Re: integration testing of xml:ID

On Fri, Jun 10, 2005 at 11:25:16PM +0200, Chris Lilley wrote:
> On Friday, June 10, 2005, 11:11:34 PM, Bjoern wrote:
> BH>   If an implementation supports CSS 2.1 and also supports
> BH>   xml:id, then the xml:id attribute must be treated as an
> BH>   ID for the purposes of CSS selector processing.
> 
> I must say that this is exactly the sort of conditional conformance
> requirement that i would like to see (and the same for DOM and XPath).

  At the CSS-2.1 level, yes. At the xml:id level, no. You are mixing
layers, this is confusing as I pointed out in a separate mail, and
won't reach the right audience. If we try to add any such boilerplate for 
any kind of spec based on XML which may have used attributes of type ID
this will be a spec with 4 pages of content and 50 pages of warnings
for each and every kind of applications built on top of the parser.
The introduction address all of those in generic terms.

> BH> to CSS 2.1 would in some way change conformance requirements for
> BH> implementations that support CSS 2.1 and also support xml:id? If
> BH> it does, could you please clarify why and why xml:id does not
> BH> contain a conformance requirement to this effect already?
> 
> The same wording could usefully be added to, for example, CSS3 as well.

  That is where it would make sense, as well as a note for example in
the CSS-2.1 errata. It will reach the right audience and will be effective
there, not in the xml:id spec.

Daniel

-- 
Daniel Veillard      | Red Hat Desktop team http://redhat.com/
veillard@redhat.com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/