Kobo Firmware 3.5.0

[murg]

Originally Posted by davidfor

"The rollout is on pause"

And no, I don't know when that will change.

First post of the Direct Links thread has been updated.

[Anak]

Originally Posted by davidfor

Kobo have stated that firmware 3.5.0 will be released for the Glo, Aura and Aura HD. They state the improvements as:[list][*]New Adobe RMSDK

• Support for new Adobe DRM
• Long paragraph bug is fixed!

Well, I'm not so sure that the Long paragraph bug has been completely fixed and the end of chapter bug is still there.
I'm currently reading a book (epub) that shows the similar LPB symptoms.
Every few pages there is a gap of the size of two text lines. None of the code used by the elements (css/htm/xml) explains this behaviour. I.e: no deliberate margin/padding settings.

The only difference this book has to my other books that it uses different values for the @page rule.

All my books are standard edited with

@page{margin:0;}

and the LPB seems to be fixed.

The problematic book I currently read uses these values:

@page{margin:4pt 0 3pt;}

Guess, LPB is triggered by a combination of font family, font size, line height and @page rule

Note: setting reading settings to "full screen" mode (show header and footer disabled) seems to solve the LPB.

[arspr]

Anak check my previous posts about widows and orphans (before all the absurd discussion about 0 value started). Possibly you are seeing the same issue. LPB is possibly gone but w&o don't seem to work absolutelly fine and sometimes extra space is left at the end of the page.

[violent23]

Originally Posted by arspr

Anak check my previous posts about widows and orphans (before all the absurd discussion about 0 value started). Possibly you are seeing the same issue. LPB is possibly gone but w&o don't seem to work absolutelly fine and sometimes extra space is left at the end of the page.

I have definitely seen this. I have widows and orphans set to 1 and even on 3.5.0 (where the LPB is supposed to be fixed) still get single lines that could fit at the bottom of the page moved to the next, usually at an end of a chapter. It also does appear as if font family, font size, line height does have an effect on this but I am unable to figure out the specifics. Using a patched firmware also seems to exacerbate the issue depending on the custom font family, font size, line height settings but could also be because of the extra space added to the screen from lowering the page number/position footer with the custom footer patch.

Patched or un-patched the problem appears and no amount of combinations seem to fix it.

[davidfor]

Anak: None of the books I have with long paragraphs are showing the problem that was visible before 3.5.0. That there might be some other pagination issue, doesn't surprise me. Or that fiddling with the reading settings affect this. Any change, including the @page or widows and orphans could change were the page breaks will be. And without seeing the book and the settings you are using, it isn't possible to try and judge what is happening.

[arspr]

Just in case (as looking for it can be a real PITA). Here is the link to my post where w&o, or as davidfor says, who knows what other possible thing, is shown as somehow buggy. There's some unneeded extra space at the bottom of the page, but in no way as extremely big as the long paragraph issue and, more over, without any kind of long paragraph involved...

[cybmole]

Originally Posted by arspr

Anak check my previous posts about widows and orphans (before all the absurd discussion about 0 value started)...

that would be the section where it is pendantically & nitpickingly asserted that zero is not a nacceptable value and that only foolish amateurs like & myself, & microsoft, think differently

http://msdn.microsoft.com/en-us/libr...(v=vs.85).aspx
....
Internet explorer development center

...

orphans: integer
Property values

integer
A String that specifies or receives the minimum number of lines to print at the bottom of a page.
CSS information
...
Initial Value 0

so there you have it, ignore that whole absurd section like the man says

[Anak]

Originally Posted by davidfor

Any change, including the @page or widows and orphans could change were the page breaks will be. And without seeing the book and the settings you are using, it isn't possible to try and judge what is happening.

I agree, however my book does not have any user defined widows or orphans settings in the css. The book should use the default settings of 2.
I'll add the widows and orphans properties (set to 1 as that is the lowest valid integer value) and check if it 'behaves' differently.
If you want a copy of the book I can sent you one.

UPDATE: adding widow and orphans values of 1 to the book seems to solve my reported issue.
I also have tested it with 0 (an invalid value) but it does solve the reported issue.
The problem is back when I manually set the values to default (2).
If any member of the Kobo development team reads this topic: consider to change the default values for widows and orphans to 1. Yes, I'm aware that is not in line with the standard but it improves the overall behaviour of your readers.

I used the word 'solve' above but 'improve' is probably a more accurate description. Different pages are filled with text more evenly (the same amount of text on every page).

Originally Posted by arspr

Just in case (as looking for it can be a real PITA). Here is the link to my post where w&o, or as davidfor says, who knows what other possible thing, is shown as somehow buggy. There's some unneeded extra space at the bottom of the page, but in no way as extremely big as the long paragraph issue and, more over, without any kind of long paragraph involved...

Yep, that books looks very much like mine and happens indeed without any kind of long paragraph involved.
That is also why I described it as similar to LPB. Thanks for the link, I doubt if I had found it using the forums search options.

