Kindle to utilize epub format in late 2022

[JSWolf]

Originally Posted by celinego

CathyA or JSWolf .... another little problem I need help on please.

So yesterday was happily using the Modify Epub plugin and sending epubs to my kindle and kindle app on iphone. One particular book isn't working though. I'm getting the standard email from Amazon saying it doesn't accept epubs but later says in addition it now accepts epubs. Don't think it's too big a file either as I presume I'd get a max size exceeded in the send process. Calibre says book sent.
So I converted the epub to mobi and sent it and that worked. Amazon also then sent me an email regarding from August can't send mobis any more. Any ideas why this one modified epub isn't sending? The only other thing the Amazon email says is to check the document isn't password protected or encrypted - neither applies. The book is The Brethren by Bob Woodward & Carl Armstrong in case anyone wants to try it. Many thanks in advance for any suggestions.

Load this ePub into the Calibre editor. Use both the Calibre ePub checker and epubckeck to see if anything needs to be fixed and fix it. When both checkers show no errors, then try sending it again. Also make sure there is an NCX ToC with the ePub.

[JSWolf]

Originally Posted by fendesa

try to reformatting the Epub in another Epub and than send to kindle (the 2nd Epub) I resolve the strange symbol and compress a little bit the size of the ebook. I think in the future update they resolve this problem of bad formatting.

Inviato dal mio MAR-LX1B utilizzando Tapatalk

Better solution (IMHO) is to use the Modify ePub plugin to make sure the ePub is in UTF-8. Funky punctuation is an encoding issue.

[JSWolf]

Originally Posted by Geoff-D

So I did an epub to epub conversion in Calibre and that works perfectly. I emailed the converted file directly rather than let Calibre do it.

Thanks for the tip.

That is not the best way to fix the problem. The fix is to use Modify ePub to make sure the ePub is UTF-8.

[ickam]

Originally Posted by j.p.s

They finally realized their policy of rejecting EPUB was stupid.

Their automatic conversion engine has worked on EPUB for years, and people have been bypassing the policy by renaming EPUB to have a zip, txt, or png extension for years.

Every epub that I tried to convert since they announced epub support for send to kindle has come out with messed up apostrophes and special characters, although I didn't change the extension. Anything I can do to fix that?

[DNSB]

Originally Posted by ickam

Every epub that I tried to convert since they announced epub support for send to kindle has come out with messed up apostrophes and special characters, although I didn't change the extension. Anything I can do to fix that?

I'm going to assume that you have access to calibre, calibre's ModifyEpub plugin, Sigil and Sigil's EpubCheck plugin.

• Import the epub into calibre
• Run the ModifyEpub plugin with the Encode HTML in UTF-8 option checked.
• Open the epub with Sigil. Say no to the Automatically Fix errors
• Run Mend or Mend and Prettify from the Tools menu
• Run EpubCheck. If any errors, fix them
• Save the epub and then send it to Amazon

See attached images.

Edit: This is the process I've used and so far my limited testing hasn't had failure or odd looking output.

[JSWolf]

Originally Posted by ickam

Every epub that I tried to convert since they announced epub support for send to kindle has come out with messed up apostrophes and special characters, although I didn't change the extension. Anything I can do to fix that?

Go back and read this thread in it's entirety. The answer is there.

[JSWolf]

Originally Posted by DNSB

I'm going to assume...

How many times is the same answer going to be given for those who do not want to red the entire thread?

[jhowell]

Originally Posted by JSWolf

How many times is the same answer going to be given for those who do not want to red the entire thread?

Well there are over 200 posts and it isn’t clear that this thread is the right place to look for an answer.

Once sending of MOBI is eliminated this issue is going to come up frequently. Perhaps a sticky thread on how to fix EPUBs for sending to Kindles would help.

[lkmiller]

Of course, the first epub I try sending to my Kindle didn't work. It was a DRM-free epub from Tordotcom. Took the same file, changed the extension to png, and that worked fine.

Tried another Tordotcom epub and that worked fine without the png trick.

