[Device Plugin] Kobo Touch Extended Driver

[arspr]

Originally Posted by anacreon

Mostly, when I then reconverted my epubs, it looked good, but I have two problems

1. all the beautiful centred separations (stars, small pictures and such) become left justified
2. in the books where there are negative indents, which remain readable as epubs, I loose the first characters of every line.

Without a test book I cannot speak about 1, but 2 is probably the expected behaviour.

I mean, if you have 0 as the left margin/padding value, a negative indentation is in fact telling the reader to display some characters OUTSIDE the screen.

You need to add some positive left margin/padding so you have available room for the negative indentation...

Originally Posted by anacreon

body { orphans: 1; widows: 1; }

Just for your knowledge. This setting is useless in Kepubs because it always behaves in this way. Kobo ACCESS renderer doesn't support widows and orphans, so it always behaves as if they are disabled no matter what you set.

(Nevertheless it's a good setting if you want to be sure that they are going to remain disabled even if they worked sometime somewhere).

[njess]

I've just downloaded this plugin but when I read a book on my Aura the pages are all mixed up? How come? Is there anything I can do about it?
I can see that the pages are in the right order it's just the numbers that are completely off-they can go something like this: 1,2,3,54,21 and so on.

[jgoguen]

That sounds like an issue with the TOC. Can you attach an image of a page that's out of order and the page right before it?

[njess]

Yes, here is the page I'm reading (to the left) and the previous one was the one to the right. I can see that the pagenumbers skip when you get to the next chapter. Ie the numbers are in order in the chapter but when I get to the next one the numbers jump up or down.

I don't know how to attach a pic of it since it says that it should be a http.

[GeoffR]

Originally Posted by njess

I've just downloaded this plugin but when I read a book on my Aura the pages are all mixed up? How come? Is there anything I can do about it?
I can see that the pages are in the right order it's just the numbers that are completely off-they can go something like this: 1,2,3,54,21 and so on.

I have seen this in Kobo-synced kepubs too, so I don't think it is anything to do with this driver.

From what I can tell, the problem is that the kepub reader is much fussier than the epub reader about how the entries in the book's toc and spine are arranged, it expects the toc and spine to match exactly on the ordering of entries, any inconsistency can cause problems with page numbering. It also expects each entry to reference content in a seperate .html file, and if not then the in-book stats won't work properly.

I don't know of any automated way to fix it that always works, it is a matter of editing the toc and spine to remove inconsistencies, and splitting up the .html files so that each toc entry has its own file (An epub to epub conversion in Calibre can sometimes help with splitting the files, but it doesn't always work.)

Edit: These problems didn't usually show in Kobo-synced kepubs in the past, because the kepub reader only displayed the number of pages per chapter. It was only when Kobo added the new page numbering options that the inconsistencies between toc and spine became more of an issue.

Edit2: Here is an example from this book. When I read the kepub version synced direct from Kobo, the page numbers were fine for most of the book, but they were completely wrong when reading the Lexicon. What seems to have happened is that the publisher decided to move the Lexicon (contained in 9780061743948_chapter_34.html) from the back of the book to the front, updated the TOC but forgot to update the spine.

Spoiler:

Extract from OPF <manifest>:

		<item href="9780061743948_chapter_33.html" id="x_9780061743948_36" media-type="application/xhtml+xml"/>
		<item href="9780061743948_chapter_34.html" id="x_9780061743948_37" media-type="application/xhtml+xml"/>
		<item href="9780061743948_aboutauthor.html" id="x_9780061743948_38" media-type="application/xhtml+xml"/>

Extract from OPF <spine>

		<itemref idref="x_9780061743948_36"/>
		<itemref idref="x_9780061743948_37"/>
		<itemref idref="x_9780061743948_38"/>

Extracts from toc.ncx

            <navPoint id="np-6" playOrder="6">
              <navLabel>
                <text>Lexicon</text>
              </navLabel>
              <content src="9780061743948_chapter_34.html"/>
            </navPoint>

            <navPoint id="np-39" playOrder="39">
              <navLabel>
                <text>Chapter 8</text>
              </navLabel>
              <content src="9780061743948_chapter_33.html"/>
            </navPoint>

            <navPoint id="np-40" playOrder="40">
              <navLabel>
                <text>About the Author</text>
              </navLabel>
              <content src="9780061743948_aboutauthor.html"/>
            </navPoint>

[jgoguen]

@njess: If you use the Quick Reply at the bottomif the page to post (or if you edit your post) there's a "Go Advanced" button to click on. That will give you more options, including a "Manage Attachments" button that lets you upload files.

