RDF Core WG Review of Requirements for XML Schema 1.1

RDF Core WG Review of 

  Requirements for XML Schema 1.1
  W3C Working Draft 21 January 2003
  http://www.w3.org/TR/2003/WD-xmlschema-11-req-20030121/

RDF Core's primary concerns on XML Schema 1.1 are that the maximum
number of constructs are named by URI references, that the identity
and equality relations are clearly defined and that relationships
between value spaces of datatypes are clarified.

Thank you

Dave, for RDF Core Working Group

-----

1. First class objects (RQ-23)
  http://www.w3.org/TR/2003/WD-xmlschema-11-req-20030121/#N400183

The lack of URIrefs for user defined simple datatypes presents a
substantial difficulty for the use of XML Schema datatypes in RDF and
OWL.

We strongly ask that:
   Either RQ-23 be raised to a requirement, or a new requirement of
   URIs for simple types is added.

Not speaking for the  WebOnt WG, but we note additionally that it has
postponed an issue on incorporating XML Schema complexTypes within OWL:
  http://www.w3.org/2001/sw/WebOnt/webont-issues.html#I4.3-Structured-Datatypes
which would also benefit greatly from unique URI references for each
user defined XML Schema datatype.  The reason for postponing was
given as:
  [[ XML Schema WG hasn't yet decided how XML schema
     components fit into URI space.]]
  -- http://lists.w3.org/Archives/Public/www-webont-wg/2002Oct/0157.html
so would probably also support this requirement being strengthened.


2. Systematic treatment of fundamental facets (RQ-24)
  http://www.w3.org/TR/2003/WD-xmlschema-11-req-20030121/#N40024F

Equality is a fundamental issue in the use of XML Schema datatypes in
the semantic web, particularly OWL, and to a lesser extent RDF.  

We ask that you would:

  Define the equality relation clearly.

    In particular whether xsd:anyURI and xsd:string may or may not be equal.

  Define the identity relation clearly.

    In particular we understand from XML Schema 1.0 that this is the
    same as the mathematical relation of quantitative equality.  The
    RQ24 wording tends to suggest that there is a difference.

We refer you to Jeremy Carroll's study of equality of XML Schema datatypes
  http://lists.w3.org/Archives/Public/www-archive/2002Nov/att-0092/02-index
for further analysis and information in particular:

  "Equality"
  http://lists.w3.org/Archives/Public/www-archive/2002Nov/att-0092/02-index#issue-equality



3. Add URI datatype (RQ-108)

We moderately supporting this since URIs are used distinguished from URIrefs.
Such as in xml:base attribute values?



4. Subclass relationships between value spaces

We would like to be able to make RDF subclass relationships between
the value spaces of XML schema datatypes (which are RDF Classes in
the RDF datatypes semantics).

We thus need clarity on where there is intersection on datatype value
spaces, in particular we have considered xsd:anyURI and xsd:string -
how can these be related?

Please consider this further discussion by Jeremy Carroll in
  http://lists.w3.org/Archives/Public/www-archive/2002Nov/att-0092/02-index
at "Value space of anyURI":
  http://lists.w3.org/Archives/Public/www-archive/2002Nov/att-0092/02-index#issue-anyURI-values


5. Conjunction of ranges

If an RDF property has two ranges that indicate datatypes (by
knowlege of classes) such as an xsd:integer and xsd:unsignedLong, is
the resulting conjuctive datatype an xsd:unsignedInt?  

This is a more minor issue that may be answered by the answer to the
subclass question 4 above.

Received on Tuesday, 25 March 2003 10:10:12 UTC