- From: Jos De_Roo <jos.deroo.jd@belgium.agfa.com>
- Date: Fri, 25 Oct 2002 11:11:32 +0200
- To: "Brian McBride <bwm" <bwm@hplb.hpl.hp.com>
- Cc: "Jeremy Carroll" <jjc@hplb.hpl.hp.com>, "pat hayes" <phayes@ai.uwf.edu>, w3c-rdfcore-wg@w3.org, w3c-rdfcore-wg-request@w3.org
>>I guess it's the one from
>>http://lists.w3.org/Archives/Public/www-webont-wg/2002Sep/0308.html
>>
>>1:[[
>>
>>eg:prop rdfs:range eg:A .
>>eg:A rdfs:subClassOf eg:B .
>>
>>entails
>>
>>eg:prop rdfs:range eg:B .
>>]]
>
>Which looks clearly false to me, so I'd better ask to find out what I'm
>missing.
I don't support it either, but it could make sense
to have IFF (if and only if) semantics such that
Range(P,C) iff (forall x,y P(x,y) -> C(x))
and in that case rdfs:Resource would for instance
be the range of any property
>Consider:
>
> IEXT(A) = {a}
> IEXT(B) = {a, b}
>
>Then if I say that prop can take any member of A as its value, it can also
>take any member of B, because B happens to be a superclass of B.
>
>Wierd!
>
>Isn't it to stop that sort of thing happening that we switched domain and
>range to conjunctive semantics, i.e. so that an inferencing engine finding
>a range constraint would know that all values must be a member of that
>class; there is no way to add members to the range?
well, it all depends on the chosen semantics for range
and I support the oneway IF as well
-- ,
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/
Received on Friday, 25 October 2002 05:13:25 UTC