Re: follow your nose, HTML widgets, GRDDL spec#issue-tx-elt, TAG issue xmlFunctions-34 [was: [RDFa] Use Case...]

On Tue, 2006-12-05 at 18:29 -0500, Ben Adida wrote:
> Dan Connolly wrote:
> > 
> > Oh. I sometimes get the impression that RDFa is intended not just
> > for HTML but as a mix-in for other XML vocabularies too.
> 
> It may be used in other vocabularies, yes, but those vocabularies will
> likely have to choose how to include it.

Ah. So it's sorta like xml:base in that any host language _may_
use it, but it doesn't automatically become part of all
host languages.

You might want to borrow some text from the XML base
spec and/or cite it as precedent.
 C Impacts on Other Standards (Non-Normative)
 http://www.w3.org/TR/2001/REC-xmlbase-20010627/#impacts

[...]
> it would definitely have to be part of the schema for the XML language
> you're thinking about. For example, it could well be done using a GRDDL
> transform specified in the namespace document.

In that case, there is some urgency; currently, the GRDDL
implementations I know about never fetch http://www.w3.org/1999/xhtml ;
they assume they know what's on the other end of that URI,
and in particular, that there are no GRDDL transformation
links there. We haven't baked that into the spec yet,
but we should soon, as we're aiming for last call
in the next month or so.

If that's going to change, it will impact GRDDL interoperability.

It would help a lot if you could update the XHTML namespace
document ASAP, or at least choose the URI of the transformation
and put an XSLT 1.0 implementation there.

I took a snapshot copy of Fabien's RDFa2RDFXML.xsl
and put it in the GRDDL test suite...
 http://www.w3.org/2001/sw/grddl-wg/td/RDFa2RDFXML.xsl
 http://www.w3.org/2001/sw/grddl-wg/td/testlist1#rdfa1

but I heard from Ivan that it produces different results
from Elias's parser. Did he report that problem? Hmm...
I can't find it in the -tf archives.

I suppose we can tweak our GRDDL implementations to do
late-binding on the XHTML namespace document. It's
a bit of a pain to work out a practical caching policy
that's not "don't fetch the XHTML namespace document",
but maybe it's The Right Thing.


> > GRDDL uses a namespace-qualified attribute on the root
> > element of an XML document as its hook, in the most general
> > case; and in the specific case of DTD-constrained HTML, it uses
> > the head/@profile hook.
> 
> Right, so you still need access to HEAD/PROFILE, right?

Well, either the profile of the source document or of its
namespace document.

>  Since that value
> of PROFILE will be different from one application to another, obviously
> (unless it's the RDFa transform, for example... but then we're back to
> our use case :)
> 
> -Ben
-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E

Received on Tuesday, 5 December 2006 23:59:01 UTC