- From: Brian McBride <bwm@hplb.hpl.hp.com>
- Date: Thu, 21 Nov 2002 14:52:45 +0000
- To: "Jos De_Roo" <jos.deroo.jd@belgium.agfa.com>
- Cc: "RDFCore Working Group" <w3c-rdfcore-wg@w3.org>
At 14:31 21/11/2002 +0100, Jos De_Roo wrote: >[...] > > > If the test case is of the form > > > > <a> <b> "10"^^xsd:integer . > > > > datatype entails (xsd) > > > > <a> <b> "010"^^xsd:integer . > > > > and bearing in mind the we expect these to be machine checkable, do this > > suggest that a datatype aware RDF processor must generate the infinity of > > > different representations of the number 10. I might be wrong here - Jos > > seems to have done it without this problem. > >right, this is the traditional forward/backward chaining stuff >we do the latter and so start from the entailed graph and try >to find evidence for it in the premised graphs, which is quite >obvious here as the term "010"^^xsd:integer unifies with the >term "10"^^xsd:integer Aha! I see how you got away with it. A forward chaining reasoner such as cwm would have a problem, presumably. Brian
Received on Thursday, 21 November 2002 09:51:14 UTC