Re: subgraph/entailment

On Wed, Sep 07, 2005 at 01:37:48 -0400, Bijan Parsia wrote:
> I fear that people will quite naturally expect  "?p rdf:type 
> rdf:Property" to simply *mean* "fetch me all the properties mentioned 
> in this document". It's very intuitive to read it that way and it's a 
> useful query.

I agree, but I dont think youre going to get that from many compliant
implementations. 3store, in its standard compliance mode will require

	GRAPH [] { ?p rdf:type rdf:Property . }

c.f. complaints from me about the implications of the default graphs
situation. If you store inferred triples like that in the background graph
then you have no way to determine thier provenance and/or it breaks
orthogonality with the other grpahs.

- Steve

Received on Thursday, 8 September 2005 08:53:57 UTC