[arspr]

An issue I've been suffering for some time and I don't know if it's possible to be easily solved or not.

Can KTE upload books with the same author and title anyhow?

I have suffered two cases:

• Books in different translations which maintain the title.
• Books in different editions (even in the same language).

The problem seems to be caused by the way the books are named in the device (Title - Author scheme). With this scheme, only one version of each book can exist and any new version you try to send to the device just overwrites the previous one.

I don't know if there's an option I'm not using properly or even if some improvement would be possible in the driver...

[Terisa de morgan]

I don't think driver can "guess" this. I'm afraid that, if you want to keep different versions, you should add any information to the book in the device (you can configure the name). But I'm not sure if, after this, the Kobo SW will be able to recognize the books as different.

[jgoguen]

First, my initial reaction is to put this in the main driver. But Terisa is right, the best way to do this would be to maintain it by yourself. You could use a metadata plugboard to keep the same titles in calibre but send a different title to the device.

[davidfor]

The Kobo software has no problem with multiple books with the same title and author. The key in the database is the file name and path. As long as you put the different editions in different places on the device, they will be treated as different books. You can change the save template to include whatever it is that is different about the editions so that sending one won't clobber the other.

[arspr]

Ooops. Then my question is how/where can I configure the name the driver uses?

(I had already "solved" the issue by sending the first book, then renaming it in the device through Windows Explorer, then sending the second one... But it feels quite hackish...)

[Terisa de morgan]

In the device settings there's a field called "Saving template" (or something similar, I'm using the Spanish interface). My input there is:

{tags}/{author_sort} {publisher:|-|}- {title}

If you want to differentiate them, you should have something like this, for example:

{title}{colEsp} -{author}

where colEsp is the column where you have the element to identify them. With this, the files go to different name and the title in the ereader is the same.

[arspr]

Thanks guys. (I don't know how I haven't seen that option )

It should be more than enough.

I'm on holidays the whole Holy Week, so I'll test it next one...

EDITED----

Obiously, fully confirmed.

I've just changed the default

{author_sort}/{title} - {authors}

into

{author_sort}/{title} - {authors} - {publisher}

For the time being it's perfect. If I find an extremely strange case where this info is duplicated, I'll made further tweaks but obviously no change is needed in the driver...

[monry]

Hi! Maybe someone will be able to help me. In some books in footer instead of time to the end of chapter Kobo gives me time to the next picture/image (like it would read it as the marker of chapter end). Does anyone know if there is a simple solution to this?
And now I sideloaded another ebook and in this one (with structure chapter-subchapter) the estimated times are not working properly - through the entire first subchapter it gives me 0 min..
Edit 2: OK, so problem with 0 min affects most ebooks in my library, no matter if there are subchapters or not. Sometimes it starts to give estimated time, but after jump to the beginning of chapter there is 0 min once more. Tried to factory reset reader, update it to 3.13.1 - nothing helps.

[davidfor]

Originally Posted by monry

Hi! Maybe someone will be able to help me. In some books in footer instead of time to the end of chapter Kobo gives me time to the next picture/image (like it would read it as the marker of chapter end). Does anyone know if there is a simple solution to this?

Recent firmware added options in the reading settings about what to show in the footer. The choices are page number, percent read and time left. And there is a choice as to whether to display them for the current chapter or the whole book. Then when reading, if you do a long press on the number displayed, it will change the displayed value.

And now I sideloaded another ebook and in this one (with structure chapter-subchapter) the estimated times are not working properly - through the entire first subchapter it gives me 0 min..
Edit 2: OK, so problem with 0 min affects most ebooks in my library, no matter if there are subchapters or not. Sometimes it starts to give estimated time, but after jump to the beginning of chapter there is 0 min once more. Tried to factory reset reader, update it to 3.13.1 - nothing helps.

Kepubs are supposed to be one ToC entry per internal file. If you have more than one ToC entry pointing to a file, the per chapter page numbering isn't correctly handled. I assume this also affects the remaining time, but I haven't tested that part of it.