own reader: click “Read” on the book’s screen. If the window is wide enough, you get two pages side by side. Next time the book opens where you stopped, and once you’ve started it the button says “Continue reading”.</p>\n<h2 id=\"which-books-open-in-the-reader\">Which books open in the reader</h2>\n<p>The reader only shows EPUB, but that covers more books than it sounds. When a MOBI, AZW3, PDF or other older file comes into the library, Readwyn makes an EPUB 3 copy of it, and the reader opens that copy. The original stays as it was, next to it (<a href=\"/blog/convert-old-ebook-formats-keep-original/\">more on the conversion</a>).</p>\n<p>A book with no EPUB at all has “Open” instead of “Read”. That hands the file to its default app, and the “…” menu names the app, for example “Open in Books”. Comics are the exception: they get “Read” too, and it opens them in a page reader of their own.</p>\n<h2 id=\"the-page\">The page</h2>\n<p>This isn’t a web view. Readwyn lays the book out itself: paragraphs, headings, lists, block quotes, images, rules and simple tables, with italic, bold, small caps, superscripts and subscripts. Most of the publisher’s CSS is ignored, but the simple rules for alignment, indents, margins and font style are kept.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader.webp?h=8f10e162ef8d8cef7c82\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader.webp?h=9141239fb26850877292\" alt=\"A two-page spread of Middlemarch in the Readwyn reader, with two highlights and the popover of a note\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Two pages side by side, with the running head and the time left in the chapter</figcaption>\n</figure>\n<p>You get two pages once the window is at least 1000 px wide, and one page below that. In an English book, a word that doesn’t fit at the end of a line is broken between syllables with a hyphen.</p>\n<p>At the top of each page is a running head: the book title on the left page, the chapter path (“Book II · Chapter XX”) on the right. A page where a chapter, or any other entry in the contents, begins has no running head. The page number sits at the bottom.</p>\n<h2 id=\"turning-pages\">Turning pages</h2>\n<p>You can turn a page with:</p>\n<ul>\n<li>the arrow keys, Space, Page Up or Page Down,</li>\n<li>the two round buttons beside the sheet,</li>\n<li>a swipe on the trackpad,</li>\n<li>a click on the left or right part of the page, if “Turn the page with a click on its left or right side” is on in Settings › Reading.</li>\n</ul>\n<p>The turn itself can slide, cover, fade, curl, flip, or not animate at all. Curl rolls the sheet over from its right edge; flip swings it on the spine like a stiff card.</p>\n<h2 id=\"the-text-panel\">The Text panel</h2>\n<p>Click “Text” in the top bar for the type settings:</p>\n<ul>\n<li>the font: Literata, Georgia, Palatino, Charter, Avenir or Helvetica,</li>\n<li>the size, in 1 pt steps (19 pt to start with),</li>\n<li>the line spacing: Tight, Normal or Loose,</li>\n<li>the margins: Narrow, Normal or Wide,</li>\n<li>the page turn,</li>\n<li>the page theme: App, Light or Dark (<a href=\"/blog/dark-mode-for-ebooks-on-mac/\">how the dark page works</a>).</li>\n</ul>\n<p>They apply to every book on this Mac. Settings › Reading has the same type settings with a preview, along with the click-to-turn checkbox and “Always one page”, for when you don’t want a spread even in a wide window.</p>\n<h2 id=\"getting-around-a-book\">Getting around a book</h2>\n<p>“Contents” docks the table of contents on the left, with a page number beside each entry. The entry you’re in reads “· you are here”. Type in “Find a chapter” to narrow the list to the entries whose title contains your text.</p>\n<p>Footnotes open in place. Click the reference and the note appears in a popover just under it, with a “Go to the note” button.</p>\n<p>Along the footer runs the progress track, with a tick for each top-level entry in the contents. Press or drag on it to jump. Below it you’ll see the page (“p. 196 of 880”), the percentage, and something like “18 min left in chapter”. The time left comes from your own reading speed (<a href=\"/blog/how-long-to-read-a-book/\">how Readwyn measures it</a>).</p>\n<p>To search the open book, use “Find” in the top bar (<a href=\"/blog/find-text-in-ebook-on-mac/\">finding text in an ebook</a>).</p>\n<h2 id=\"what-gets-saved\">What gets saved</h2>\n<p>While you read, Readwyn saves your place, your progress (in whole percent), the date you last read and how long you’ve spent. A book with no status, or with “To read” or “Next up”, moves to “Reading” as soon as progress is recorded. If you’d rather none of this were saved, open the book in <a href=\"/blog/read-a-book-privately/\">private reading</a>.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-guest-admin-access-levels/","url":"https://readwyn.com/blog/read-guest-admin-access-levels/","title":"Read, Guest, and Admin: access levels for your ebook library","summary":"What a paired Mac can and can't do at the Readwyn levels Guest, Read, and Admin, and how to change a level or answer a request for more access.","content_html":"<p>Every Mac you pair with your libraries gets one of three levels: Guest, Read (the default), or Admin. The Mac that holds the libraries is always Admin for itself.</p>\n<h2 id=\"the-three-levels\">The three levels</h2>\n<p>The level menu in Settings › Other devices sums up each one in a line:</p>\n<table><thead><tr><th>Level</th><th>What the menu says</th></tr></thead><tbody>\n<tr><td>Guest</td><td>“Read books. Nothing is saved; every book opens privately.”</td></tr>\n<tr><td>Read</td><td>“Also keep progress, highlights, notes and My Words, keep books offline and sync a Kobo.”</td></tr>\n<tr><td>Admin</td><td>“Everything, also adding and removing books, editing details, libraries and these settings.”</td></tr>\n</tbody></table>\n<p>Admin gets a second line as a warning: “Only for a device you own. It can delete books and libraries for every device.”</p>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-settings-devices.webp?h=5e2bf654f800618406e3\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-settings-devices.webp?h=1263ffab1becc7426e6c\" alt=\"Settings, Other devices: Paired devices with Studio Mac mini at Admin, Living room iMac at Guest, and Old MacBook Pro at Read, each with a level menu and Remove\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>Each paired device has its own level</figcaption>\n</figure>\n<h3 id=\"guest\">Guest</h3>\n<p>A <a href=\"/blog/guest-access-to-ebook-library/\">Guest Mac</a> can browse the lists, open any book or comic, look up words, and search inside a book. Every book opens in private reading, so nothing it does is saved. Its book screen offers “Read privately” and leaves out the controls that would change anything. There's no Stats and no year in review, and it can't keep books offline.</p>\n<h3 id=\"read\">Read</h3>\n<p>A Read Mac can also see and change your reading data, just as you can on the Mac that holds the libraries:</p>\n<ul>\n<li>progress, the reading status, and shelves, even for many books at once,</li>\n<li>the rating, the review, and its spoiler flag,</li>\n<li>highlights, notes, and My Words,</li>\n<li>keeping books <a href=\"/blog/keep-book-offline-on-laptop/\">offline</a> with “Available offline”,</li>\n<li>a Kobo sync and sending books to a Kobo, with “Remove from device”,</li>\n<li>the reading stats and year in review.</li>\n</ul>\n<p>What it can't do is add or remove a book, convert it, edit its details or its cover, or fetch metadata.</p>\n<h3 id=\"admin\">Admin</h3>\n<p>An Admin Mac can also change the libraries themselves and Readwyn server's settings. That covers adding, removing, converting, and editing books, fetching metadata, importing from calibre, creating and renaming libraries, and running compaction. Connecting or disconnecting a reading service, and pairing or removing a device, need Admin.</p>\n<h2 id=\"what-the-other-mac-sees\">What the other Mac sees</h2>\n<p>On a paired Mac, Settings › Other devices has a “Your access” section, marked “Set on Sid’s iMac”, that says what the level allows. At Read it says “Read, keep your place, highlight, keep books offline and sync a Kobo.” A Read Mac never shows “Edit details”: the app leaves out each control the level doesn't allow. The library switcher's menu gives the level as well.</p>\n<h2 id=\"changing-a-level\">Changing a level</h2>\n<p>In Settings › Other devices on the Mac that holds the libraries, every device under “Paired devices” has its own level menu. While its window is open, a paired Mac asks for its level every 30 seconds, so it picks up a change without reconnecting.</p>\n<p>A pairing code carries a level too. Set it in “It can” on the “Pair a device…” sheet before you hand the code over.</p>\n<h2 id=\"asking-for-more\">Asking for more</h2>\n<p>A Guest or Read Mac can ask to move up one level: “Your access” has “Ask for Read access…” on a Guest Mac and “Ask Sid’s iMac for Admin access…” on a Read Mac. On the Mac that holds the libraries, the request shows up both as a notification with Allow and Deny and as an entry in “Paired devices”. Allow gives the level that was asked for. Deny leaves the level as it was, and the other Mac shows that the request was turned down. A request waits for 24 hours, and a new one replaces the old.</p>\n<h2 id=\"reader-apps\">Reader apps</h2>\n<p>A reader app such as KOReader, which reads your libraries over <a href=\"/blog/what-is-opds/\">OPDS</a>, isn't a paired Mac. It gets the Guest level, so its feeds show no status, progress, rating, or review. The one exception is each reader app's “Currently reading” setting, which lets that app see the books you're reading, though not your progress in them.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-manga-right-to-left-on-mac/","url":"https://readwyn.com/blog/read-manga-right-to-left-on-mac/","title":"Read manga from right to left on a Mac","summary":"How Readwyn knows that a manga reads from right to left, and what changes in its page reader: the page order, the arrow keys, the swipe, and the progress track.","content_html":"<p>A manga turns from right to left in Readwyn's page reader when the <code>ComicInfo.xml</code> inside its archive has <code>Manga</code> set to <code>YesAndRightToLeft</code>. Open one and the earlier page sits on the right, the left arrow key goes forward, and the progress track fills from the right. It works the same for a CBZ, a CB7, or a CBT.</p>\n<h2 id=\"where-the-direction-comes-from\">Where the direction comes from</h2>\n<p>Readwyn reads <code>ComicInfo.xml</code> when you add the comic and keeps the direction with the book. Any other value of <code>Manga</code> means left to right, and so does a missing or broken <code>ComicInfo.xml</code>. If an archive holds more than one <code>ComicInfo.xml</code>, the rules for which one counts are in the <a href=\"/blog/comicinfo-xml/\">ComicInfo.xml post</a>.</p>\n<p>Comics that arrive with a calibre import are no different here. Their title, authors, and series come from calibre, but the direction still comes from the archive.