RE: RDDL for names (was Re: XML Schema built-in data type namespaceURI.)

[Jonathan Borden <jborden@mediaone.net>]

Henry S. Thompson wrote:

>
> The concept is the same, but that's _only_ because they both are
> connected to http://www.w3.org/TR/xmlschema-2/#unsignedInt.

We agree. The context of my issue with this is the failure of an explicit
identification of the qname "xsd:unsignedInt" with a URI reference. If this
identity were in place (i.e. that we define the qname as a shorthand for the
URI) then we can allow 'URI calculus' to assert these equivalences.


>
> To try to bring your and Simon's threads together, I think what we're
> struggling with is the difference between the concept "unsignedInt" as
> defined by the W3C XML Schema spec. and various representations and/or
> manifestations of it, and then trying to generalise from this
> particular case to the general case of elements and attributes and
> types in any namespace.
>
> There are in general a _range_ of formal things you may want to say
> about some concept:  a W3C XML Schema simple type definition for it;
> an RDF Schema description of it; a Java implementation of it; . . .
>
> So it's clear that Simon was on the right track in one sense -- we've
> got the same issue at the level of individual names that RDDL is
> intended to solve for namespaces, so why not extend RDDL to handle
> this as well?
>
> What this would mean is that not only does
> e.g. http://www.w3.org/2001/XMLSchema resolve to (_inter alia_) an
> RDDL document, but that http://www.w3.org/2001/XMLSchema#unsignedInt
> resolves to a <resourceSubdirectory> (to make something up) in that
> RDDL document which in turn identifies the various sub-parts of its
> <resources> which correspond to this particular name.

This is an important idea. Actually every namespace is a collection. The
former namespace name is a URI proper and might resolve to a RDDL document.
The latter is a URI reference, and hence its own namespace. What ought it
resolve to? Would <div id="unsignedInt"> be a reasonable choice? Such an
<xhtml:div> element would contain a collection of <rddl:resource> elements
in the same manner as a RDDL document contains a collection of rddl:resource
elements.

RDDL currently provides for rddl:resources being children of div. (for
example how all the XHTML DTD modules are contained), but I think this usage
of <div> is important enough to merit explicit discussion in the next draft.
Not only does it solve this issue with XML Schema, but more generally the
issue of namespace names which are URI references having non-null fragment
ids.

-Jonathan