Re: [XHTML 2] 25.2 role and QNames in attribute context (PR#7757)

"Steven Pemberton" <steven.pemberton@cwi.nl>
> The advantages of QNames in this context are many:
>
> 1. Classical usage (such as rel="index") can stay the same.

having 2 attributes, rel and relNS also supports this if relNS is defaulted 
the same as no qname.

rel="index" relNS="

> 2. It is extensible, so that you don't have to reissue the language for 
> new values.

rel and relNS supports this.

> 3. It allows the user or communities to add new values without fear of 
> clashing with other users or communities (and at least two communities, 
> WAI and DI, are already doing this).

rel and relNS supports this.

> 4. It allows you to use existing values from other communities, such as 
> rel="dc:rights".

rel and relNS supports this.

> 5. Standard methods of discovering what a value means have been developed.

rel and relNS support this.

> 4. It is low-hurdle, since there is no need to register values with a 
> central authority.

rel and relNS support this.

> 4. It is a relatively simple mechanism, widely used in W3C technologies.

QNames _in attribute context_ are not simple, and are not widely used, I 
cannot find a single W3 Recommendation that uses QNames in attribute 
context.

> It is not true that the browser has to support XML Schema in order for it 
> to work.

Browsers currently support XSLT, to use XSLT with QNames in Attribute 
context the browsers either need specialist knowledge of the doctype, or 
they need support for XML Schema.

> All the browser has to know, if it needs to do anything special with these 
> values at all, is that the attribute is a QName.

Please stop using QNames in attribute context, rel and relNS attributes 
achieve this and meet all your above advantages. Alternatively change XSLT 
to support QNames in attribute context, XHTML is an XML language and must be 
compatible with XSLT, it is currently not.

Cheers,

Jim.

Received on Thursday, 26 January 2006 12:27:40 UTC