- From: Kang-Hao (Kenny) Lu <kennyluck@csail.mit.edu>
- Date: Thu, 15 Mar 2012 16:06:31 +0800
- To: WWW Style <www-style@w3.org>
- CC: "L. David Baron" <dbaron@dbaron.org>, ML publc-i18n-bidi <public-i18n-bidi@w3.org>
(Cc +public-i18n-bidi)
(12/03/15 6:46), L. David Baron wrote:
> The 'image-orientation' property defined in
> http://dev.w3.org/csswg/css3-images/#image-orientation should
> specify what images it applies to. Does it apply only to replaced
> elements, or does it apply to other images (e.g., background images)
> as well?
If 'image-orientation' applies to other images, say, at least images
specified with 'content', I think we should think about whether it's a
good idea or not to fold 'ltr/rtl' of 'image()' into a value of
image-orientation, given my concern about the current syntax of
directional images in [1]. In particular, if we make 'image-orientation'
inheritable (why is it not right now, by the way?), it be can naturally
inherited into '::marker', like
ol, ul {
list-style-image: url(arrow-ltr.png);
image-orientation: flip;
}
On the other hand, we can fold 'image-orientation' into the 'image()'
syntax somehow. The advantage of this is that Web authors can have full
control over which images (no matter it is a background-image or
border-image) 'image-orientation' applies. The obvious drawback is that
for the most common use case in the near feature
img { image-orientation: from-exif; }
has to be as complicated as something like
ima { content: replaced image(attr(src) from-exif) }
and seems a bit far from reality.
Or do we think these two use cases are too far in concept so they should
be addressed in different syntax?
[1] http://lists.w3.org/Archives/Public/www-style/2012Mar/0243
Cheers,
Kenny
Received on Thursday, 15 March 2012 08:07:08 UTC