- From: Dan Brickley <danbri@w3.org>
- Date: Mon, 21 Oct 2002 05:42:44 -0400 (EDT)
- To: Dave Beckett <dave.beckett@bristol.ac.uk>
- cc: Patrick Stickler <patrick.stickler@nokia.com>, w3c-rdfcore-wg <w3c-rdfcore-wg@w3.org>
On Mon, 21 Oct 2002, Dave Beckett wrote: > > >>>Patrick Stickler said: > > > > > > Seeing as how datatyping is now a part of the RDF core > > rather than a part of RDFS, should we change rdfs:Datatype > > to rdf:Datatype? > > We have to tread as carefully as possible since we want to be able to > justify keeping the rdf: namespace URI the same. If it seems the rdf > namespace gets too many new semantic terms (Properties and Classes) > then the namespace looks increasingly like it would have to change. > > At present, we've removed some syntax things - rdf:aboutEach* > and added some new syntax things - rdf:nodeID, rdf:datatype, so > we are on the line. > > If we were changing the namespace(s), we'd surely split the rdf/xml > syntax terms / semantic terms completely. But not in this round of > specs and not at this stage in document writing. > > Just my thoughts. I agree entirely. Goofing around with the 1999 namespace too much (for some measure of excess) is impolite. I think we can get away with it a little since the M+S spec didn't say much one way or the other about the contents of its namespaces. Putting new classes and properties into RDFS's namespace is cleaner, changing '99 only to reflect syntactic fixes to RDF/XML. Dan -- mailto:danbri@w3.org http://www.w3.org/People/DanBri/
Received on Monday, 21 October 2002 05:42:49 UTC