Re: XHTML m12n XSD

[Rick Jelliffe <ricko@allette.com.au>]

From: Bullard, Claude L (Len) <clbullar@ingr.com>

> What would be the role of the RDF would
> in the XHTML schema?

Some kind of RDF Schema, for example.

> Why was the section on XLinks not done?  Time
> or conceptual problems?

Just division of labour.

> I think the XHTML Schema should be a top priority.
> Really, it might help XHTML become popular. :-)

How kind. Well, *if* that use of <redefine> turns out to be OK, it may be
better at helping XML Schemas become popular: I think the M12n work is
really good for exposing the kinds of things Schema languages like XML
Schemas and RELAX would do well to support (Schematron is not in that list,
because we already have phases and abstract rules in 1.5 which will do a lot
of what is needed, and because I have had no requests to move Schematron to
"completeness" or "universality"; I personally would prefer to right-size it
for small validation, so modularization is not important.  But I saw an XML
Schema business schema with more than 5000 types here in Taiwan last week!
That is a different world to markup languages.

 > From: Sean B. Palmer [mailto:sean@mysterylights.com]
>
> In conclusion, an XML Schema version of m12n is at best extremely
> experimental, and makes for a volatile subject.

I concur with Sean. But I don't know if the implication that m12n can only
be done by DTDs or their exact duplicates-in-XML is correct: XHTML M12n has
an abstract model, and to some extent an XHTML M1n2 in XML Schemas (X24s ?
:-) may also reveal where that model is not abstract enough.  Perhaps the
reasonable test is this: is there enough information in an X24s schema to
generate a good XHTML M12n DTD using, e.g. XHTML?   I think it looks more
like it can be so than it cannot be so.

Cheers
Rick Jelliffe