Re: [CSS21] Please endorse xml:id

On Fri, 1 Jul 2005, Norman Walsh wrote:
>
> / Ian Hickson <ian@hixie.ch> was heard to say:
> | I haven't yet seen any technical reason for requiring xml:id support of 
> | CSS implementations. It doesn't improve _CSS_ implementations at all.
> 
> If vendor A has a CSS-for-XML product that supports xml:id and vendor B 
> has a CSS-for-XML product that does not, then id selectors for 
> well-formed XML documents are not interoperable across those products.

Sure. Same applies to XML namespaces and XML Base, or XHTML. For example 
if vendor A has a CSS-for-XML product that supports XHTML and vendor B has 
a CSS-for-XML product that does not, then :link and :visited selectors for 
well-formed XHTML documents are not interoperable across those products.

But that doesn't mean that CSS should require XHTML. The same argument 
could be made for XLink, for SVG, for XForms, etc.

I don't see why xml:id is special here.


> Chris asked me to make a comment along these lines in response to an 
> email exchange that we had on the public-xml-id comments list. I have 
> done so. I gather that the CSS WG's consensus is to reject the comment. 
> I have neither the energy nor the inclination to pursue this further so, 
> if and when you do reject the comment, you may record, as a matter of 
> process, that I am satisfied with that resolution.

Okie dokie. Thanks.


> CSS makes explicit reference to XML IDs and consequently I think it 
> should recommend support for xml:id. If xml:id made explicit reference 
> to styling, I believe I'd be entirely comfortable saying that CSS should 
> be supported.

xml:id does make explicit mention of CSS, in Appendix C. Indeed, xml:id's 
mention of CSS is longer than CSS's mention of xml:id. :-)

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Friday, 1 July 2005 17:07:02 UTC