RE: XSLT2 draft (16th August) XHTML output

> > 20.2 XHTML Output Method
> 
> The XHTML output description specifies that the /> form
> should not be used if the declared content model (by 
> implication, in XHTML 1.0 DTD) is non empty. However the 
> first bullet simply says that a space should be placed before 
> the /> for "empty elements" which would seem to include all 
> empty elements rather than just those declared EMPTY in XHTML.
> 
> The first bullet point requiring a space before /> should
> probably be explictly restricted to elements in the XHTML 
> namespace with which are declared EMPTY in (one of) the XHTML 1.0 DTD.

Yes, I think this is essentially a question of editorial clarification:
perhaps the second bullet point should come before the first.
> 
> Would it be possible for the XHTML output method to force
> that the xhtml namespace is default and so output XHTML 
> elements unprefixed. If the aim is to work in legacy HTML 
> browsers outputting as <kjfkcwwa:br /> which would be allowed 
> by the current rules will not realy work.

This seems a reasonable suggestion. I'll register it as an issue so it gets
onto the WG agenda. It's not trivial to specify this, because namespace
prefixes are allocated as part of the namespace fixup process during result
tree construction, not during serialization. It may be simplest to express
it in the form of a suggestion to implementors.
> 
> similar comment on the fith bullet
>   "If there is a head element"
> does this mean any element with local name "head" or an
> element with local name name  "head" and namespace uri 
> "xhtml-namespace".
> 
I think it means an element with local name "head" in the XHTML namespace,
and will add a clarification to that effect unless anyone in the WG objects.

Thanks for the comments, as always.

Michael Kay

Received on Wednesday, 28 August 2002 11:28:19 UTC