W3C home > Mailing lists > Public > www-style@w3.org > April 2013

Re: [css3-regions] flow-into: <ident> content-only

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Fri, 19 Apr 2013 15:28:41 -0700
Message-ID: <CAAWBYDDJ2-3fK=h3C+xq3Wiy+sOqaRjZzxKigkg-ecz7kzBsUQ@mail.gmail.com>
To: Brad Kemper <brad.kemper@gmail.com>
Cc: Alan Stearns <stearns@adobe.com>, "www-style@w3.org" <www-style@w3.org>
On Fri, Apr 19, 2013 at 3:19 PM, Brad Kemper <brad.kemper@gmail.com> wrote:
> I think that would be odd, if the border was left behind, and an un-border-able image was sent into the flow. Which box would be affected by something like 'img { width:100px; }', and does the answer change if you had 'box-sizing:border-box'?

It's no odder than "<p style='flow-into: foo contents;'>foo</p>",
which leaves the border behind and flows an unborderable anonymous
text node into the flow.  We already have a concept of "things that
aren't boxes, but still live in the box tree", this would just be
adding a new one.

> Also, this would be inconsistent with the way ::after cannot add content inside the image's content box, because we don't treat that content-box as a container.

Yes, if we ended up adopting this concept, either ::before/::after
would start working, or we'd special-case them to not work (but maybe
allow other things to work, like setting <img
style="display:list-item;list-style-position:inside;">).

> I think it would be better if replaced elements just ignored the 'contents' element keyword, and just always acted as though it was 'element'.

Possible, yeah.

~TJ
Received on Friday, 19 April 2013 22:29:28 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:39:10 UTC