Glo Kobo Firmware 2.3.1

[jusmee]

Originally Posted by davidfor

. If I don't get dragged out to do the Christmas shopping (and I cannot find an smiley that describes that horror).

My sympathies if you do. I have managed to avoid it myself

[GadgetMan]

I just bought a new Kobo Glo.. I am going to leave it Virgin.. no side loads etc.. I think that way it will work fine.. mine came with the lastest firmware and it works perfect.. I think doing side-loads, mods etc may cause problems and contamination.. maybe not.. but just a thought..
DG

[jusmee]

Originally Posted by GadgetMan

I just bought a new Kobo Glo.. I am going to leave it Virgin.. no side loads etc.. I think that way it will work fine.. mine came with the lastest firmware and it works perfect.. I think doing side-loads, mods etc may cause problems and contamination.. maybe not.. but just a thought..
DG

How are you going to borrow free books from local libraries? What about being able to buy a book cheaper at other places than Kobo. Frightening you into lock-in to the Kobo store is, I suspect, what they want.

[GadgetMan]

Originally Posted by jusmee

How are you going to borrow free books from local libraries? What about being able to buy a book cheaper at other places than Kobo. Frightening you into lock-in to the Kobo store is, I suspect, what they want.

I will not borrow free books and if I pay a little more from Kobo for a Book, Magazine or Newspaper so be it.. the experience of reading on the Kobo is worth the trade off.. I primarily have it to buy a Top Selling book.. period.. but I hear what you are saying.. I have Kindle and iPad so I can get the other stuff on that Gadget..
DG

[EldRick]

(except for one niggle, which is the large/inconsistent amount of whitespace at the bottom of some pages)

This one is certainly a large-annoyance issue on every Kobo in the world, and is Long overdue for a fix.

If davidfor is correct, and the issue is related to programming stupidity in handling page number boundaries (possible), then a far more appropriate solution would be to deal with the page numbers like my Sony PRS-950, i.e. let text flow to the bottom of the screen and present page numbers as 215/216 when there is a page number break that does not match a screen break.

Then Kobo programmers can focus their limited attention on returning to the correct paragraph within the book, regardless of font changes and ignore page-numbers for all other purposes than that for which they are useful - displaying page-numbers that match the printed book.

There is no other reason for displaying the page-numbers than to allow readers to match up an e-book with the printed version (and that's only semi-sensible anyway, since publishers will often publish different versions, such as hardback and softback).

However, page numbers are useful in an academic environments for citations and for helping students find the same page when some are using e-readers and some using hardcopy, so you have to show them - just don't let them play any role in setting screen boundaries.

[davidfor]

Originally Posted by EldRick

This one is certainly a large-annoyance issue on every Kobo in the world, and is Long overdue for a fix.

If davidfor is correct, and the issue is related to programming stupidity in handling page number boundaries (possible), then a far more appropriate solution would be to deal with the page numbers like my Sony PRS-950, i.e. let text flow to the bottom of the screen and present page numbers as 215/216 when there is a page number break that does not match a screen break.

The page numbering has nothing to do with the gaps at the bottom of the page. To me, this is a problem in the book. And I know that no-one will agree with me, but, I'll explain.

The gap is caused by the settings of the widows and orphans style setting. The "widow" is that last line from a paragraph that appears at the top of the next page. The "orphan" is the solo line at the end of a page. CSS styles support setting the number of lines for both of these. The default is either 2 or 3. Kobo seem to be using 3.

What this means is that if a paragraph goes over the end of the screen, and it has less lines than the sum of the widows and orphans settings, it will be pushed onto the next screen. And there will be a big gap at the bottom of the screen. Exactly how big that gap is depends on a lot of things. As well as the widows and orphans settings, it depends on the font and size, line spacing and margins between paragraphs. And probably some other things.

So, the device is respecting the design of the book. It does this in the same way as it respects the other settings. To me, this is the correct thing to do. The book designer decided to do it this way. At least I hope it was a conscious decision and not that they forgot. Or worse, didn't know they could do this.

Then Kobo programmers can focus their limited attention on returning to the correct paragraph within the book, regardless of font changes and ignore page-numbers for all other purposes than that for which they are useful - displaying page-numbers that match the printed book.

There is no other reason for displaying the page-numbers than to allow readers to match up an e-book with the printed version (and that's only semi-sensible anyway, since publishers will often publish different versions, such as hardback and softback).

However, page numbers are useful in an academic environments for citations and for helping students find the same page when some are using e-readers and some using hardcopy, so you have to show them - just don't let them play any role in setting screen boundaries.

I don't completely agree. The page numbers give a good idea of the position in the book and the length. The percent read is handy, but 50% of "The Cat in the Hat" is a lot different to 50% of "War and Peace".

[murg]

Originally Posted by davidfor

So, the device is respecting the design of the book. It does this in the same way as it respects the other settings. To me, this is the correct thing to do. The book designer decided to do it this way. At least I hope it was a conscious decision and not that they forgot. Or worse, didn't know they could do this.

I think that the problem with the widows/orphans handling isn't that Kobo is paying attention to stated values in the epub, it's that it is setting non-zero values as the default.

If the default settings are being used, then the book designer isn't involved in the formatting. And face it, almost all ebooks aren't 'designed', they're mashed together with a minimal attempt at design, especially in regards to page layout.

Additionally, the whole widows/orphans thing is mostly an artifact from physical books. On an ereader, they really aren't that necessary.

[GeoffR]

Originally Posted by davidfor

The gap is caused by the settings of the widows and orphans style setting.

[...]