Originally Posted by cybmole

that would be the section where it is pendantically & nitpickingly asserted that zero is not a nacceptable value and that only foolish amateurs like & myself, & microsoft, think differently

http://msdn.microsoft.com/en-us/libr...(v=vs.85).aspx
....
Internet explorer development center

...

orphans: integer
Property values

integer
A String that specifies or receives the minimum number of lines to print at the bottom of a page.
CSS information
...
Initial Value 0

so there you have it, ignore that whole absurd section like the man says

Yes, css2 always accepted the 0 value for widows and orphans, but is that is now changed to 1 (in line with css3). The W3C CSS Validation Service now rejects widows or orphans set to 0 but always accepted this value before. That is the standard now, whether i like it or not and is also a different discussion.

[arspr]

Hey guys, please do not start again.

Just think a little and you'll notice that 0 has no logical and reasonable meaning at all (I don't care about CSS definition, spec or whatever). Maybe it is an artificial and conventional way of saying "off". But saying "off" is the same as saying 1 in a strict mathematical sense, not 0.

How can you assure you leave 0 lines in a page?, does it mean your whole paragraph must be rendered in a single page (no leftovers in the previous or following pages)?, what do you do if your paragraph is longer than a page?, how can you strictly follow a 0 instruction?

Maybe it works, maybe it has always worked and maybe it will forever work... Or not... 1 is just safer and more logical. And it doesn't matter if you, (or I), like it or not. It's an absolute mathematical truth (with Gödel's permission). Period.

But if you want to use 0, use it, no problem at all, they are your books... Nobody is forcing you to use 1, it was just a piece of advice. But I repeat DO WHATEVER YOU WANT WITH YOUR BOOKS. (Yes, intentional uppercase).

[cybmole]

http://csshumor.com/good-condom-orph...erflow-hidden/

.goodCondom {
orphans: 0;
widows: 0;
break-after: avoid;
overflow: hidden;
}

[Rev. Bob]

Originally Posted by Anak

If any member of the Kobo development team reads this topic: consider to change the default values for widows and orphans to 1. Yes, I'm aware that is not in line with the standard but it improves the overall behaviour of your readers.

In your opinion.

Others, like myself, are perfectly happy with the occasional blank line at the end of a screen if it means not having one line of a paragraph on a different screen from the rest of it; it's easier for us to read the paragraph as a whole that way. It's the way tree-books have been done for decades, if not centuries, and the only reason you don't notice it there is that there are subtle typographical tricks that can be used to camouflage the missing line. However, those tricks don't work unless the layout is fixed and immutable, which is why they don't work on ebooks or websites.

Frankly, the occasional blank line isn't worth breaking standards over.

[arspr]

Anak, as the trouble is probably related to w&o calculations, of course setting them to 1 (or 0 for square-headed people) "fixes" it. But that is not the solution... There's a bug about the extra space and masking it is not acceptable (as in kepubs where w&o doesn't work in any way).

[JSWolf]

Originally Posted by Rev. Bob

Frankly, the occasional blank line isn't worth breaking standards over.

I think it is worth it. It just looks better when the end of the page is more consistent and that looks better.

[Anak]

Originally Posted by Rev. Bob

In your opinion.

Others, like myself, are perfectly happy with the occasional blank line at the end of a screen if it means not having one line of a paragraph on a different screen from the rest of it; it's easier for us to read the paragraph as a whole that way. It's the way tree-books have been done for decades, if not centuries, and the only reason you don't notice it there is that there are subtle typographical tricks that can be used to camouflage the missing line. However, those tricks don't work unless the layout is fixed and immutable, which is why they don't work on ebooks or websites.

Frankly, the occasional blank line isn't worth breaking standards over.

Sure, it is just my opinion and I was trying to find some common ground as there is no perfect solution.
If it were just an occasional blank line I would not be bothered at all but it just happens to frequently to call it occasional (it varies, but on average every 3-4-5 pages).
And because there are no subtle typographical tricks to camouflage this in "digital print" (stricktly limited to the epub format here. To avoid other discussions.) it also becomes too visible. Which is fine for some, and bothering for others. Setting the values to one would not make reading more 'difficult' but it does improve the overall look-and-feel of displayed text on e-readers.
But yes, that is just my take.

[murg]

Originally Posted by cybmole

that would be the section where it is pendantically & nitpickingly asserted that zero is not a nacceptable value and that only foolish amateurs like & myself, & microsoft, think differently

http://msdn.microsoft.com/en-us/libr...(v=vs.85).aspx
....
Internet explorer development center

...
so there you have it, ignore that whole absurd section like the man says

Trusting Microsoft on any kind of standard is just plain wrong.

They've always redefined web standards to be what they want them to be in their products. I'm starting to run into all kinds of websites that won't run in IE at all, and that's in a version that supposed to be web-standards compliant.