GoogleCodeExporter
commented
9 years ago [deleted comment]Closed GoogleCodeExporter closed 9 years ago
GoogleCodeExporter
commented
9 years ago [deleted comment][deleted comment][deleted comment][deleted comment][deleted comment]Original comment by d...@google.com on 28 Mar 2013 at 5:29
GoogleCodeExporter
commented
9 years ago Original comment by mgarrish on 28 Mar 2013 at 8:56
GoogleCodeExporter
commented
9 years ago I would be more comfortable to use a property in OPF rather than CSS for this
purpose for the following reasons:
1. CSS may or may not define such property in future. Such an unknown extension
should be done in OPF rather than proprietary extensions to CSS.
2. CSS Page is still under big discussion and is still very early stage for how
to define in browser (rather than PDF generation software.) It's safer not to
rely too much on such an early spec.
3. Implementation-wise, this is done during paging calculation rather than
layout-within-pages, and therefore RS is easier to handle if defined in OPF.Original comment by kojii...@gmail.com on 31 Mar 2013 at 8:37
GoogleCodeExporter
commented
9 years ago [deleted comment]By the method which Brady wrote, it does not operate well.
<html>
<head>
<style>
.vert {-epub-writing-mode: vertical-rl;}
p {margin-left: auto; margin-right: auto;}
</style>
</head>
<body>
<p class="vert">ABC</p>
</body>
</html>
By the method which Brady wrote, it does not operate well.
A more complicated method has been tried, as below;
<html>
<head>
<meta charset="utf-8" />
body.hori {
-epub-writing-mode:horizontal-tb;
}
div.vert {
width:auto;
height:auto;
margin-right:auto;
margin-left:auto;
-epub-writing-mode:vertical-rl;
}
</head>
<body class="hori">
<div class="vert">
<p>ABC</p>
</div>
</body>
</html>
but even by this method, the text of the overflowing portion will be out of
display.
So, I proposed to specify this as a new function in 3.0.1.
I first thought of doing this in CSS, but now I agree to the way Ishii-san has
written: to use a property in OPF.
ex)
<spine page-progression-direction="rtl">
<itemref idref="XHTML001a" properties="vertical-align:middle" />
....
</spine>
Off course, a suitable prefix is needed.Original comment by ko...@voyager.co.jp on 1 Apr 2013 at 10:04
GoogleCodeExporter
commented
9 years ago I'm afraid this is a new feature request and it is too short to start the
discussion as 301.
I assume content creators will design pages based on this property, but the
pages will not be rendered correctly on existing EPUB 3 viewers. I don't think
the new property is existing viewer friendly.
The property behaves like as CSS properties, so its processing model should be
defined as well. Without evaluating the processing model, it is hard to judge
whether it is easy to implement or not, I think.
Original comment by tkanai...@gmail.com on 9 Apr 2013 at 4:45
GoogleCodeExporter
commented
9 years ago We're probably not using CSS property model but OPF property model, please stay
tuned for the proposal to come out to discuss technical details further.
This issue has been under discussion for more than two years among us, and I
hear a lot of complaints from many publishers and vendors. I don't think it's
too short. Also, we will take care of backward compatibility issue as much as
we can, at least to the level the new property won't make things worse.
Thanks for the opinion anyway.Original comment by kojii...@gmail.com on 9 Apr 2013 at 12:55
GoogleCodeExporter
commented
9 years ago After discussing with Koike-san and some back-and-forth, I'm almost back to
Koike-san's original proposal except a few minor differences.
The proposal is to add a itemref property:
Name: align-x-center
Description: The content document is layout as usual, then the resulting box is centered horizontally. If the content document consists of multiple pages, all pages are centered.
Usage: Package Document spine itemref element
Cardinality: Zero or one
Example:
<spine page-progression-direction="rtl">
<itemref idref="XHTML001a" properties="align-x-center" />
....
</spine>
FAQ:
Q: Why "vertical-align" is renamed to "align-x"?
A: In order to make it clear that the direction is physical, not logical.
Q: Why "center", not "middle"?
A: In order to align the value names in CSS. In CSS, "middle" indicates the
center of x-height due to historical reasons, and "center" is the center.
Q: Why use "-", not ":" as in Koike-san's original proposal?
A: Considering future possibility to add this as meta property, syntax is
aligned with "Specifying name-value pairs on the spine itemref element"
http://www.idpf.org/epub/fxl/#syntax-itemref
With this syntax, in future when needed, we could define a meta property named "align-x" with "center" value.
Q: Why physical, not logical?
A: To make calculation slightly easier, and also because there's no real use
cases to align center vertically.
There was a discussion in Japan that why this issue only came up in Japan.
Our best guess is that, when layouting a small object in a page with lots of margins,
it makes most sense to center horizontally, and bottom margin being larger than top.
In horizontal flow, this layout can be done by text-align:center with appropriate margins.
In vertical flow, CSS can't handle this property, and this property came in to help.
Q: Why not to define in CSS?
A: This property is a document property, not an element property, and
inheritance/cascading is not needed at all.
This property has nothing to do with layout within page boxes.
It is also easy to implement outside the rendering engine.
Such property is better defined in OPF than in CSS.Original comment by kojii...@gmail.com on 2 May 2013 at 7:49
GoogleCodeExporter
commented
9 years ago Not only in Japanese Books, Books in Traditional Chinese also designed in that
way.
I've made several books with "Naka-Tobira". Some of them positioned by margin
proportion, you have to do responsive design for multi-screen. And once reading
system changed their way to render, I have to adjust again.
align-x-center could do great help for ebook design and capability. I support
Koike-san and Ishii-san's proposal.
Adding align-y-center also do great help for book design in easy way.Original comment by bobbyt...@wanderer.tw on 2 May 2013 at 8:32
Attachments:
GoogleCodeExporter
commented
9 years ago 
Although I am sympathetic, I think that we should ask W3C CSS WG.Original comment by eb2m...@gmail.com on 6 May 2013 at 11:36
GoogleCodeExporter
commented
9 years ago A first draft of the prose has now been committed for review:
https://code.google.com/p/epub-revision/source/detail?r=4657
The rendered html can be accessed here:
https://epub-revision.googlecode.com/svn-history/r4657/trunk/build/301/spec/epub
30-publications.html
Please review the prose for accuracy and completeness, in particular the
informative note at the end. If any additional changes are necessary, please
note them in this issue.Original comment by mgarrish on 15 Jun 2013 at 5:55
GoogleCodeExporter
commented
9 years ago Noticed that the fragment identifier got stripped from the URL in #17. Direct
link to the rendered section is:
https://epub-revision.googlecode.com/svn-history/r4657/trunk/build/301/spec/epub
30-publications.html#layout-property-align-x-centerOriginal comment by mgarrish on 11 Jul 2013 at 2:33
GoogleCodeExporter
commented
9 years ago The second paragraph has been updated to the following, as decided on the
2013-07-11 telecon:
For reflowable content, Reading Systems that support this property must center
each virtual page.
https://code.google.com/p/epub-revision/source/detail?r=4686
Original comment by mgarrish on 12 Jul 2013 at 11:49
GoogleCodeExporter
commented
9 years ago Original comment by markus.g...@gmail.com on 17 Jul 2013 at 8:30
Original issue reported on code.google.com by
ko...@voyager.co.jpon 9 Apr 2012 at 7:38