Re: aspect-ratio property

On Wed, Oct 6, 2010 at 5:17 PM, Tab Atkins Jr. <jackalmage@gmail.com> wrote:

> [From another thread]
>
> On Wed, Oct 6, 2010 at 5:03 PM, fantasai <fantasai@inkedblade.net> wrote:
> > On 10/06/2010 01:56 PM, Tab Atkins Jr. wrote:
> >> I'm also curious just in general how much we should pay attention to
> >> horizontal block flow.  For my aspect-ratio draft, I'm currently just
> >> making it always a width/height ratio.  Should I pay attention to
> >> block flow and instead make it an inline-flow-dimension /
> >> block-flow-dimension ratio?
> >
> > In general, all CSS3 drafts should be written to be well-defined for
> > both horizontal and vertical writing modes.
>
> The draft is well-defined, yes.  It specifically calls out the fact
> that, when both width and height are underdefined, either the width
> *or* the height will be resolved first (width first in the default
> block flow, height first in a horizontal block flow), and the
> interactions of aspect-ratio are well-defined in either case.
>
> But should I instead define the ratio with logical dimensions, such
> that a ratio of "2/1" means "twice as wide as it is tall" in a
> vertical block flow, but "twice as tall as it is wide" in a horizontal
> block flow?  Is there a general rule?  Just best judgement?
>
> I'm just not sure what's "smarter" here.  If I change block-flow, do I
> expect elements to rotate their dimensions significantly as well?  Is
> the answer the same for a <video> and a <p>?  Should I have a flag
> optionally switching it from physical to logical?
>

I think providing just one of these is sufficient for now. I don't feel
strongly as to whether it should be physical or logical here. I can see
use-cases for both. Although, most use-cases I can think of involve images
or video elements, which don't rotate in vertical writing mode. So, physical
makes more sense to me here.

The original link to the draft no longer works. Here's the new one:
http://www.xanthir.com/blog/b4810.

This seems like a great addition. It has clear use-cases and is easy to
implement.

Received on Tuesday, 18 October 2011 03:33:40 UTC