Didn't look closely at the formatting. Just trying to see if it worked at all.

I don't usually use this myself, but I'm wondering how this change to allowing epub/not allowing mobi will effect places that email books to your Kindle for you (i.e. Weightless Books or NetGalley).

[johnmatushek]

I emailed 12 epub files to my Kindle today. 9 worked and 3 failed. I loaded the three failing files into the Kindle Previewer program and they all had errors in the TOC. I found two cumbersome ways to fix them:
1. In Calibre’s TOC editor, delete all entries and rebuild it from scratch.
2. Use Calibre to convert from epub to mobi, delete the epub, then convert from mobi to epub.

Is there a simpler way to verify and fix the TOC using Calibre?

[Quoth]

You can convert epub to epub. The source is renamed Original_Epub.
There are settings in Conversion to build TOC, fix CSS, do file breaks at headings (so an Amazon created version has page breaks) etc.

The TOC editor in epub Editor is simple to use and can automatically make a new TOC from headings.

[JSWolf]

Originally Posted by johnmatushek

I emailed 12 epub files to my Kindle today. 9 worked and 3 failed. I loaded the three failing files into the Kindle Previewer program and they all had errors in the TOC. I found two cumbersome ways to fix them:
1. In Calibre’s TOC editor, delete all entries and rebuild it from scratch.
2. Use Calibre to convert from epub to mobi, delete the epub, then convert from mobi to epub.

Is there a simpler way to verify and fix the TOC using Calibre?

Use the Calibre editor to build the ToC. You do it is automatically if possible, manually if need be. Converting to Mobi and then to ePub is a really bad idea as that seriously changes the code.

[JSWolf]

Originally Posted by Quoth

You can convert epub to epub. The source is renamed Original_Epub.
There are settings in Conversion to build TOC, fix CSS, do file breaks at headings (so an Amazon created version has page breaks) etc.

The TOC editor in epub Editor is simple to use and can automatically make a new TOC from headings.

Use the Calibre editor to fix the ToC. ePub > ePub is not always a good idea.

[johnmatushek]

Originally Posted by JSWolf

Use the Calibre editor to build the ToC. You do it is automatically if possible, manually if need be. Converting to Mobi and then to ePub is a really bad idea as that seriously changes the code.

The TOC editor in Calibre is a powerful tool that I often use. But it does not seem like the ideal solution to this particular problem. The way I found to correct the problem is to delete all TOC entries then rebuild from scratch. However, this is time consuming and error prone. In my experience, most books do not have proper headings, so the automated technique doesn’t work. Recreating the TOC from regular expressions is notoriously hit-or-miss.

Calibre has some features for checking TOC already. When the TOC editor is first invoked, it flags certain invalid entries with a red icon. I usually just delete those. However, this error check doesn’t seem to catch all the errors that Amazon does.

I would like to see the TOC editor beefed up to find all the types of errors that Amazon might fail on. Ideally, I’d like a single button that says “fix the TOC.”

[Quoth]

Impossible if the ebook is badly formatted.

If the Calibre TOC tool is not automatic then the headings HTML etc is rubbish and you may not have file/page breaks where you want them.

I've seen Chapter heading that are just <p> and some style maybe even with a span and div too. You only need an "h" tag, example using two lines
<h2 class="chapter">Chapter 2<br/>A New Hope</h2>
(This allows h1 for parts, sections, books etc. The Title should just be a p tag, unless there are multiple books in the ebook, then it's h1.
if one click doesn't make a decent TOC, then the ebook is sort of broken anyway.

In LO Writer Leveln in paragraph style translates to <hln class="whatever">
Then the CSS .whatever will have the font size, face, weight, indent, padding, margins, alignment. You need padding-top for space at top of a new page. margin-top only works if there is a prior paragraph. In LO Writer style "padding" is under borders and you set "none", untick Syncronise and then can set Padding top for a heading that does a page break.
Some formats/platforms ignore page break directives. The trick is to have Calibre do a file break at those headings (which can be automatic).