- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Mon, 21 Nov 2011 08:15:58 -0800
- To: Florian Rivoal <florianr@opera.com>
- Cc: www-style@w3.org
On Mon, Nov 21, 2011 at 8:06 AM, Florian Rivoal <florianr@opera.com> wrote: > On Mon, 21 Nov 2011 16:50:59 +0100, Tab Atkins Jr. <jackalmage@gmail.com> > wrote: >> >> As Boris is saying, the 2.1 processing model for images in 'content' >> is to render them as an anonymous inline box in the element. They are >> then unsizeable by traditional means. >> >> Content 3 is *way* out of date here (if you're looking at the ED, it >> has that big obsoletion notice on it). > > Right, I saw that notice. But webkit (at least in its chrome incarnation) > follows this obsolete ED when it comes to images in non-pseudos, which is > causing some (admittedly minor) interoperability problem for Opera, since > we don't. > > I am fine considering this a bug in the spec and in chrome/webkit, but > I wonder if webkit people are interested in touching this as long as > css3-content is in its current state. Oh, huh, I didn't realize we supported that. We're buggier than you let on, though - we appear to *only* support the <image> argument for content on non-pseudos. Do you know who depends on this? We should fix this bug, even if it's just by jumping ahead and switching to the 'replaced <image>' grammar. >> The correct solution, when I >> or someone else has time to pick the spec up, is to add another value >> to 'content', something like: >> >> ::before { content: replaced url("foo"); } >> >> which would make the element into a replaced element. > > That sounds reasonable. Was css3-content likely to get an active editor > any time soon? Not in the near future. It's on mine and fantasai's lists, but we're both busy with other things right now and that's a pretty involved spec. ~TJ
Received on Monday, 21 November 2011 16:16:58 UTC