RE: Answer to the question: What is a "value" to RDF

> -----Original Message-----
> From: ext jos.deroo.jd@belgium.agfa.com
> [mailto:jos.deroo.jd@belgium.agfa.com]
> Sent: 16 November, 2001 23:43
> To: Stickler Patrick (NRC/Tampere)
> Cc: Graham.Klyne@MIMEsweeper.com; melnik@db.stanford.edu;
> phayes@ai.uwf.edu; w3c-rdfcore-wg@w3.org
> Subject: RE: Answer to the question: What is a "value" to RDF
> 
> 
> 
> [...]
> >
> > Am I the only one to see a conflict here? (though that may
> > not be so surprising ;-)
> >
> > I.e. what is the interpretation of the following knowledge:
> >
> >    x s:age "P12Y" .
> >    s:age rdfs:range xsd:duration .
> >    s:age rdfs:subPropertyOf xsd:duration .
> >
> > According to the S proposal, either range or subproperty
> > relations can define type. Which is it going to be? Both?!
> 
> 
> I think that in the S-idiom it is (just)
> 
>   x s:age [ xsd:duration "P12Y"] .

By which you mean, in NTriples:

   x s:age y .
   y xsd:duration "P12Y" .

But that is redundant, since the semantics of
xsd:duration have already be infused into the
semantics of s:age.

With regards to data typing semantics only, you are 
saying the equivalent of

   x xsd:duration y .          (from subPropertyOf)
   y xsd:duration "P12Y" .     (from range)

Such redundancy seems to point to some problems
that perhaps have not been identified yet with this
proposal.

> and
>   s:age rdfs:range [ is rdfs:domain of xsd:duration ] .
>
> and
>   xsd:duration rdfs:range [ is rdfs:domain of xsd:string ] .
> 
> (datatypes in the S-idiom are properties, that S-imple ;-)

This all seems quite clever and may be good fun, but it's
IMO a bit too radical a change to impose on the RDF community
and also totally blurs the distinction between data type
and property, which I don't consider to be correct.

Cheers,

Patrick

Received on Monday, 19 November 2001 05:36:51 UTC