</p>\n<h2 id=\"what-flips\">What flips</h2>\n<ul>\n<li>With two pages open, the earlier page is on the right and the later one on the left.</li>\n<li>The left arrow key, the left turn button, and a swipe to the right go to the next page.</li>\n<li>The progress track at the bottom fills from the right.</li>\n<li>The top bar shows the pill “Right to left”.</li>\n</ul>\n<p>Pages pair up and spreads show alone just as in any other comic, only mirrored. The pairing rules are in <a href=\"/blog/double-page-spreads-in-comics/\">how Readwyn shows double-page spreads</a>.</p>\n<h2 id=\"checking-a-book-s-direction\">Checking a book's direction</h2>\n<p>On the book screen, the Details tab has a “Reading” row where an ebook would show its word count. For a manga it says “Right to left · from ComicInfo.xml”, and for every other comic “Left to right”.</p>\n<h2 id=\"progress-and-your-place\">Progress and your place</h2>\n<p>Reading right to left doesn't change what Readwyn records. Progress is the percentage of pages before the one you're on, in whole percent, so it's the same in either direction. Time read counts the same as for any other comic, and the time left is the pages still to go at your comics reading speed.</p>\n<p>When you stop, Readwyn saves the later page of the spread, which in a manga is the one on the left. Next time, the reader finds that page again by its file name in the archive.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-palmdoc-ebooks/","url":"https://readwyn.com/blog/read-palmdoc-ebooks/","title":"How to read old PalmDOC ebooks today","summary":"Old PalmDOC books (.pdb or .prc) still open: Readwyn converts each to EPUB 3 when you add it, for reading on a Mac or a Kobo, and keeps the original.","content_html":"<p>Readwyn converts a PalmDOC book (<code>.pdb</code> or <code>.prc</code>) to EPUB 3 when you add it, so “Read” opens it on your Mac and “Send to device” puts it on a Kobo. Your PalmDOC file stays in the library as it was.</p>\n<p>Drop the files on the window, then click a book and “Read”. A file with DRM won’t convert, and its book keeps only the PalmDOC file. Old books sitting in a calibre library go through the same conversion when you import it (<a href=\"/blog/import-calibre-library/\">How to import your calibre library into Readwyn</a>).</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader.webp?h=8f10e162ef8d8cef7c82\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader.webp?h=9141239fb26850877292\" alt=\"Middlemarch in the reader: two pages side by side, with the chapter path at the top, page numbers at the bottom, and the time left in the chapter\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A converted book reads like any EPUB</figcaption>\n</figure>\n<p>You can highlight, take notes, look up words and search the text as in any other book.</p>\n<h2 id=\"from-plain-text-to-chapters\">From plain text to chapters</h2>\n<p>A PalmDOC book is usually plain text in the Windows-1252 character set, sometimes followed by a list of bookmarks. With no markup to go on, Readwyn makes the paragraphs itself: at blank lines if the file has them, otherwise one paragraph for each line. Each bookmark becomes an entry in the table of contents, and a bookmark at the start of a paragraph starts a new chapter. No character of the text is lost or added, and <code>epubcheck</code> finds no errors in the result.</p>\n<p>The converter is tested on ten PalmDOC books, used exactly as found, each at least 25,000 words long (Readwyn counts that as 100 pages). After conversion, every one was also opened in the reader, and two of its spreads were captured and reviewed.</p>\n<p>Some PalmDOC files hold an HTML document instead; Mobipocket Creator wrote them like that. Two of the ten test books turned out to be this kind, both <code>.prc</code> files from MobileRead. Readwyn converts those like a MOBI, with the bookmarks as the table of contents.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/read-pdf-as-reflowable-text-on-mac/","url":"https://readwyn.com/blog/read-pdf-as-reflowable-text-on-mac/","title":"Read a PDF book as reflowable text on a Mac","summary":"A PDF book converted to EPUB 3 reflows in your own font and size, without its running headers and page numbers. A scanned PDF stays as it is.","content_html":"<p>Once a PDF is in your Readwyn library, “Read” on its book screen opens it as reflowable text. Readwyn converts the PDF to EPUB 3 as it adds the book, so the <a href=\"/#reader\">reader</a> can set the text as ordinary pages in the font and size you choose, at the width of your window. The PDF itself stays in the library, untouched.</p>\n<h2 id=\"opening-it\">Opening it</h2>\n<p>Click “Add books”, or drag the PDF from Finder onto the window. On the book screen, “Read” (“Continue reading” once you've started) opens the converted text. In the Details tab, “Files” lists the PDF as “Original” and the EPUB 3 as “For reading”.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-select.webp?h=f7d09bb574dbc078aa37\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-select.webp?h=30358a89252416889c61\" alt=\"A two-page spread in the reader, with a passage selected and the four highlight colours above it\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A book in the reader. A converted PDF reads the same way.</figcaption>\n</figure>\n<p>From then on the reader treats the book like any other EPUB. Change the font or size in the Text panel and the pages reflow to fit. A window at least 1000 px wide shows two pages side by side, unless Settings › Reading is set to “Always one page”; a narrower one shows a single page. In an English book, a word that doesn't fit at the end of a line is hyphenated between syllables. The footer shows your page, the percent read and the time left in the chapter.</p>\n<h2 id=\"taking-the-page-apart\">Taking the page apart</h2>\n<p>A PDF pins every line to a fixed spot on the page, so the conversion has to work out what the lines were:</p>\n<ul>\n<li>Running headers, running footers and page numbers are dropped. A page number in brackets, which marks a page of the printed edition, stays where the page broke.</li>\n<li>Lines are joined into paragraphs, across page breaks and across columns. A page set in two columns reads one column, then the other.</li>\n<li>A word split by a hyphen at the end of a line becomes one word again, unless the book uses the hyphenated form elsewhere and never the joined one, or the second half starts with a capital.</li>\n<li>A footnote at the bottom of a page moves to the end of the paragraph that the page break interrupted, so it doesn't land in the middle of a sentence.</li>\n<li>A line whose parts are more than two ems apart becomes a table row.</li>\n</ul>\n<p>Italic, bold, superscripts and subscripts survive, and so do the headings and the images (apart from JPEG 2000 and JBIG2 images). The table of contents comes from the PDF's outline, or from its headings if it has no outline. Links inside the book still work.</p>\n<h2 id=\"tested-on-50-pdfs\">Tested on 50 PDFs</h2>\n<p>The conversion was tested on 50 PDFs from Project Gutenberg, each compared with the EPUB 3 edition of the same book. Across all 50, the word error rate (words that are wrong, missing or extra) was 9.15%, and the mean paragraph-break F1 score was 0.753. In the 42 books with headings, 91% of the reference headings came out as headings on average.</p>\n<p>On the test Mac, converting one book at a time, half the books took a hundredth of a second or less, and all but one took under a tenth. The slowest, a Portuguese dictionary of 2,164 pages, took 0.45 seconds. <a href=\"/blog/convert-pdf-to-epub-without-calibre/\">Convert PDF to EPUB without calibre</a> has every measure and the baseline.</p>\n<h2 id=\"scans-and-encrypted-pdfs\">Scans and encrypted PDFs</h2>\n<p>There's no OCR, so a scanned PDF with no text layer won't convert, and nor will an encrypted one. The book then keeps just the PDF. Its book screen shows “Open” instead of “Read”, which opens the PDF in its default app; the “…” menu names that app.</p>\n<h2 id=\"on-a-kobo\">On a Kobo</h2>\n<p>“Send to device” sends the converted EPUB 3 to a Kobo, as a KEPUB if “Convert EPUB to KEPUB for Kobo e-readers” is on.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/reading-speed-words-per-minute/","url":"https://readwyn.com/blog/reading-speed-words-per-minute/","title":"Find your reading speed in words per minute","summary":"Readwyn works out your words per minute from your own page turns, starting from 250, and shows it as “Reading speed” in Stats.","content_html":"<p>Every library in Readwyn starts at 250 words per minute, and your own page turns take over from there, so there's no test to sit. “Reading speed” in Stats shows your words per minute for whichever range you've picked at the top of the screen.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-stats.webp?h=6be17290a8c05ed4461f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-stats.webp?h=4b431a86ee04dd47470d\" alt=\"Reading stats for October 2026: 21 h 46 min read, a bar for each day, and five figures, among them Reading speed 284 words per minute\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Reading speed among the figures of a month</figcaption>\n</figure>\n<h2 id=\"which-gaps-become-samples\">Which gaps become samples</h2>\n<p>Each gap between two page turns in the reader can become a sample. To count towards your speed, it has to:</p>\n<ul>\n<li>last 5 minutes or less, like any <a href=\"/blog/track-reading-time-automatically/\">recorded reading time</a></li>\n<li>move forward in the book (going back adds no words)</li>\n<li>work out at 50 to 1,000 words per minute</li>\n</ul>\n<p>Anything faster is a skim, a jump through the contents, or a jump to a search result; anything slower is a pause on one page.</p>\n<p>Readwyn counts the words between where you were and where you are now, in reading order, in the book's own text. A page of short dialogue counts fewer words than a dense page of description.</p>\n<h2 id=\"the-formula\">The formula</h2>\n<p>Each library works out its own speed from its own samples:</p>\n<p><code>(W + 250 × 10) / (T + 10)</code></p>\n<p>where <code>W</code> is the words and <code>T</code> the minutes of all your samples. The default of 250 words per minute counts as if you'd already read for 10 minutes at that pace. In practice:</p>\n<ul>\n<li>With no samples, the speed is 250, and the time-left estimates use that.</li>\n<li>Ten minutes at 350 words per minute gives 300.</li>\n<li>Ten hours at 350 gives 348.</li>\n</ul>\n<p>So a few pages on the first day hardly move the figure, and a few hours of your own reading outweigh the default.</p>\n<p>Stats uses the samples from the reading records that start in the range. If there aren't any, the figure shows “—” with the line “Not enough reading this week” (“this month” and “this year” for the other ranges, and “this time” under All time).