Firmware 4.16.13337

[JSWolf]

Originally Posted by jackie_w

That's hard to say at the moment. Footers is probably the biggest area of change.

Some of the old patches will run without error in 4.16.x it's just that they don't actually do anything to the new page layout. My 'Custom menubar - reduce height by 33%' (and the 50% equivalent) are examples of this. I did find some new CSS to patch but in 4.16.x it looks hideous when patched. 4.17 may work better but it's equally possible the patch won't be needed at all in 4.17.

I haven't had much time to play yet because I'm busy with calibre 4 beta but I'm hoping to get back to it next week.

I am hoping the footer can be patched. I'm not all that bothered if the footer text cannot be changed. But I would like to reduce the space and drop it to the bottom of the screen.

I do use your reduce height by 33% patch. So I'm hoping it can be made to work with 4.17 unless it's not needed.

Which patches work but don't do anything? Is the reason they don't do anything because they are no longer needed or because of changes to the code?

[JSWolf]

Originally Posted by Calenorn

Hello davidfor,

In general, what is the most likely way to edit / repair an epub that has this problem.

Thanks in advance!

Use the calibre Editor to edit the ToC to add in an enter for every HTML file that has not got an entry in the ToC. You have to make sure they are in the correct order. The ToC you have to edit is the NCX ToC.

[MGlitch]

Originally Posted by JSWolf

Which patches work but don't do anything? Is the reason they don't do anything because they are no longer needed or because of changes to the code?

I suspect the ones that do nothing are ones affecting the reader UI, I.e. the reduce menu bar patch, likely the margins patches, and anything touching the header and footer.

As to why I believe Geek already mentioned that patches would have to be done to support both the old UI which is accessible by disabling the new one according to David however it causes instability, and the new UI. Which is why Geek also said unless there was a compelling reason to do more he’d only be supporting the new UI.

As to if the old patches work with 4.16 on the old UI I don’t know but I’d wager that they would need to be updated and as the old UI is now buggy it’d be a questionable idea to patch them without knowing what you’re doing

[Calenorn]

Originally Posted by JSWolf

Use the calibre Editor to edit the ToC to add in an enter for every HTML file that has not got an entry in the ToC. You have to make sure they are in the correct order. The ToC you have to edit is the NCX ToC.

Thank you, sir. So I guess "Generate ToC from files" would be the best option?

I will give that a try.

[JSWolf]

Originally Posted by Calenorn

Thank you, sir. So I guess "Generate ToC from files" would be the best option?

I will give that a try.

Sorry, but that's the worst option. You have to do it by hand in the ToC editor. The problem is that Generating a ToC relies on things not always there in the the chapter headers (and code) and because of this, you could end up with a much smaller ToC needing even more hand editing. So just do it by hand in the ToC editor and forget trying to generate.

[jackie_w]

Originally Posted by JSWolf

Which patches work but don't do anything? Is the reason they don't do anything because they are no longer needed or because of changes to the code?

It's possible that those which still run without error but don't appear to do anything once installed are actually doing what they always did, namely changing the look&feel of the old page layout not the new one.

When I did a brief test of disabling the new layout, without any patches, I had several crashes so I didn't want to try patching something so flaky in case I bricked the device.

[JSWolf]

Originally Posted by jackie_w

It's possible that those which still run without error but don't appear to do anything once installed are actually doing what they always did, namely changing the look&feel of the old page layout not the new one.

When I did a brief test of disabling the new layout, without any patches, I had several crashes so I didn't want to try patching something so flaky in case I bricked the device.

I think it's a good decision to patch the new UI and not the old UI as eventually, the old UI will be going away.

[davidfor]

Originally Posted by Calenorn

Thank you, sir. So I guess "Generate ToC from files" would be the best option?

I will give that a try.

Firstly, the changes in the firmware mean that there are less issues with kepubs that do not have a one-to-one relationship between the ToC and the internal files. In fact, I'm not sure if there are any problems now. If I can find a list of them, I will do some testing.

I'll at least partly disagree with Jon on this. Using "Generate ToC from files" should work to solve the issue. But, it uses the first text in the file for the heading, or the file name if there isn't any, so it might not produce what you want. But, it is a starting point that can be edited.

Personally, I use "Generate ToC from major headings". This generates the ToC from h1, h2 and h3 tags in the book. Then I look and see what is missing. I will sometimes use the ToC editor to add an extra entry, but, I usually add the heading to the file. If there isn't a natural heading in the file, I'll add it with a class that doesn't display the text or take space. I will also do this for some books that use images or other styles for chapter headings that the ToC generation options do not work with. That allows me to regenerate the ToC at any time without losing entries.

And Jon, before you get on your hobby-horse about wasted space and useless files, I completely and utterly disagree with your campaign to make all books look the same, remove all except absolutely necessary whitespace and remove extras that the author or publisher added. My cleanup is about making sure I can read the book. My suggestions are to help people achieve the results that they desire.

[s4nder]

How's hyphenation with the new reader? Does it properly use the dictionaries on the device?

[Strrav]

Hello, with this new firmware, is it possible to have the inverse reading for the night , white text on black screen ?tnaks

[Calenorn]

Thanks again. As you say, a starting point.

