an unfilled need

[Rick Jelliffe]

 From: Tim Berners-Lee <timbl at w3.org>

>I haven't investigated this case just now as regards where all the
documents
>are, but the RDF work certianly assumed that the
>namespace URIs point to schemas and that is clear even from the RDF
Rec.

Certainly the RDF spec says that "RDF requires the XML namespace
facility
to precisely associate each  property with the schema that defines the
property."

However, because "XML syntax  is only  one possible syntax for RDF"
the property schemas which RDF speaks about cannot be structural
schemas (except accidentally) because their instances may lose their
structures during a transformation into some internal form, or even from
the full RDF syntax to the abreviated syntax. The RDF PR says
"A syntactic schema alone is not sufficient for RDF purposes."

A property schema for RDF can be just text descriptions (e.g., Dublin
Core) but it cannot be a vanilla DTD which only models
structure. Structural schemas are neither necessary nor sufficient.
"Namespaces are simply a way to tie a specific use of a word in
context to the dictionary (schema) where the intended definition is
to be found"

RDF Schemas give the purest indication of what the RDF WG means by
schema: semantic schemas  to provide typing (i.e., class/subclass
relationships) and not syntactical constraints per se.

For RDF, the grammar of RDF sets the structural schema from the
top-level (until we get to any "XML literals" in leaves). So in RDF
any property schemas that are identified in the namespace URI
cannot define the structure anyway (except, perhaps, in ways which
are irrelevent to RDF, for example to add extra constraints on
the order that properties in the element will be specified.)

When we see, in the RDF spec
<rdf:RDF
  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  xmlns:dc="http://purl.org/metadata/dublin_core#">
  <rdf:Description about="http://www.foo.com/cool.html">
    <dc:Creator>
      <rdf:Seq ID="CreatorsAlphabeticalBySurname">
 <rdf:li>Mary Andrew</rdf:li>
 <rdf:li>Jacky Crystal</rdf:li>
      </rdf:Seq>
    </dc:Creator>
  </rdf:Description>
</rdf:RDF>
the syntactic schema (content model) for dc:Creator is specified by
rdf:RDF not by Dublin Core.

If we are to look at the example of RDF, we get the idea
that it is the semantics *in neutrality to the  structural schema*
that is the information that is "associated" with a namespace URI.
RDF explicitly makes this association and explicitly prevents
associations with pure syntactical schemas.

RDF seems to endorse "a <P> is a <P> is a <P>" or at
least be neutral.

It may be said that RDF's use of namespaces for "XML literals"
in property values (e.g., a MathML formula) is an example
of where a syntactical schema is pointed to by the namespace
URI. However, RDF says that
"This document does not address how the characteristics of
properties are expressed" thus explicitly leaving questions of
the appropriate form for the declaration or association of
syntactical schemas unresolved, on the face of it.


Rick Jelliffe



xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)