W3C home > Mailing lists > Public > www-style@w3.org > July 2012

RE: [css3-writing-modes] column progression direction

From: Koji Ishii <kojiishi@gluesoft.co.jp>
Date: Sat, 7 Jul 2012 01:15:54 -0400
To: Alan Gresley <alan@css-class.com>, fantasai <fantasai.lists@inkedblade.net>
CC: "www-style@w3.org" <www-style@w3.org>, CJK discussion <public-i18n-cjk@w3.org>
Message-ID: <A592E245B36A8949BDB0A302B375FB4E0D5E51D173@MAILR001.mail.lan>
> >> Alex Mogilevsky raised an issue in a discussion awhile back:
> >> specifically, what should the column progression direction be if the
> >> parent of the multi-column element belongs to an orthogonal flow with
> >> a block flow direction opposite to the multi-col element's inline direction?
> >>
> >> E.g. suppose I have a vertical Japanese document:
> >>
> >>
> >>    | | | | | V
> >>    | | | | | V
> >>    | | | | | V
> >>    | | | | | |
> >>    | | | | | |
> >>      <====
> >>
> >> Then I insert a horizontal (LTR) multicol element
> >>
> >>>>> -- >>>-- | | | | | V
> >> ----- ----- | | | | | V
> >> --A-- --B-- | | | | | V
> >> ----- ----- | | | | | |
> >> ----- ----- | | | | | |
> >>                  <====
> >>
> >> Is column A or B first in the logical order?
> >
> > This is a very interesting question. I think the simple answer is "column A is first."
> 
> I agree with Koji here but it makes a mess of overflow and this is why I have
> previously stated on this mailing list that a concept of logical block progression be
> introduced [1] [2]. It is only in Old Mongolian that that you don't have this conflict of
> inline overflow and block overflow [3].
> 
> 1. http://lists.w3.org/Archives/Public/www-style/2010Oct/0645.html
> 2. http://lists.w3.org/Archives/Public/www-style/2011Jun/0199.html
> 3. http://lists.w3.org/Archives/Public/www-style/2011Jun/0756.html

As I read Alan and Leif's responses, and gave more thoughts on other scenarios such as paged media, I guess I'm changing my opinion.

By now, I think the consensus is that both cases exist, depends on situation and author's intention.

When the "column A is first" type of multi-column element spans multiple pages in a paged media, it'll change page progression as well. So I think it's natural to say that, even when the element fits within a page, it is actually changing the page progression of the element.

We discussed a bit before about how to mix multiple page progressions in a document, and we gave up defining it in Writing Modes Level 3 because it's too complex.

Could we define column progression follows page progression, and postpone the definition of how to mix multiple page progression to future levels?

This gives the different answer than before; "column B first" in this example. It doesn't fulfill all the use cases, but I don't think supporting all use cases in this area is urgent. I still believe, from use case perspective, "column A first" is more common than "column B first," but the former requires us to resolve multiple page progression issue. Because the commonness of both use cases aren't high anyway, we should take simpler approach for now and extend in future if needed.

I have no idea which is simpler in terms of implementations. If that contradict, we might need to reconsider again, but if "B first" is simpler for implementations as well, I think that's the good way to go.


Regards,
Koji
Received on Saturday, 7 July 2012 05:13:05 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:56 GMT