Re: Versioning and html[5]

Matthew Raymond schrieb:
> Chris Wilson wrote:
>> Dave Raggett [mailto:dsr@w3.org] wrote:
>>> IE conditional comments for HTML are better than selector hacks, but
>>> have to be placed in the HTML and there is currently no direct
>>> equivalent within CSS.
>> We did propose this in the CSS WG a couple of years ago, and got shot down.

Note that spec'ing syntax within comments doesn't really make sense. 
Conditional comments could work for CSS as they do for HTML.

>    There's a reason for that. Prototyping using the vendor-specific
> extension naming syntax works just fine. For instance, Mozilla has a
> prototype implementation of "border-radius" under the name
> "-moz-border-radius". When "border-radius" finally becomes part of a
> recommendation, Mozilla can support it even if it doesn't work the same
> as its earlier implementation because the names are different. In fact,
> they can use the declaration of "border-radius" as a flag to turn off
> the "-moz-border-radius" style.
> 
>    Also, CSS isn't version based to begin with. It's level based. But
> even if you consider the levels nothing more than another name for
> versions, how would versioning fix problems like the bugs in Internet
> Explorer's CSS Level 1 implementation?
> 
>    But if you really want switches in CSS for IE to use, why not the
> following?
> 
> | * { -msie-version: 7.0; }

Why the star selector? An at-keyword like @-msie-navigator seems 
appropriate.

--Dao