- From: Robin Berjon <robin.berjon@expway.fr>
- Date: Tue, 26 Apr 2005 18:49:54 +0200
- To: Ian Hickson <ian@hixie.ch>
- Cc: "Henry S. Thompson" <ht@inf.ed.ac.uk>, www-style@w3.org, Norman Walsh <Norman.Walsh@Sun.COM>, public-xml-id@w3.org
Ian Hickson wrote:
> On Tue, 26 Apr 2005, Henry S. Thompson wrote:
>>That would bring it in to line with XML Namespaces 1.1 in a useful way.
>
> What's the use case? This seems like it would complicate the spec and the
> implementations, as well as make full tutorials slightly more complicated,
> without any real benefit to the end user. When would you ever use this
> namespace in CSS?
The use case is consistency. I know you disagreed last time we discussed
this but I would expect anyone using namespaces in CSS selectors to have
an existing experience with XML namespaces (since they predate by a fair
margin). Said user would therefore be confused if the way in which
namespaces are declared in CSS differed from that in which they are
declared for XML. This applies to rules for default namespaces just as
it should apply to the rules for the xml prefix:
- the xml prefix cannot be bound to anything other than
'http://www.w3.org/XML/1998/namespace';
- the 'http://www.w3.org/XML/1998/namespace' namespace IRI cannot be
mapped to any prefix other than 'xml'; and
- said binding can be defined, but needn't be.
I agree that the use cases are limited. For xml|id you have #id, for
xml|lang there's :lang(), and I don't see anyone matching anything to
xml|base. Remains the case of xml|space which could see some rules
regarding white space handling be bound to it, but that's marginal. I do
think however that adding yet another discrepancy between specs by not
applying these rules would only create confusion.
--
Robin Berjon
Research Scientist
Expway, http://expway.com/
Received on Tuesday, 26 April 2005 16:50:04 UTC