- From: Ian Hickson <ian@hixie.ch>
- Date: Sun, 15 Feb 2004 18:42:08 +0000 (UTC)
- To: Boris Zbarsky <bzbarsky@MIT.EDU>
- Cc: www-style@w3.org
On Sun, 15 Feb 2004, Boris Zbarsky wrote:
>
> Would it be at all possible to have CSS2.1 explicitly say that this is
> not supported? Otherwise, by the time the relevant CSS3 module is ready
> there will be enough UAs out there supporting it and enough author
> expectations that it works that changing the behavior will effectively
> be impossible....
Well, the question is what do you think _should_ happen with replaced
content? I've been considering whether ::before and ::after nesting (as
described in Generated Content) should maybe just be removed, in favour of
a more generic technology like XBL being used. The question then becomes,
if we leave only the CSS2.1-level ::before/::after, wouldn't it be better
to allow that for replaced content?
One way of allowing it would be that once 'content' is set to replace the
element, then ::before/::after stop applying to the element itself, and
start applying to ::outside. So:
Current specification | Proposal
------------------------------------+------------------------------------
element { content: normal; } | element { content: normal; }
element::before { } | element::before { }
element::outside { } | element::outside { }
element::outside::before { } | /* not possible without XBL */
|
element { content: url(), normal } | element { content: url(), normal }
/* element::before meaningless */ | /* n/a */
element::outside { } | element::outside { }
element::outside::before { } | element::before { }
With the inheritance always going:
container
|
element
____|____
| |
::before ::outside
...whether ::before is really inside the element or the ::outside pseudo.
Anyway, all this is still undecided. Which is why CSS2.1 will not say
anything about it.
--
Ian Hickson )\._.,--....,'``. fL
U+1047E /, _.. \ _\ ;`._ ,.
http://index.hixie.ch/ `._.-(,_..'--(,_..'`-.;.'
Received on Sunday, 15 February 2004 13:42:11 UTC