</p>\n<h2 id=\"what-the-speed-is-used-for\">What the speed is used for</h2>\n<p>It turns words into time. Under the pages, the reader shows what's left of the chapter, for example “18 min left in chapter”. The book screen estimates the rest of the book from its word count and the share you haven't read yet, for example “about 27 h”. A finished book shows no time left.</p>\n<p>The word counts behind these estimates are Readwyn's own, made on your Mac with no online service.</p>\n<h2 id=\"what-stays-out\">What stays out</h2>\n<p>Comics have pages rather than words. Their time still counts as time read, but their samples stay out of the words-per-minute figure, and they get a speed in <a href=\"/blog/reading-stats-for-comics/\">pages per minute</a> instead. Kobo time gives no samples either, since it arrives as one total per book, and private reading records nothing at all.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/reading-stats-across-libraries/","url":"https://readwyn.com/blog/reading-stats-across-libraries/","title":"Reading stats across all your libraries","summary":"The “All libraries” switch in Readwyn's Stats adds up the reading time, finished books, and streak of all your libraries together.","content_html":"<p>Stats has a scope switch at the top right, with the current library's name on one side and “All libraries” on the other. Pick “All libraries” and every figure on the screen counts across all of them. The switch only appears once you have more than one library; Settings › Libraries › “Create library…” makes another.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-stats.webp?h=6be17290a8c05ed4461f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-stats.webp?h=4b431a86ee04dd47470d\" alt=\"Reading stats for October 2026, with the range Week, Month, Year, and All time at the top, and Home library and All libraries at the top right\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The scope at the top right: one library or all of them</figcaption>\n</figure>\n<h2 id=\"what-changes-with-the-scope\">What changes with the scope</h2>\n<p>Whichever range you pick, “All libraries” changes these figures:</p>\n<ul>\n<li><strong>Time read</strong> adds up the time from Readwyn's reader in every library.</li>\n<li><strong>Books finished</strong> counts finished reads everywhere, and the covers mix the books of all your libraries, latest first.</li>\n<li><strong>Streak</strong> counts a day once your reading across the libraries comes to 5 minutes.</li>\n<li><strong>By hour of the day</strong> and <strong>By day of the week</strong> add the time of every library.</li>\n<li><strong>Top authors, genres, and series</strong> add up the time and the books from every library.</li>\n</ul>\n<p>Clicking a book opens its book screen in its own library. Clicking an author, a genre, or a series opens the library filtered to it.</p>\n<p>Each library still keeps its own books, reading data, and reading speed; the scope only adds them up for display. A <a href=\"/blog/year-in-books/\">year in review</a> that you open from Stats counts the same scope.</p>\n<h2 id=\"libraries-on-another-mac\">Libraries on another Mac</h2>\n<p>“All libraries” means every library of the Mac you're browsing. If the app shows another Mac's libraries through Readwyn server, the scope adds up that Mac's libraries, not the ones on your own Mac. Stats needs Read access, so a Mac paired with Guest access has no Stats in its sidebar.</p>\n<h2 id=\"fast-with-a-large-library\">Fast with a large library</h2>\n<p>Readwyn keeps the reading data of each library in memory until that library changes. Switching the range or the scope then recounts from memory without going back to the database, and the combined view of all libraries is kept the same way.</p>\n<p>The target is 100 ms per count, even with 60,000 books. In a test library of 60,000 books with 20,000 finished reads over 15 years, a release build on an M5 Mac counted each range in 1.5 to 6.5 ms (22 October 2026, measured in the backend without drawing the screen). The first count after a change took 50 ms, and a year in review 11 ms.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/reading-stats-for-comics/","url":"https://readwyn.com/blog/reading-stats-for-comics/","title":"Reading stats for comics, which have no word count","summary":"How Readwyn counts the time you spend on a comic, estimates the time left in pages per minute, and keeps comics out of your reading speed in words.","content_html":"<p>Time in the page reader counts just as <a href=\"/blog/track-reading-time-automatically/\">time in the ebook reader</a> does: it adds to the comic's time read and to Stats. What changes is the unit. With no words to count, Readwyn works in pages, and the time left is the pages still to go at a speed in pages per minute that it learns from the comics in your library.</p>\n<h2 id=\"how-the-time-read-is-counted\">How the time read is counted</h2>\n<p>Readwyn times the gaps between page turns, and each new spread on the screen counts as a turn.</p>\n<ul>\n<li>A gap of 5 minutes or less counts in full.</li>\n<li>A longer gap counts for nothing, since you'd most likely stepped away.</li>\n<li>The final gap, from the last turn to closing the comic, follows the same rules.</li>\n</ul>\n<p>The time is saved when you close the comic and after every 10 minutes of reading, so a crash costs you 10 minutes at most. Private reading records no time, for comics as for ebooks.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-comic-book.webp?h=3c139e32e63e50aaf717\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-comic-book.webp?h=22ee8ea9dd9bc59a3077\" alt=\"The book screen of the comic The Lantern Keeper: 40% read, p. 5 of 10, 9 min time read, about 6 min left, and a rating of four stars\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The time read and the time left of a comic</figcaption>\n</figure>\n<h2 id=\"your-speed-in-pages-per-minute\">Your speed in pages per minute</h2>\n<p>A gap where you moved forward measures your speed if it works out at 0.5 to 30 pages a minute. Gaps outside that range still count as time read; they just don't feed the speed.</p>\n<p>Your comics speed is <code>(P + 2 × 10) / (T + 10)</code> pages per minute, where <code>P</code> is the pages and <code>T</code> the minutes of every comic sample in the library. With no samples yet, it's 2 pages a minute. That default weighs as much as 10 minutes of reading, so your first few pages nudge it, and hours of reading settle it.</p>\n<p>Comic samples never touch your <a href=\"/blog/reading-speed-words-per-minute/\">speed in words per minute</a>, and ebook samples never touch your speed in pages.</p>\n<h2 id=\"the-time-left\">The time left</h2>\n<p>The footer of the page reader shows it, for example “6 min left”. It's the pages from the current one to the end, at your comics speed. Comics have no chapters, so this is always the time to the end of the book, and the last page or spread shows no time at all.</p>\n<p>The book screen gives a time left as well, for the pages you haven't read yet at the same speed. A finished comic doesn't have one.</p>\n<h2 id=\"comics-in-stats\">Comics in Stats</h2>\n<ul>\n<li><strong>Time read.</strong> Time in the page reader counts in every range (Week, Month, Year, All time) and in the charts by hour and by weekday.</li>\n<li><strong>Books finished.</strong> A finished comic counts as a finished book.</li>\n<li><strong>Pages read.</strong> A finished comic adds its exact page count.</li>\n<li><strong>Words.</strong> Comics have none. The line under Pages read only counts the books that have a word count, for example “391,000 words in 3 of 4 books”.</li>\n<li><strong>Reading speed.</strong> In words per minute, so comics don't appear in it.</li>\n</ul>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/reading-status-and-shelves/","url":"https://readwyn.com/blog/reading-status-and-shelves/","title":"How reading status and shelves differ","summary":"A Readwyn book has one reading status but any number of shelves. Statuses get tabs and shelves are filters; both sync with a Kobo and Goodreads.","content_html":"<p>A book in Readwyn has one reading status at most, and it can sit on any number of shelves. The status says where you are with it: To read, Next up, Reading, Read or Did not finish. Shelves are groups you name yourself, like “Favourites” or “Book club”. Statuses get tabs over the library; shelves live in the Filter menu.</p>\n<h2 id=\"statuses\">Statuses</h2>\n<p>A status is exclusive: set a book to Read and it leaves Reading. Each tab shows its count, “All 412”, “Reading 3” and so on, and the counts follow the other filters and the search. With “Author: George Eliot” on, “All 4” and “Read 1” count only her books.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-library-list.webp?h=8ef4e13c111a8ae0fdcd\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-library-list.webp?h=0051b07c7f0ee42ce4d5\" alt=\"The library list with the status tabs All, Reading, Next up, To read, Read, and Did not finish over it, and a Status column with Reading for three books\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The status tabs, and the status of each book in the list</figcaption>\n</figure>\n<p>To change a status, click the status button on the book screen, for example “Status: Reading”. The menu lists each status, then “No status”. For a book you haven't finished, “Read” carries a note such as “Finished today, 17 Oct”: choosing it closes the current read with today's date. “No status” takes the book out of every status tab and keeps its progress.</p>\n<p>Often you won't need the menu at all. When the Readwyn reader records progress, a book with no status, To read or Next up moves to Reading by itself.</p>\n<p>Statuses also drive the reading history: starting a book opens a new read, and finishing it closes that read with the date (<a href=\"/blog/track-rereads/\">how rereads are tracked</a>).</p>\n<p>For a batch of books, select them and click “Status” in the selection bar, which opens “Status of 6 books”.</p>\n<h2 id=\"shelves\">Shelves</h2>\n<p>The Overview tab of the book screen lists the book's shelves, with “Add to shelf”. A shelf has nothing to do with the status, so one book can be on “Favourites”, “Book club” and “Victorian” and still be Reading.</p>\n<p>With several books selected, “Add to shelf” in the selection bar opens “Add 3 books to a shelf”, with a checkbox for each shelf. If only some of the books are on a shelf, it shows a bar and a count such as “1 of the 3 books”. “New shelf” and “Add” put them all on a new one.</p>\n<p>In the <a href=\"/blog/filter-library-by-language-series-author-format/\">Filter menu</a>, shelves are listed with their counts. Pick more than one and a book matches if it's on any of them.</p>\n<h2 id=\"kobo-and-goodreads\">Kobo and Goodreads</h2>\n<p>A Kobo sync carries both. A book that is Reading in the library is in progress on the Kobo, and Read is finished. To read, Next up and Did not finish are all unread there, and a book that is unread on the Kobo comes back as To read. Each library shelf becomes a Kobo collection of the same name; a device without collections carries no shelves. The <a href=\"/blog/sync-kobo-with-mac-without-calibre/\">Kobo sync guide</a> walks through a whole sync.