W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > March 2003

RDF schema for XML datatypes?

From: Dan Connolly <connolly@w3.org>
Date: 14 Mar 2003 11:51:38 -0600
To: w3c-rdfcore-wg@w3.org
Message-Id: <1047664298.25644.73.camel@dirk.dm93.org>

EricM and everybody,

as part of last call review of our datatype design,
did we ask the XML Schema WG to make an RDF schema
available at http://www.w3.org/2001/XMLSchema-datatypes
and/or at http://www.w3.org/2001/XMLSchema ?

I think we should; i.e. W3C should endorse the
use of these names as datatypes by putting
some RDF there that says they're datatypes.

Something like what Patrick put together...

http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Nov/0654.html
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Nov/att-0654/01-XSD.rdf

I still prefer to think that datatype subtyping
works more like subProperty than like subClassOf;
e.g.
  xsdt:long rdfs:subPropertyOf xsdt:integer.
is a better way of looking at it than
  xsdt:long rdfs:subClassOf xsdt:integer.

But I'm not all *that* uncomfortable using
datatypes both as properties (whose extensions
relate values to lexical forms) and as classes
(of values).

Note XQuery is using
  http://www.w3.org/2001/XMLSchema-datatypes
as a namespace prefix for a vocabulary of constructors.
http://www.w3.org/TR/xquery-operators/#namespace-prefix

These constructors act an awful lot like RDF properties,
if you ask me.

consider...


8.4.7.1 Examples
      * fn:get-year-from-dateTime(xs:dateTime("1999-05-31T13:20:00-05:00")) returns 1999.
        

it looks very natural as...

  [ is xs:dateTime of "1999-05-31T13:20:00-05:00" ]
    fn:get-year-from-dateTime 1999.

(assuming N3 grows integer literals).



-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Friday, 14 March 2003 12:50:32 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:56:14 EDT