So, the device is respecting the design of the book. It does this in the same way as it respects the other settings. To me, this is the correct thing to do. The book designer decided to do it this way. At least I hope it was a conscious decision and not that they forgot. Or worse, didn't know they could do this..

The problem for me is that the device is respecting the (often badly chosen) settings in the book rather than respecting the reader's preferences. In firmware 2.3.1 Kobo made changes so that user preferences can override epub book settings for margins and linespacing, and that was a big improvement in my view. I think allowing user preferences for widow/orphan settings to override the book settings would be consistent with that, and would be another big improvement.

[DNSB]

Originally Posted by EldRick

Then Kobo programmers can focus their limited attention on returning to the correct paragraph within the book, regardless of font changes and ignore page-numbers for all other purposes than that for which they are useful - displaying page-numbers that match the printed book.

There is no other reason for displaying the page-numbers than to allow readers to match up an e-book with the printed version (and that's only semi-sensible anyway, since publishers will often publish different versions, such as hardback and softback).

There is no point in displaying page numbers the way you suggest. I have 4 copies of Stranger in a Strange Land. The hardcover first edition signed by RAH which I literally had to bleed for, a hard cover of the edition where some pages deleted from the earlier editions were added back in, a hard cover which is reset edition of the first printing and a paperback edition -- the latter two being gifts. The page numbers within the three hard cover editions do not agree so just which print edition should be used to match the page numbers displayed by the electronic edition?

At least if there was agreement that a page is 1024 visible characters would make the page numbering consistent across ereaders which is currently not the case.

Regards,
David

[davidfor]

Originally Posted by DNSB

I have 4 copies of Stranger in a Strange Land. The hardcover first edition signed by RAH which I literally had to bleed for,

But it was worth it, wasn't it? And yes, I'm jealous.

[Mrs_Often]

Originally Posted by davidfor

The page numbering has nothing to do with the gaps at the bottom of the page. To me, this is a problem in the book. And I know that no-one will agree with me, but, I'll explain.

The gap is caused by the settings of the widows and orphans style setting. The "widow" is that last line from a paragraph that appears at the top of the next page. The "orphan" is the solo line at the end of a page. CSS styles support setting the number of lines for both of these. The default is either 2 or 3. Kobo seem to be using 3.[...]

I think it is strange that Kobo decided on a default of 2 or 3. Using widows and orphans on an ereader is useless, as murg mildly suggested a few posts back. Actually, not just useless but confusing and annoying.

More importantly, I don't think this is the only 'problem' that's going on. After I read this information a while back in a different thread I re-coverted all my epubs to include widow and orphan values of 0. This helped a lot (but did not completely get rid of slightly premature page breaks). BUT. There is a bug of some kind that has been previously mentioned only a few times, where paragraphs that are longer than a page (or very long, I don't know the cutoff) always start on a new page. This meant that the book I was reading at the time (a Stephen King I think, with veeeerrry long paragraphs) frequently stopped only a couple of lines into a page, making me wonder if the chapter had already finished. But no, the next paragraph was just put onto the next page.

Now, I cannot remember what firmware this was (not more than 2 version back, maybe just one), but it was definitely not a widow/orphan thing.

[Corpsegoddess]

So I'm in Vancouver, Canada. Avoiding a long story, I don't deal with the Kobo desktop because it doesn't like my Kobo (or the one before it, this is a replacement. It disconnects from the Kobo after a minute or so and I can never get anything done). Anyway, as a result, I always do everything via wi-fi.

When I update via wi-fi (which I've just done for the third time in three days), I get a software download notification. Each time, it downloads and installs, but says it's still version 2.1.5. But--it also recreates the multiple shelves I've deleted--every time, after each download.

Is anybody else experiencing this? I have a Kobo Touch, by the way.

[davidfor]

Originally Posted by Lynx-lynx

How do I send you a copy of my Glo's database?

I'm quite willing because I don't keep anything on it that isn't entirely expendable.

I've had a look at the database that Lynx-Lynx sent me. Thanks for that, and my apologies for the mistake in the first message.

A little explanation first.

The reference stored for the reading position and bookmarks are made up of several parts:

- A reference to the chapter
- A reference to the file the position is in
- A point in that file

As a lot of epubs a built with one chapter per file, the first two parts usually look like they are repeating some information, but they are serving different purposes.

The point in the file is a series of number separated by slashes and colons. I'm not completely sure what the numbers mean, but it looks like they are counters for walking the HTML/XML tree. If I am right, it is a reference to the end of the first paragraph on the screen.

So, using this information, the reader application can work out where the reader is up to and where to put bookmarks. Looking at the database Lynx-Lynx supplied, the current reading position and the bookmark for the page are identical. Same chapter reference, same file and same point. Looking at my database, it is the same.

What does this tell me? Not much. If they were different, it would tell me that the calculation for the two types of position was handled in different ways. Then I might have been able to see where the two points actually were in the book and then on the screen. It looks like the code for positioning the bookmark is different from the code for getting to the last reading position and the latter is wrong. My guess would be the code is that an index isn't being incremented at the right place or someone is counting from zero instead of 1. But this doesn't tell me why it works for some people and not for others.

[murg]

Originally Posted by davidfor

But this doesn't tell me why it works for some people and not for others.

Because sometimes you feel like a nut, and sometimes you don't.

[murg]

Originally Posted by Mrs_Often

I think it is strange that Kobo decided on a default of 2 or 3. Using widows and orphans on an ereader is useless, as murg mildly suggested a few posts back. Actually, not just useless but confusing and annoying.

The 2 and 3 values are the standard CSS defaults for these values.

Specifically setting the values to 0 in the CSS body element does work.