</p>\n<p>Goodreads sync carries shelves and status too, but only an exclusive Goodreads shelf sets the status. A normal shelf whose name looks like a status, such as “next-up”, is skipped: the sync neither sets the status from it nor removes it.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/reading-time-from-two-macs/","url":"https://readwyn.com/blog/reading-time-from-two-macs/","title":"Reading time from two Macs, counted once","summary":"How Readwyn measures the time you read on two paired Macs in one place, with no second total on the laptop, and what happens to it when the connection drops.","content_html":"<p>When you read on two paired Macs, Readwyn measures all your reading time in one place: on the Mac that holds the libraries, say your desktop. The reader on your laptop only tells the desktop which page it's showing, and the desktop does the timing. The laptop keeps no total of its own, so there's no second figure to add on top.</p>\n<h2 id=\"one-figure-for-each-book\">One figure for each book</h2>\n<p>A book has one time read, and its book screen shows the same figure on <a href=\"/blog/share-ebook-library-between-two-macs/\">either Mac</a>. Reading on the desktop itself and reading on the laptop add to that same figure.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book.webp?h=6c9b53f864b64d2872b9\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book.webp?h=1970bca0dae546ce847c\" alt=\"The book screen of Middlemarch: 22 percent read, about page 196 of 880, 7 h 40 min time read, about 27 h left, and a rating of four stars\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Time read is one figure for the book, on every Mac</figcaption>\n</figure>\n<h2 id=\"what-counts\">What counts</h2>\n<p>Time is measured between page turns. A gap of 5 minutes or less counts in full. A longer one counts for nothing, since you'd most likely left the book. The stretch from the last page to closing the book follows the same rule.</p>\n<p>Each Mac keeps its own reading session for a book. Closing the book on the laptop leaves the desktop's session open, and the other way round. A book you left open on the desktop adds nothing while you read on the laptop: after 5 minutes without a page turn there, that whole stretch is dropped.</p>\n<p>The time is saved when you close the book, and every 10 minutes of reading while it stays open, so a crash costs you 10 minutes at most.</p>\n<p>Private reading never counts, because it saves no position, and that covers everything read on a <a href=\"/blog/read-guest-admin-access-levels/\">Guest Mac</a>.</p>\n<h2 id=\"when-the-connection-drops\">When the connection drops</h2>\n<p>If the laptop leaves the network without closing its books, the desktop notices within about 15 seconds, ends the laptop's sessions, and saves their time. The time up to the drop is kept, and nothing is added for the time the laptop was gone.</p>\n<p>A short drop needn't stop you reading. Once the connection is back, the session starts again on the desktop, so the time after the reconnect counts and the outage still doesn't.</p>\n<p>Reading from an <a href=\"/blog/keep-book-offline-on-laptop/\">offline copy</a> while the desktop is out of reach records no time. Your page turns wait on the laptop and reach the desktop later, in one burst. They move your place in the book but add nothing to your time.</p>\n<h2 id=\"stats-on-either-mac\">Stats on either Mac</h2>\n<p>Stats shows the time from Readwyn's own reader, and says so: “Time from the Readwyn reader”. Both Macs count the same records. Only one thing depends on the Mac showing the stats, and that's its time zone. A day starts at midnight in that Mac's time zone, even when the libraries live on the other Mac.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-stats.webp?h=6be17290a8c05ed4461f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-stats.webp?h=4b431a86ee04dd47470d\" alt=\"Reading stats for October 2026: 21 h 46 min read, a bar for each day, a daily average of 59 min, and the line Time from the Readwyn reader\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>Stats count the same records on every Mac</figcaption>\n</figure>\n<p>Your reading speed comes from the same records: the words you read forward in each counted gap, as long as the pace falls between 50 and 1,000 words per minute. Pages read on either Mac feed one speed, and the <a href=\"/blog/how-long-to-read-a-book/\">time left</a> of each book uses it.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/save-looked-up-words/","url":"https://readwyn.com/blog/save-looked-up-words/","title":"Save the words you look up and review them later","summary":"Readwyn saves a looked-up word to My Words in one click. The Notebook then lists your words by headword and by book, and a Kobo sync can share them.","content_html":"<p>Words you save while reading go to the Notebook, under “My Words”, each with the book it came from. Saving one takes a click on “Save to My Words” at the foot of the <a href=\"/blog/look-up-words-in-ebooks-on-mac/\">lookup popover</a>, and a second click on the same button takes it out again.</p>\n<h2 id=\"save-a-word\">Save a word</h2>\n<p>Nothing pops up when you save. The only sign is the button itself, which turns into “In My Words” with a filled bookmark, and its tooltip changes to “Remove from My Words”.</p>\n<p>There’s no button for a word the dictionaries have no entry for, or anywhere in <a href=\"/blog/read-a-book-privately/\">private reading</a>. When the popover shows two readings, such as “gave up” and “gave”, each one has its own button and saves its own text.</p>\n<p>Readwyn keeps the word as it’s printed, so “forlornness” stays “forlornness” rather than becoming the headword “forlorn”. Case doesn’t matter. Once you’ve saved “Forlornness”, the button already reads “In My Words” when you meet “forlornness” on a later page or in another book, and My Words keeps the capitalisation, the book and the date of your first save.</p>\n<h2 id=\"review-your-words\">Review your words</h2>\n<p>Click “Notebook” in the sidebar, then “My Words” in the toolbar. Words are listed newest first and grouped by headword, so “forlornness” and “forlorn” share one group under “forlorn”. Until you save something, the view says “No words yet”.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-notebook-words.webp?h=cd3bb9313f6f7a11a159\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-notebook-words.webp?h=b4ba5a2db81a6383cd4a\" alt=\"My Words in the Notebook: All books and four books at the left, and word groups at the right, with the full entry of forlorn open and two saved terms under it\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>One group for each headword, with the book and the date of each saved term</figcaption>\n</figure>\n<p>Each group shows the headword and its pronunciation, the first sense (two lines at most), and every saved term with the book it came from. The date is on the right, such as “9 Oct”, or “12 Mar 2025” for an earlier year.</p>\n<p>Click a group to open its full entry in place, with a speak button after the pronunciation and the dictionary’s name at the bottom. Click again to close it. Hover over a group and “Remove” appears (“Remove 2 words” if it holds two terms). It removes every term in the group.</p>\n<p>A saved word that no dictionary has an entry for gets a group of its own, which says “No entry in the dictionaries on this Mac.” On Linux, Wiktionary is the only dictionary. Until you download it, the popover has no entry to show but still offers “Save to My Words”, and a word saved then shows that line (with “this computer” for “this Mac”). It picks up its entry once Wiktionary is ready.</p>\n<h2 id=\"words-by-book\">Words by book</h2>\n<p>The column on the left starts with “All books” and your total, then lists every book you’ve saved words from, each with its cover and count. Click a book to see only its words.</p>\n<p>“Search my words” narrows the list as you type, matching both the saved terms and the headwords. “Copy all” puts every headword on the clipboard with its first sense and its terms.</p>\n<h2 id=\"where-the-words-live\">Where the words live</h2>\n<p>My Words belongs to the app rather than to one library, so “All books” counts words from every library. If you delete a book, or a whole library, its words stay, just without a book.</p>\n<p>No definitions are stored. Each time the view shows a word, Readwyn looks it up again in your dictionaries, in the <a href=\"/blog/dictionary-order/\">order you’ve set for them</a>.</p>\n<h2 id=\"with-a-kobo\">With a Kobo</h2>\n<p>A Kobo keeps its own My Words list, and a <a href=\"/blog/kobo-my-words-on-mac/\">Kobo sync</a> copies words between the two lists in whichever direction you run it. It only covers books that are both on the Kobo and in the library. Before anything changes, the review sheet lists each word under its book, and a word whose book is missing on the other side goes under “Left out”, with the reason.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/send-books-to-kobo-from-mac/","url":"https://readwyn.com/blog/send-books-to-kobo-from-mac/","title":"How to send books to a Kobo from a Mac","summary":"In Readwyn, “Send to device” in the “…” menu of a book copies it to a plugged-in Kobo, as a KEPUB unless you turn the conversion off.","content_html":"<p>A sync never adds books to a Kobo, so there’s one way to put a book on it: plug the Kobo in, open the book in Readwyn, and pick “Send to device” from the “…” menu on the book screen. The book goes over as a KEPUB unless you’ve turned that off.</p>\n<h2 id=\"send-a-book\">Send a book</h2>\n<ol>\n<li>Connect the Kobo with a USB cable. A card for it shows up in the sidebar.</li>\n<li>Open the book from your library.</li>\n<li>Click “…” (“More actions”) on the book screen. “Send to device” is in the menu, with the name of the connected Kobo under it.</li>\n<li>Click it.</li>\n</ol>\n<p>For a book that’s already on the Kobo, the item says “Send to device again”.</p>\n<h2 id=\"what-arrives-on-the-kobo\">What arrives on the Kobo</h2>\n<p>The book lands in the Kobo’s <code>Books</code> folder, named after its first author and title:</p>\n<ul>\n<li><code>Books/&lt;first author&gt; - &lt;title&gt;.kepub.epub</code> for a KEPUB,</li>\n<li><code>Books/&lt;first author&gt; - &lt;title&gt;.epub</code> if you’ve turned the conversion off,</li>\n<li><code>Books/&lt;first author&gt; - &lt;title&gt;.cbz</code> for a comic.</li>\n</ul>\n<p>“Convert EPUB to KEPUB for Kobo e-readers” in Settings › E-readers controls the conversion, and it’s on by default. It wraps each sentence in a <code>koboSpan</code> and leaves the text itself alone. With it off, Readwyn copies the EPUB as it is. One Kobo can also have its own setting: “Convert EPUB to KEPUB when sending”, under “Options for this device” on the Sync screen. <a href=\"/blog/kepub-or-epub-for-kobo/\">KEPUB or EPUB</a> covers which one to pick.</p>\n<p>Readwyn converts a book in an older format, such as MOBI, AZW3, or PDF, to EPUB 3 when it imports it. Where that conversion worked, “Send to device” sends the EPUB 3. The original stays in the library untouched.