One last question, to see if I understand (then I will stop hijacking the thread):

If an html file is removed from (or added to) the epub, AND the toc.ncx file is NOT adjusted to match, THEN the kobo will have difficulty remembering the page being read when the file is closed and re-opened.

Does that sound right?

Originally Posted by davidfor

Firstly, the changes in the firmware mean that there are less issues with kepubs that do not have a one-to-one relationship between the ToC and the internal files. In fact, I'm not sure if there are any problems now. If I can find a list of them, I will do some testing.

I'll at least partly disagree with Jon on this. Using "Generate ToC from files" should work to solve the issue. But, it uses the first text in the file for the heading, or the file name if there isn't any, so it might not produce what you want. But, it is a starting point that can be edited.

Personally, I use "Generate ToC from major headings". This generates the ToC from h1, h2 and h3 tags in the book. Then I look and see what is missing. I will sometimes use the ToC editor to add an extra entry, but, I usually add the heading to the file. If there isn't a natural heading in the file, I'll add it with a class that doesn't display the text or take space. I will also do this for some books that use images or other styles for chapter headings that the ToC generation options do not work with. That allows me to regenerate the ToC at any time without losing entries.

And Jon, before you get on your hobby-horse about wasted space and useless files, I completely and utterly disagree with your campaign to make all books look the same, remove all except absolutely necessary whitespace and remove extras that the author or publisher added. My cleanup is about making sure I can read the book. My suggestions are to help people achieve the results that they desire.

[geek1011]

Originally Posted by davidfor

Firstly, the changes in the firmware mean that there are less issues with kepubs that do not have a one-to-one relationship between the ToC and the internal files. In fact, I'm not sure if there are any problems now. If I can find a list of them, I will do some testing.

I usually edit my books to be one file per chapter and a TOC entry per file, and to remove overly specific font styling, but one thing I noticed the most is page numbering jumping between chapters, and stats bring messed up.

P.S. One pet peeve of mine is the books with a whole file just for the title of the chapter, and with the content in another file, not in the TOC, but referenced using the spine.

And Jon, before you get on your hobby-horse about wasted space and useless files, I completely and utterly disagree with your campaign to make all books look the same, remove all except absolutely necessary whitespace and remove extras that the author or publisher added. My cleanup is about making sure I can read the book. My suggestions are to help people achieve the results that they desire.

I partly disagree with this, but only with regards to fiction. I believe publishers get too much freedom with EPUB, and tend to abuse it (especially for fiction). The font overriding wars is an example of this, and using divs for everything and using CSS to make it online, increase the font size, look like a paragraph, and so on.

Originally Posted by Strrav

Hello, with this new firmware, is it possible to have the inverse reading for the night , white text on black screen ?tnaks

Once I have the patches, there will be a convenient way to do this. For now, you can enable dev mode (search devmodeon), and enable it from device info, but note that this is now even longer to get to from reading a book. I haven't tried them, but there are also night mode mods.

[rashkae]

Originally Posted by Calenorn

Thanks again. As you say, a starting point.

If an html file is removed from (or added to) the epub, AND the toc.ncx file is NOT adjusted to match, THEN the kobo will have difficulty remembering the page being read when the file is closed and re-opened.

Does that sound right?

If the book is being edited by Calibre editor, removing one of the html files will automatically remove any TOC entry that points to it. Adding an HTML file will trigger the Kobo bugs. However, the problem you describe, not remembering the page at some points, will only happen with Epub files (and only applies to Firmware version 4.15). It will not affect files sent to the Kobo with the Kobo Touch Extended Plugin, or otherwise processed into Kepub Files. The only known drawback of not having all files in the TOC of a Kepub is the kobo will count pages out of order.

Personally, for people who want to sideload Epub files, I would suggest keeping the Kobo on firmware 4.14. I fully expect this problem will be addressed in the next update. This might be a problem for users who both want to sideload Epub *and* synchronize books with Kobo cloud.

[Calenorn]

Originally Posted by rashkae

If the book is being edited by Calibre editor, removing one of the html files will automatically remove any TOC entry that points to it. Adding an HTML file will trigger the Kobo bugs. However, the problem you describe, not remembering the page at some points, will only happen with Epub files (and only applies to Firmware version 4.15). It will not affect files sent to the Kobo with the Kobo Touch Extended Plugin, or otherwise processed into Kepub Files. The only known drawback of not having all files in the TOC of a Kepub is the kobo will count pages out of order.

Personally, for people who want to sideload Epub files, I would suggest keeping the Kobo on firmware 4.14. I fully expect this problem will be addressed in the next update. This might be a problem for users who both want to sideload Epub *and* synchronize books with Kobo cloud.

Hmmm. So is the day coming when side-loaded books will have to be kepub to play nicely with Kobos? I have a lot of old epubs. A LOT. Sure don't want to have to convert them all!

[Sirtel]

Originally Posted by Calenorn

Hmmm. So is the day coming when side-loaded books will have to be kepub to play nicely with Kobos? I have a lot of old epubs. A LOT. Sure don't want to have to convert them all!

Why not? It's a bulk job in Calibre. It's not like you have to convert them one by one.

Even better, using the extended driver, the books will be converted to kepub automatically when sending them to the device. You won't have the kepubs littering your library.