W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > November 2001

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

From: <Patrick.Stickler@nokia.com>
Date: Wed, 21 Nov 2001 09:15:04 +0200
Message-ID: <2BF0AD29BC31FE46B78877321144043114C0BF@trebe003.NOE.Nokia.com>
To: phayes@ai.uwf.edu
Cc: Graham.Klyne@MIMEsweeper.com, w3c-rdfcore-wg@w3.org
> >This is the interpretation that is suggested in my
> >recent recommendation
> >
> >http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Nov/0579.html
> >
> >In a nutshell, it is the "object slot" of the statement
> >that denotes the value, not the "thing" that is in the
> >slot (literal, uriref, or bNode).
> 
> 1. That isn't a new interpretation; all the proposals that require 
> the MT extension (P and P++ of those on the table, and possibly X) 
> all assume that the graph node itself is the primary denoting entity. 
> They have to, since the same literal label might have different 
> meanings in different contexts. That is why they require that RDF 
> graphs not be restricted to be tidy on literal nodes.

Great. Then at least that is agreed upon.

> 2. That isn't what Graham is talking about here (I think).
> 
> >  > [Later]
> >  >
> >>  It also occurs to me that we could have three parallel
> >>  namespaces; e.g.
> >>
> >>      xsd:integer a rdf:Class.     (Lexical space of decimal
> >>  integers, per
> >>  XML schema)
> >>      xsdv:integer a rdf:Class.    (Value space of integers,
> >>  per XML schema)
> >>      xsdp:integer a rdf:Property.
> >
> >And precisely who is the authority defining these classes
> >which have meaning only to RDF
> 
> Obviously, the RDF Core WG. It's our language, and we can say 
> what it means.

I think this is opening up a can of worms...  insofar as
it implies that anyone who wishes to use some other data type
scheme other than XML Schema basic types has to do more than
just allocate a URI for each type, but must allocate three
different URIs -- and users have to then keep clear in their
heads the nuances of meaning for each.

Not a very appealing scenario...

> >, and how does that sync up
> >with the use of XML Schema simple class usage outside of
> >RDF, and does that mean that *every* data type scheme must
> >define three parallel taxonomies?
> >
> >This is what I mean by the S proposal not being fully
> >understood as far as implications to the big picture.
> 
> The S proposal does not involve the introduction of new 
> vocabulary in this way.

So exactly what are xsdv:integer and xsdp:integer, if not
new vocabulary terms?

Patrick
Received on Wednesday, 21 November 2001 02:15:21 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:42:45 EDT