</p>\n<p>Comics always go as CBZ, never as KEPUB. A CBZ is copied as it is. For a CB7 or CBT, Readwyn already keeps a CBZ copy to read it from, with page images byte for byte the same as the original’s, and that copy is what goes to the Kobo.</p>\n<p>The first sync or send with a Kobo also leaves one small file on it, <code>.readwyn/device-id</code>, which is how Readwyn recognises this Kobo later (<a href=\"/blog/how-readwyn-writes-to-kobo-safely/\">more on how Readwyn writes to a Kobo</a>).</p>\n<h2 id=\"why-the-book-isn-t-on-the-kobo-yet\">Why the book isn’t on the Kobo yet</h2>\n<p>Sending a book only copies the file. The Kobo adds it to its own list of books when it leaves USB mode, so eject it with “Eject device” on the Sync screen.</p>\n<p>If you click it while Readwyn is still writing, an alert says “Readwyn is still writing to the Kobo”. “Wait” closes the alert. “Stop and eject” gives the write up to 30 seconds to finish and then ejects. If the write is still going after that, the Kobo stays connected and you can try again once it’s done.</p>\n<h2 id=\"check-it-or-take-it-off-again\">Check it, or take it off again</h2>\n<p>On the book screen’s Details tab, the row “On the device” shows the copy on the connected Kobo: its format, the device, and the date you sent it.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book-details.webp?h=3f5c56cb7a1e9927112f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book-details.webp?h=647f1b629b6b477420d1\" alt=\"The Details tab of Middlemarch: the original MOBI, the EPUB 3 converted from it, and a KEPUB on a Kobo Clara 2E, sent 2 Oct 2026, with the button Remove from device\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The Details tab: the original, the EPUB 3, and the copy on the Kobo</figcaption>\n</figure>\n<p>“Remove from device” in that row deletes the file from the Kobo. Nothing else in Readwyn deletes a book from a device. The book itself stays in your library, along with its highlights and reading data, and the Kobo drops it from its list the next time it leaves USB mode.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/share-ebook-library-between-two-macs/","url":"https://readwyn.com/blog/share-ebook-library-between-two-macs/","title":"Share one ebook library between two Macs","summary":"Two Macs can use one Readwyn library with nothing to sync: one Mac keeps the books and reading data, and the other opens them over your home network.","content_html":"<p>One Mac keeps the books, your progress, your highlights, and My Words. The other pairs with it and opens those same libraries over your home network. Because they live in one place, there's nothing to sync between the two Macs.</p>\n<p>To set it up, turn on “Readwyn server” in Settings › Other devices on the Mac that keeps the books and choose your network in “Listen on”. Then <a href=\"/blog/pair-second-mac/\">pair the second Mac</a> once. The server keeps going after you quit Readwyn and starts again when you log in.</p>\n<p>Once paired, the second Mac opens with the first Mac's libraries. Its library switcher names the other Mac, says whether it's connected, and shows this Mac's level. The libraries the second Mac had before aren't touched, and “Use the libraries on this Mac” takes you back to them at any time. You see one Mac's libraries at a time; Readwyn never merges two.</p>\n<h2 id=\"what-both-macs-share\">What both Macs share</h2>\n<p>Either Mac shows the same books, covers, and details. Your progress and the status of each book are the same on both, and so are your highlights, notes, My Words, time read, and reading stats.</p>\n<p>Each Mac checks for changes every 2 seconds, so a highlight you make on one shows up on the other almost at once. An open book or an open edit sheet isn't reloaded until you close it.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-notebook.webp?h=c5d907fb6a66e4603293\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-notebook.webp?h=758885220be4b17df8b7\" alt=\"The Notebook: the books with highlights at the left, and the highlights of Middlemarch at the right, grouped by chapter, with their colours, pages, and a note\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The Notebook lists the same highlights on either Mac</figcaption>\n</figure>\n<p>Some things belong to the Mac in front of you rather than to the library. The theme and the reader settings are per Mac, and so are the dictionaries and their order (each Mac downloads Wiktionary for itself). A spoken word uses the voices installed on the Mac you're reading on.</p>\n<p>What the second Mac may change depends on <a href=\"/blog/read-guest-admin-access-levels/\">its level</a>, which you set on the first Mac: Guest, Read, or Admin. Read is the default. A Read Mac saves your place and your highlights and can sync a Kobo plugged into it, but adding, removing, or editing books needs Admin.</p>\n<h2 id=\"limits\">Limits</h2>\n<p>Both Macs have to be on the same home network. There's no access from outside it and no hosted service: your library stays on your own Mac. If that Mac can't be reached, the second one shows the last lists it saw and the books you <a href=\"/blog/keep-book-offline-on-laptop/\">kept with “Available offline”</a>.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/sync-kobo-on-linux/","url":"https://readwyn.com/blog/sync-kobo-on-linux/","title":"How to sync a Kobo on Linux","summary":"Readwyn on Linux syncs a Kobo as it does on a Mac. It finds the Kobo where udisks mounts it and backs up the Kobo’s database to ~/.local/share/readwyn.","content_html":"<p>Plug the Kobo in with a USB cable and let your desktop mount it, as it would any USB drive. Readwyn finds it where udisks puts it, the sidebar shows the Kobo card, and from there a sync works as it does on a Mac: open “Sync”, review a direction, apply.</p>\n<p>The app runs on x86_64 and arm64. The <a href=\"/download/#linux\">download page</a> has the deb and rpm packages, and <a href=\"/blog/install-readwyn-on-ubuntu/\">Install Readwyn on Ubuntu</a> and <a href=\"/blog/install-readwyn-on-fedora/\">Install Readwyn on Fedora</a> go through the steps.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/linux/dark-library.webp?h=02876295485e7080879c\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/linux/light-library.webp?h=24919ab2ffd9d07527ff\" alt=\"The library of Readwyn on Linux, with the window controls at the right of the top bar and the card Kobo Clara 2E, Connected · 13 changes, Review changes in the sidebar\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The Kobo card on Linux</figcaption>\n</figure>\n<p>The card gives the Kobo’s name and how many changes are waiting in both directions, for example “Connected · 13 changes”, and “Review changes” opens the Sync screen. A normal session is Device → Readwyn, then Readwyn → Device, and each direction <a href=\"/blog/kobo-sync-preview/\">shows its changes</a> before it writes any. Books go onto the Kobo with “Send to device” on the book’s page, <a href=\"/blog/send-books-to-kobo-from-mac/\">as on a Mac</a>.</p>\n<h2 id=\"if-the-kobo-doesn-t-show-up\">If the Kobo doesn’t show up</h2>\n<p>Readwyn looks in the folders where udisks mounts removable devices:</p>\n<table><thead><tr><th>System</th><th>Folder</th></tr></thead><tbody>\n<tr><td>Ubuntu, Debian</td><td><code>/media/&lt;user&gt;</code></td></tr>\n<tr><td>Fedora</td><td><code>/run/media/&lt;user&gt;</code></td></tr>\n</tbody></table>\n<p><code>&lt;user&gt;</code> is your login name. If the card doesn’t appear, check that the desktop actually mounted the Kobo in one of these folders.</p>\n<p>There’s no sandbox on Linux, so unlike the Mac App Store version, Readwyn doesn’t ask you to grant access to the Kobo.</p>\n<h2 id=\"where-the-backups-go\">Where the backups go</h2>\n<p>Before each write, Readwyn copies the Kobo’s database into its app folder. On Linux that folder is <code>$XDG_DATA_HOME/readwyn</code>, which is <code>~/.local/share/readwyn</code> unless you’ve changed it, so the backups end up in <code>~/.local/share/readwyn/kobo-backups/</code>. The “Backups” section of Settings › E-readers has a button that opens it. The backup lands there even when your libraries live on another computer, such as a Mac or a headless <code>ryn-server</code>: that computer runs the sync on <a href=\"/blog/how-readwyn-writes-to-kobo-safely/\">a copy of the Kobo’s files</a>, and the Linux machine with the Kobo backs up the database and writes the changes.</p>\n<p>Where the Mac app says “this Mac”, the Linux app says “this computer”, so the backup note there reads “Before every write, Readwyn saves a copy of the e-reader’s database on this computer.”</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/sync-kobo-with-mac-without-calibre/","url":"https://readwyn.com/blog/sync-kobo-with-mac-without-calibre/","title":"How to sync a Kobo with a Mac without calibre","summary":"Sync a Kobo with a Mac in Readwyn: plug it in over USB, run Device → Readwyn and Readwyn → Device, and review every change before it is written.","content_html":"<p>Plug the Kobo into your Mac with a USB cable and Readwyn opens its Sync screen with the changes in each direction already counted. It reads the Kobo’s own database, so calibre doesn’t need to be installed at all. If you came from calibre, your library is imported once, and calibre’s <code>metadata.db</code> is never read after that.</p>\n<h2 id=\"when-you-plug-it-in\">When you plug it in</h2>\n<p>Plugging the Kobo in only reads it. Nothing is written until you apply a sync. A card for the Kobo appears in the sidebar with a count of pending changes, for example “Connected · 13 changes”, and its “Review changes” button opens the Sync screen.</p>\n<p>At the top of that screen is a drawing of your model, its name, “Connected over USB”, and a line with the number of books on it, the free space, the firmware version, and the date of your first sync.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-sync.webp?h=3748d9769c62e82b8755\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-sync.webp?h=25b1a7e75f31a1231622\" alt=\"The Sync screen of a Kobo Clara 2E: the device, the cards Device → Readwyn with 8 changes and Readwyn → Device with 5 changes, and the books that are not in the library\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The Sync screen: one card for each direction</figcaption>\n</figure>\n<p>If you’d rather the Sync screen didn’t open by itself, set “When an e-reader connects” in Settings › E-readers to “Stay where I am”. The sidebar card still shows the Kobo and its changes. In the Mac App Store build, the first connection also asks you to <a href=\"/blog/kobo-access-mac-app-store/\">give Readwyn access to the Kobo</a>.</p>\n<h2 id=\"two-directions\">Two directions</h2>\n<p>Every sync runs one way, and Readwyn is always one end of it. Under “Choose what to sync” there are two cards, each with its own count. Both directions carry the same things; only the source changes. “Device → Readwyn” copies the Kobo’s values into the library, and “Readwyn → Device” copies the library’s values to the Kobo.