Re: More on XSD in RDF

[Patrick Stickler, Nokia/Finland, (+358 40) 801 9690, patrick.stickler@nokia.com]


----- Original Message ----- 
From: "ext Dan Connolly" <connolly@w3.org>
To: "Brian McBride" <bwm@hplb.hpl.hp.com>
Cc: "Patrick Stickler" <patrick.stickler@nokia.com>; "ext Jeremy Carroll" <jjc@hplb.hpl.hp.com>; <w3c-rdfcore-wg@w3.org>
Sent: 25 November, 2002 18:15
Subject: Re: More on XSD in RDF


> On Mon, 2002-11-25 at 09:42, Brian McBride wrote:
> [...]
> >    o I suggest:
> > 
> >    - we use non controversial examples of schema datatypes in our test 
> > cases - stay away from the stuff that tests understanding of schema 
> > datatypes more than rdf
> 
> As I explained earlier, I disagree with shying away from
> contraversial issues that are relevant to our design, i.e.
> keeping them out of the test suite.

This has nothing to do with our design, but with ambiguity in
the XML Schema datatyping specs. I read them as saying that the
following assertions are not licensed:

   xsd:decimal rdfs:subClassOf xsd:float .
   xsd:float rdfs:subClassOf xsd:decimal .

i.e. "10.0"^^xsd:float  !=  "10.0"^^xsd:decimal

and if we're going to have test cases for the above, they should
be negative entailment tests, since IMO XML Schema doesn't
license the datatype entailment

IF
{
   thing:A some:Property "10.0"^^xsd:float .
   thing:B some:Property "10.0"^^xsd:decimal .
}
THEN
{
   thing:A some:Property _:x .
   thing:B some:Property _:x .
}

It is NOT the duty of the RDF Core WG to resolve the above
ambiguity or fix XML Schema datatypes so that applications can
fairly consider that "10.0"^^xsd:float  ==  "10.0"^^xsd:decimal,
even if it is relevant to the RDF community.

This seems more like an issue for the SW Coord. Group or the TAG
to resolve.

It is IMO enough if we simply say somewhere "here be dragons".

> If we're shy about these issues, let's get them out of
> our design; c.f. rdfs:format
> http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Oct/0031.html

Surely you're not proposing we throw out the typed literals and
go only with lexical constraints?!

Again, this has nothing to do with the RDF design. In fact, the present
datatyping solution provided by RDF is helping the XML community by
shining a light on these problemmatic issues and "encouraging" the
XML Schema folks to fix them.

Patrick

Received on Tuesday, 26 November 2002 02:24:27 UTC