</p>\n<p>Where the two sides disagree, the receiving side ends up with the sender’s value. So the usual session is Device → Readwyn first, then Readwyn → Device, and the post on <a href=\"/blog/sync-reading-progress-kobo-and-mac/\">reading progress</a> shows what happens the other way round. You can also run just one.</p>\n<h2 id=\"review-then-apply\">Review, then apply</h2>\n<p>The card’s button, for example “Review 8 changes”, doesn’t apply anything yet: it opens a <a href=\"/blog/kobo-sync-preview/\">review sheet</a> that lists every change book by book, each row with its own checkbox. Most rows start ticked. Two kinds don’t: one whose receiving side was changed more recently, and a progress or time-read row that would move the Kobo back. Untick anything you don’t want and click the main button, say “Apply 7 changes”.</p>\n<p>Afterwards a message gives the count, for example “Applied 7 changes from Kobo Clara 2E.” After a Readwyn → Device sync, which writes to the Kobo, it adds “The device database is backed up.”</p>\n<p>When you’re done, click “Eject device” at the top of the Sync screen. Readwyn won’t eject the Kobo in the middle of a write.</p>\n<h2 id=\"what-gets-synced\">What gets synced</h2>\n<p>A sync only touches books that are both in your library and in the Kobo’s database. For each of them it carries:</p>\n<ul>\n<li>highlights and notes, with their colours,</li>\n<li>status, progress with the last-read date, rating, and time read,</li>\n<li>shelves, which become Kobo collections of the same name,</li>\n<li>the words of My Words.</li>\n</ul>\n<p>It never adds a book to the Kobo and never deletes one. Putting a book on the Kobo is a separate step, “Send to device” on the book screen (<a href=\"/blog/send-books-to-kobo-from-mac/\">sending books to a Kobo</a>).</p>\n<h2 id=\"books-that-calibre-put-on-the-kobo\">Books that calibre put on the Kobo</h2>\n<p>If you used calibre to load your Kobo, there’s a file called <code>metadata.calibre</code> on it. Readwyn uses that file to work out which book on the Kobo is which book in your library: the calibre import kept each book’s calibre UUID for exactly this. If the file is missing or damaged, Readwyn falls back to matching on the title and first author.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/sync-reading-progress-kobo-and-mac/","url":"https://readwyn.com/blog/sync-reading-progress-kobo-and-mac/","title":"Sync your reading progress between a Kobo and a Mac","summary":"Run Device → Readwyn before Readwyn → Device. A sync copies the sender’s progress even when it’s behind, so the wrong order can move the Kobo back.","content_html":"<p>Sync in two steps, in this order: Device → Readwyn first, so the percent read and last-read date of each book come from the Kobo into your library, then Readwyn → Device, so whatever you read on the Mac goes back. Do it the other way round and the Kobo can jump back to wherever the Mac last left the book.</p>\n<h2 id=\"what-travels-with-progress\">What travels with progress</h2>\n<p>For every book that’s on the Kobo and in your library, a sync carries:</p>\n<ul>\n<li><strong>Progress</strong>, with the last-read date. The Kobo keeps it as a whole percent.</li>\n<li><strong>Status.</strong> “Reading” becomes reading on the Kobo and “Read” becomes finished. “To read”, “Next up”, and “Did not finish” all go over as unread, and an unread Kobo book comes back as “To read”.</li>\n<li><strong>Time read</strong>, the time you spent in the book (more in <a href=\"/blog/kobo-time-read/\">Kobo time read</a>).</li>\n<li><strong>Rating.</strong> Readwyn has half stars and the Kobo whole ones, so half a star rounds up.</li>\n</ul>\n<p>A book you’ve never opened on the Kobo has nothing to send to the library. Going the other way, it counts as unread at 0%, so any progress the library has goes over as a change.</p>\n<h2 id=\"why-the-order-matters\">Why the order matters</h2>\n<p>A sync doesn’t look for the furthest point. It copies, so where the two sides differ, the sending side’s value replaces the other one, even when the sender is behind. So if the library says 18% and the Kobo says 22%, Readwyn → Device moves the Kobo back to 18%.</p>\n<p>The review sheet is there to catch that. It shows each value with both sides, and on the Device → Readwyn tab the same book appears as “Update · Progress” with “18% → 22%”. Over on Readwyn → Device, the row warns you in plain words: “The Kobo is ahead. This moves it back to 18%. Run Device → Readwyn first to keep 22%.” A row whose receiving side changed more recently shows when, for example “changed here later, on 9 Oct”. Both kinds start unticked, so nothing goes backwards unless you tick it.</p>\n<p>Once Device → Readwyn is applied, its card says “Up to date” and the Readwyn → Device card counts again. That card also names the books whose progress or time read just came forward from the Kobo, so you can see they won’t be moved back.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-sync-done.webp?h=9a3e0d60ef8718c47169\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-sync-done.webp?h=4d9602c3976b420b2d79\" alt=\"The Sync screen after Device → Readwyn applied 7 changes. The card Readwyn → Device says: Counted again after the last sync: Middlemarch no longer moves back\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>After Device → Readwyn, the other direction counts again</figcaption>\n</figure>\n<p>To stop the Kobo overwriting books you’ve finished, turn on “Keep finished books”. Device → Readwyn then leaves alone any book the library shows as 100% read: its status, progress, and time read. It can be set for one Kobo under “Options for this device”, for every device in Settings › E-readers, or for a single review in the review sheet.</p>\n<h2 id=\"progress-from-the-mac\">Progress from the Mac</h2>\n<p>Readwyn’s own reader records how far you’ve got as a whole percent of the book’s text, with the date. A book with no status, “To read”, or “Next up” turns into “Reading” at that point. Private reading records nothing, so there’s nothing new for a sync to send.</p>\n<p>You can also set progress without opening the book. On the book screen’s Details tab, “Your reading” has a percent field and a “Set” button. The note under it says the reader will open at the start of that page. For a book that’s on the plugged-in Kobo, it adds that the Kobo gets the new value at the next Readwyn → Device sync.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/track-reading-time-automatically/","url":"https://readwyn.com/blog/track-reading-time-automatically/","title":"Track your reading time automatically","summary":"Readwyn counts every gap of up to 5 minutes between page turns in its reader, and shows the total on the book screen and in Stats.","content_html":"<p>There's no timer to start or stop. Open a book in Readwyn's reader and turn pages as you normally would. Any gap of up to 5 minutes between two page turns counts as reading, and a longer one counts for nothing. The total shows on the book's screen as “time read”, and Stats charts it by day and by hour.</p>\n<h2 id=\"how-a-session-is-counted\">How a session is counted</h2>\n<p>A session runs from the moment you open a book in the reader until you close it. Each new spread the reader shows is a page turn, and Readwyn times the gaps between them. A gap over 5 minutes usually means you were away, or the window was in the background, so none of it is kept. The last stretch, from the final page turn to closing the book, follows the same rule. Times are rounded to whole seconds.</p>\n<p>Closing the book saves the session. If you keep reading, Readwyn also saves after every 10 minutes of counted reading, so a crash can't cost you more than 10 minutes. Each record keeps the moment it started, which is how Stats knows the day and the hour.</p>\n<h2 id=\"where-you-see-it\">Where you see it</h2>\n<p>On the book screen, the time read sits between your estimated page and the estimate of what's left.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book.webp?h=6c9b53f864b64d2872b9\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book.webp?h=1970bca0dae546ce847c\" alt=\"The book screen of Middlemarch: 22% read, about page 196 of 880, 7 h 40 min time read, about 27 h left, and a rating of four stars\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The time read and the time left on the book screen</figcaption>\n</figure>\n<p>Stats adds it up for the range you choose at the top of the screen, with a bar for each day of a week or a month and a bar for each month of a year. “Time from the Readwyn reader”, at the right of the range's title, tells you where the bars come from.</p>\n<p>The same page turns also give your <a href=\"/blog/reading-speed-words-per-minute/\">reading speed</a>: a forward gap that works out at 50 to 1,000 words per minute becomes a sample. Comics follow the same rules for time, though their <a href=\"/blog/reading-stats-for-comics/\">speed and time left work in pages rather than words</a>.</p>\n<h2 id=\"when-nothing-is-recorded\">When nothing is recorded</h2>\n<ul>\n<li><strong>Private reading.</strong> “Read privately” (⌥⌘R) in the “…” menu of the book screen saves no time read, no progress, and no dates.</li>\n<li><strong>A Guest Mac.</strong> Every book opens in private reading there.</li>\n<li><strong>Reading offline.</strong> If the Mac that holds the library can't be reached and the book opens from its copy on this Mac, no time is recorded.</li>\n</ul>\n<h2 id=\"kobo-time-and-other-macs\">Kobo time and other Macs</h2>\n<p>A Kobo keeps its own reading time for each book, and a sync from the device replaces Readwyn's figure with the Kobo's. That time has no date, so Stats only adds it under All time, where the line under the chart says how much of it is undated.</p>\n<p>If you read on a laptop from a desktop Mac's library, the time is measured on the desktop, from the page turns that the laptop's reader sends. Each Mac gets a session of its own, so opening or closing the book on one doesn't end the session on the other.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/track-rereads/","url":"https://readwyn.com/blog/track-rereads/","title":"Track rereads: a history of every time you read a book","summary":"Readwyn keeps a reading history for each book: every read with its dates, its outcome, and its own rating, so a reread does not overwrite the first time you read it.","content_html":"<p>Every time you start a book again, Readwyn opens a new read instead of overwriting the old one. Each read keeps its own dates, how it ended (finished or not) and, if you finished it, the rating you gave it then. You'll find the whole history on the Overview tab of the book screen, newest read first.</p>\n<h2 id=\"where-to-see-it\">Where to see it</h2>\n<p>Open a book and look under “Reading history”, below the description and the shelves. Each line is one read:</p>\n<ul>\n<li>“Reading now” and the day you started: “Since 2 Oct 2026”</li>\n<li>“Finished”, with both dates (“12 Mar – 30 May 2019”) and that read's rating</li>\n<li>“Did not finish”, with its dates</li>\n</ul>\n<p>Another Mac paired with Read or Admin access sees the same history, because the reads belong to the library. A Guest Mac doesn't see it.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book.webp?h=6c9b53f864b64d2872b9\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book.webp?h=1970bca0dae546ce847c\" alt=\"The book screen of Middlemarch. Under Reading history: Reading now, since 2 Oct 2026, and Finished, 12 Mar to 30 May 2019, with four stars\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>A reread of Middlemarch, with the first read and its rating below it</figcaption>\n</figure>\n<h2 id=\"starting-a-reread\">Starting a reread</h2>\n<p>Reads follow the book's status, so there's nothing separate to log. Click the status button on the book screen and choose “Reading”: if the book was finished, that starts its next read and leaves the earlier ones alone. Choosing “Read” ends it as finished. If the book isn't finished yet, the menu shows the date under “Read”: “Finished today, 17 Oct”. “Did not finish” ends the read as abandoned, and an abandoned read gets no rating.</p>\n<h2 id=\"one-rating-for-each-read\">One rating for each read</h2>\n<p>Each finished read keeps the rating the book had when you finished it. Which read a later rating goes to depends on when it arrives:</p>\n<ul>\n<li>While the book stays finished, changing its rating changes the rating of the last read too, so you can finish a book today and rate it tomorrow.</li>\n<li>Once you start a reread, changing the stars leaves your earlier reads alone. The reread takes whatever rating the book has when you finish it.</li>\n<li>If a Goodreads or Kobo sync brings a finish and a rating at the same time, the rating goes to the read it finishes, and your earlier reads keep theirs.</li>\n</ul>\n<p>Setting the rating itself (half stars, ⌥-click) is covered in <a href=\"/blog/ratings-and-reviews-with-spoiler-flag/\">Ratings and reviews with a spoiler flag</a>.</p>\n<h2 id=\"reads-from-before-readwyn\">Reads from before Readwyn</h2>\n<p>A book imported from calibre comes with one read at most. The importer builds it from the book's Goodreads shelf and its Kobo last-read date, if your calibre library has those columns. A finished book arrives with one finished read, dated if there was a last-read date, and with the book's rating. A Goodreads sync adds finished reads with their dates too, so the history can go back years.</p>\n<h2 id=\"rereads-in-your-stats\">Rereads in your stats</h2>\n<p>Stats count reads, not books. Finish a book twice in a year and <strong>Books finished</strong> counts it twice. <strong>Average rating</strong> is the mean over your rated finished reads, so a reread you liked less pulls it down without changing the rating of the first read. In <a href=\"/blog/year-in-books/\">your year in books</a>, the fastest finish is measured from the start and finish dates of each read, and the highest rated book comes from the ratings of the reads.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/update-readwyn-on-linux/","url":"https://readwyn.com/blog/update-readwyn-on-linux/","title":"Update Readwyn on Linux without a package repository","summary":"Readwyn has no apt or dnf repository, so the Linux app checks readwyn.com once a day and shows a card when a new package is ready to go over the old one.","content_html":"<p>A system update won't touch Readwyn, because there's no apt or dnf repository for it. The Linux app looks at readwyn.com itself, at most once a day (the Mac app doesn't check at all). When there's a new version, a card in the sidebar says so, and you install the new package over the old one. Nothing in your libraries or settings changes.</p>\n<h2 id=\"the-card\">The card</h2>\n<p>It sits above your e-reader's card and reads “Readwyn 0.2.0 is out”, with “You have 0.1.0. Your libraries and settings stay.” underneath.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/linux/dark-update-app.webp?h=284d320c96e1f3ca7a1b\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/linux/light-update-app.webp?h=f48b251e0f78b4d51a66\" alt=\"The Readwyn library on Linux; in the sidebar, above the Kobo Clara 2E card, a card says Readwyn 0.2.0 is out, You have 0.1.0. Your libraries and settings stay, with the buttons Download and Later\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The card for a new version</figcaption>\n</figure>\n<p>“Download” opens the download page in your browser. “Later” hides the card until the next version comes out.</p>\n<h2 id=\"installing-the-new-version\">Installing the new version</h2>\n<p>Download the new package for your system and processor, and install it over the old one:</p>\n<table><thead><tr><th>System</th><th>How</th></tr></thead><tbody>\n<tr><td>Ubuntu and Debian</td><td><code>sudo apt install ./readwyn_&lt;version&gt;-1_amd64.deb</code></td></tr>\n<tr><td>Fedora</td><td>Double-click the <code>.rpm</code> file, or <code>sudo dnf install ./readwyn-&lt;version&gt;-1.x86_64.rpm</code></td></tr>\n</tbody></table>\n<p>On arm64, use the arm64 file name. Ubuntu's App Center won't put a new version over an installed one (it shows “Installed” instead), which is why the table gives the <code>apt</code> command: it replaces the old version.</p>\n<p>In testing, a 0.2.0 package went on over 0.1.0, and the settings, the list of libraries and the books came through unchanged.</p>\n<h2 id=\"what-the-check-sends\">What the check sends</h2>\n<p>The app fetches one small file, <code>https://readwyn.com/linux/latest.json</code>, over HTTPS with a plain GET. The only headers it sets itself are <code>User-Agent: Readwyn/&lt;version&gt;</code> and <code>Accept: application/json</code>. There's no query string, body or cookie: no library data goes with the request, and nothing that identifies your computer or you.</p>\n<p>The file gives the latest version of the app and of the server, and the page to download each from. Versions are compared as numbers, not as text, so 0.10.0 counts as newer than 0.9.0.</p>\n<h2 id=\"how-often\">How often</h2>\n<p>Between checks, the app answers from the last copy it read. If readwyn.com can't be reached, it tries again an hour later; if the site answers with an error, the last copy stands for another day.</p>\n<h2 id=\"a-server-on-another-machine\">A server on another machine</h2>\n<p>If the app is connected to a Linux server that runs an older version (<code>ryn-server</code> on a NAS, say), the same card names the server instead: “Readwyn server 0.2.0 is out”, then “Sid’s NAS has 0.1.0. Install the new package there; its libraries stay.” Install the new package on that machine the same way you installed the first one (the file names are in <a href=\"/blog/headless-ebook-server-on-linux/\">Run a headless ebook server on Linux</a>). The package restarts the running server for you.</p>\n<p>The card shows one notice at a time. If both are due, the app's comes first, and the server's appears once you press “Later”.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-is-opds/","url":"https://readwyn.com/blog/what-is-opds/","title":"What is OPDS, and how do reader apps use it?","summary":"OPDS is a catalogue format that reader apps browse to find and download books. Readwyn serves an OPDS 1.2 catalogue of your libraries on your home network.","content_html":"<p>OPDS is a format for catalogues of ebooks. A catalogue is a set of web feeds that a reader app on a phone, a tablet, or an e-reader can browse: lists of authors, series, and books, each book with its cover, its details, and links to download its files. Readwyn serves an OPDS catalogue of your own libraries, so reader apps such as KOReader, Thorium, Moon+ Reader, Panels, and Chunky can browse them and download books over your home network.</p>\n<h2 id=\"two-kinds-of-feed\">Two kinds of feed</h2>\n<p>An OPDS 1.2 catalogue (<a rel=\"external\" href=\"https://specs.opds.io/opds-1.2\">the specification</a>) is made of Atom feeds, and Readwyn’s come in two kinds:</p>\n<ul>\n<li>A <strong>navigation feed</strong> lists other feeds: your libraries, or the authors of a library.</li>\n<li>An <strong>acquisition feed</strong> lists books. Each book entry links to its cover and its files.</li>\n</ul>\n<p>A reader app starts at the root of the catalogue and follows the links. In Readwyn, the root’s address ends in <code>/opds</code>, for example <code>http://192.168.1.20:7314/opds</code>.</p>\n<h2 id=\"readwyn-s-feeds\">Readwyn’s feeds</h2>\n<table><thead><tr><th>Feed</th><th>Kind</th><th>What it lists</th></tr></thead><tbody>\n<tr><td>The root</td><td>navigation</td><td>Your libraries, each with its book count</td></tr>\n<tr><td>A library</td><td>navigation</td><td>Recently added, Currently reading, By author, By series, By shelf, and All books</td></tr>\n<tr><td>Recently added</td><td>acquisition</td><td>The books, newest first</td></tr>\n<tr><td>All books</td><td>acquisition</td><td>The books by title</td></tr>\n<tr><td>By author, By series, By shelf</td><td>navigation</td><td>The authors, series, or shelves, each with a book count</td></tr>\n<tr><td>One author or shelf</td><td>acquisition</td><td>Its books by title</td></tr>\n<tr><td>One series</td><td>acquisition</td><td>Its books by their number in the series</td></tr>\n<tr><td>Search</td><td>acquisition</td><td>The books whose title, authors, or series match</td></tr>\n</tbody></table>\n<p>If you keep one library, the root opens it at once. Currently reading shows up only for a reader app that you’ve allowed to see it.</p>\n<h2 id=\"search-and-pages\">Search and pages</h2>\n<p>Every feed links to itself and to the start of the catalogue, and a feed inside a library links up to that library. Each one also points to the search, in two forms: an OpenSearch description, and a template that puts your words in the address.</p>\n<p>Long feeds are split into pages of 50 books, or 100 names. A page links to the first, last, previous, and next pages, and gives the total number of results.</p>\n<h2 id=\"a-book-entry\">A book entry</h2>\n<p>A book entry in Readwyn’s catalogue carries the title, authors, date added, languages, publisher, publication date, ISBN, genres, a cover and a thumbnail, and the description. OPDS 1.2 has no element for a series, so Readwyn puts the series and its number at the start of the text, for example “The Expanse, book 2”.</p>\n<p>Each format of the book is one download link, with its file type and size: the EPUB 3 that Readwyn converted comes first, then the original file (<a href=\"/blog/download-books-to-moon-reader-from-mac/\">which one to take</a>). A comic has only its CBZ, CB7, or CBT archive.</p>\n<h2 id=\"how-readwyn-serves-it\">How Readwyn serves it</h2>\n<figure class=\"plate whole\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-reader-apps.webp?h=e0024127c108d4d70c9a\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-reader-apps.webp?h=b3168403e87d03df89f1\" alt=\"Settings › Other devices › Reader apps: the switch Share with reader apps (OPDS) is on, the address http://192.168.1.20:7314/opds with Copy and port 7314, a warning about the Wi-Fi, and three reader apps\" width=\"780\" height=\"936\" loading=\"lazy\">\n</picture>\n<figcaption>The catalogue in Settings › Other devices</figcaption>\n</figure>\n<p>Readwyn speaks OPDS 1.2 over plain HTTP, not OPDS 2.0. The catalogue is read only: a reader app can download books and covers but can’t change anything. Every request has to carry one reader app’s user name and password (HTTP Basic), which Readwyn makes for each app in “Add a reader app…”. Reader apps have the Guest level, so the feeds show no status, progress, rating, or review.</p>\n<p>The catalogue has its own port, 7314 by default, next to the 7313 that paired Macs use, and it works on your home network only. The Readwyn app serves it while “Readwyn server” and “Share with reader apps (OPDS)” are both on (<a href=\"/blog/opds-catalogue-for-koreader/\">the steps, with KOReader</a>). The headless server for Linux, <code>ryn-server</code>, can’t serve reader apps yet.</p>\n<p>The catalogue stays quick in a big library. Readwyn timed every kind of feed page against a library of 60,000 books: with the library’s files already in the system’s file cache, no page took more than 45 ms.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/what-the-calibre-import-keeps/","url":"https://readwyn.com/blog/what-the-calibre-import-keeps/","title":"What the calibre import keeps: IDs, covers, ratings, reviews, and Kobo data","summary":"Everything Readwyn brings over from a calibre library: book IDs, ISBNs and other identifiers, covers, ratings, reviews, past reads, Kobo and Goodreads data, and custom columns.","content_html":"<p>Each book arrives with its calibre book ID, its ISBN and other identifiers, its cover and files, your rating and review, your past reads, and its Kobo and Goodreads data. Nothing in the calibre folder changes, and the book files come over exactly as they were. Before you start, the <a href=\"/blog/import-calibre-library/\">Import from calibre</a> sheet sums most of this up under “What comes over”.</p>\n<figure class=\"plate mid\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-import.webp?h=a063f7fc58b1b4e4e384\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-import.webp?h=d0691a59955fe72517c9\" alt=\"The Import from calibre sheet, with the list What comes over: book files and covers, book IDs and identifiers, ratings and reviews, Kobo reading data, Goodreads shelves and dates, and MOBI and AZW3 converted to EPUB 3\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>“What comes over”, as the sheet lists it</figcaption>\n</figure>\n<h2 id=\"ids-and-identifiers\">IDs and identifiers</h2>\n<p>Each book keeps the ID it had in calibre, and its Details tab says where it came from: “imported from calibre, book 1207”. A book you add later never takes the ID of one you removed.</p>\n<p>Every identifier calibre holds for a book comes across (<code>isbn</code>, <code>goodreads</code> and the rest). You’ll find them afterwards on the left of the Edit details sheet (<a href=\"/blog/move-large-calibre-library-keep-isbns/\">more on the ISBNs</a>). calibre’s internal <code>uuid</code> comes too, though the sheet doesn’t show it.</p>\n<h2 id=\"older-formats\">Older formats</h2>\n<p>A MOBI, AZW3, PDF or other older file is left alone, and an EPUB 3 copy is made next to it for reading. The Details tab lists both, each with its size and date. Books with DRM aren’t converted. When a file won’t convert, the book keeps only the original, Readwyn records why, and a later version of the converter has another go.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-book-details.webp?h=3f5c56cb7a1e9927112f\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-book-details.webp?h=647f1b629b6b477420d1\" alt=\"The Details tab of Middlemarch: the original MOBI file, and the EPUB 3 converted from the MOBI for reading\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The original file and the EPUB 3 made from it</figcaption>\n</figure>\n<h2 id=\"ratings-reads-and-dates\">Ratings, reads and dates</h2>\n<p>Ratings and reviews arrive with each book. So do your Goodreads shelves and dates, and a finished read for each book on your read shelf. Those reads are why Reading stats can count books you read years ago, not only the ones since the import. If you connect Goodreads later (Settings › Libraries, “Connect…”), the sync carries shelves, rating, the review with its spoiler flag, progress and the date read.</p>\n<p>Published dates keep the day you typed in calibre, whatever your Mac’s time zone (<a href=\"/blog/calibre-published-dates-time-zones/\">why that needs care</a>).</p>\n<h2 id=\"kobo-reading-data\">Kobo reading data</h2>\n<p>The Kobo reading data that calibre kept for each book comes too: how far you’d read, when you last read it, how long you’d spent on it, and the Kobo’s estimate of the time left. That hidden <code>uuid</code> is how Readwyn recognises the books calibre put on your Kobo, so they line up with your library the first time you plug the Kobo in. A Kobo sync after the import then brings the newest highlights, notes and progress from the device. Run “Device → Readwyn” first, so the Kobo’s newer positions reach the library before anything goes the other way (<a href=\"/blog/leave-calibre-keep-kobo-reading-history/\">the order, step by step</a>).</p>\n<h2 id=\"custom-columns\">Custom columns</h2>\n<p>A custom column with no match in Readwyn is stored with the book as an extra field named after the column’s label, with <code>#</code> in front. Numbers in a custom series column survive too, so “Discworld [3]” and “Discworld [1.5]” keep theirs. Where the series has no number, only the name comes across.</p>\n<h2 id=\"comics\">Comics</h2>\n<p>Comics (CBZ, CB7 and CBT) keep the details you gave them in calibre. If calibre has no page count for one, the import counts the pages of the archive, and it takes the reading direction from <code>ComicInfo.xml</code>. A comic isn’t converted: the archive is its only file.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]},
{"id":"https://readwyn.com/blog/year-in-books/","url":"https://readwyn.com/blog/year-in-books/","title":"Look back on your year in books","summary":"From 1 December to 7 January, Readwyn's library offers a card-by-card story of your reading year, ending in a summary you can save as a PNG.","content_html":"<p>From 1 December to 7 January, the library shows a card called “Your 2026 in books” with a “See your year” button. It opens a short story of your reading year, one card at a time, from the hours you read to a summary you can save as an image. At any time of year, the “Year in review” section at the bottom of Stats lists each past year with a finished book or some time read, such as “2025 in books”, and opens its story. In December the current year joins the list.</p>\n<h2 id=\"when-the-card-appears\">When the card appears</h2>\n<p>The dates go by your Mac's clock. In December the card covers the year that's still going; from 1 to 7 January it covers the year that has just ended. Its line gives the gist, for example “214 hours, 38 books, and your top author. A short story of your year.”</p>\n<p>Opening the story or dismissing the card (the close button's tooltip says “Not now”) puts it away until next season. The card is tied to this Mac, even when the app is showing another Mac's libraries.</p>\n<h2 id=\"the-cards\">The cards</h2>\n<p>The story fills the window, with “Your 2026 in books” at the top, one card in the middle, and “← → or Space to move · Esc to close” underneath. A bar of segments shows how many cards there are. The cards alternate between your accent colour and a plain panel, though the summary is always plain. In order:</p>\n<ol>\n<li><strong>Totals.</strong> Hours read, books finished, and their pages, with three covers.</li>\n<li><strong>First and last.</strong> The first and the last book you finished that year.</li>\n<li><strong>Top author.</strong> By time read, or by books finished if there's no time, with up to three covers.</li>\n<li><strong>Top genre.</strong> Its share of your reading time, with your top four genres as bars.</li>\n<li><strong>Longest book.</strong> The finished book with the most pages.</li>\n<li><strong>Fastest finish.</strong> The fewest days from starting a book to finishing it, counting both days.</li>\n<li><strong>Highest rated.</strong> Your highest rating of the year, and any other reads that got the same.</li>\n<li><strong>Most highlighted.</strong> The book you highlighted most that year, with one of the highlights.</li>\n<li><strong>Longest streak.</strong> Your longest run of days with at least 5 minutes read, drawn as dots.</li>\n<li><strong>When you read.</strong> Your busiest month and the hour you read most.</li>\n<li><strong>My Words.</strong> How many words you saved, with the first of the year and the book it came from.</li>\n<li><strong>Summary.</strong> The year's figures and five covers.</li>\n</ol>\n<p>A card with nothing to show is skipped. A year with only one finished book has no First and last card, and a year without highlights has no Most highlighted.</p>\n<p>The arrow keys, Space, a click on the card, and the turn buttons move through the story. Escape closes it.</p>\n<h2 id=\"saving-the-summary\">Saving the summary</h2>\n<p>The last card has “Save as image…”, which opens the macOS save panel with the name <code>Readwyn 2026 in books.png</code>, in Pictures. Readwyn paints the summary card at twice its size, 840 × 1280 pixels, as a PNG with transparent corners. That image is the only part of year in review that leaves your Mac, and only when you save it.</p>\n<h2 id=\"where-the-numbers-come-from\">Where the numbers come from</h2>\n<p>The story takes its figures from the Year range of Stats, so the Totals card counts the same hours, books, and pages as Stats does for the same libraries. The card rounds the hours down, so it can show one hour fewer than Stats.</p>\n<figure class=\"plate\">\n<picture>\n  <source srcset=\"https://readwyn.com/captures/mac/dark-stats-year.webp?h=c0ea8ac829fdb8e9570c\" media=\"(prefers-color-scheme: dark)\">\n  <img src=\"https://readwyn.com/captures/mac/light-stats-year.webp?h=eb5f65b0c94273d46135\" alt=\"Reading stats for 2026: 249 h read, a bar for each month, 38 books finished, 12,904 pages read, a streak of 7 days, a speed of 287 words per minute, and the covers of the books finished\" width=\"1280\" height=\"800\" loading=\"lazy\">\n</picture>\n<figcaption>The Year range of Stats, which the story draws on</figcaption>\n</figure>\n<p>Which libraries count depends on where you open the story. The card in the library always takes in all your libraries together. From Stats, the story follows the <a href=\"/blog/reading-stats-across-libraries/\">scope that Stats is set to</a>: the current library or “All libraries”.</p>\n<p>Books finished include reads that came in with a calibre library or from Goodreads, so they can go back further than your time with Readwyn. The hours only count time from Readwyn's own reader.</p>\n","date_published":"2026-10-31T00:00:00Z","authors":[{"name":"Siddharth Venu"}]}
]
}