- From: Shadi Abou-Zahra <shadi@w3.org>
- Date: Mon, 17 Nov 2008 14:01:42 +0100
- To: Johannes Koch <johannes.koch@fit.fraunhofer.de>
- CC: ERT WG <public-wai-ert@w3.org>
Hi Johannes, OK, I get you now and I agree too. It is indeed clearer and is simpler for sub-classing. For example, you can create a my:probablyPass class from earl:cannotTell pretty easily because it is defined as a class. This is an essential functionality. Thanks, Shadi Johannes Koch wrote: > > Hi Shadi > > Shadi Abou-Zahra schrieb: >> >> Hi Johannes, >> >> Johannes Koch wrote: >>> >>> Shadi Abou-Zahra schrieb: >>> >>>> Carlos A Velasco wrote: >>> >>>>> 1) Whether we make the schema "pure" RDFS. We can eliminate the OWL >>>>> "thingies." We could make all instances from Outcome and TestMode >>>>> simple >>>>> RDF Resources. >>>> >>>> I thought we had come to this conclusion but I may be wrong. >>> >>> On the one hand, the range of earl:outcome makes clear that every >>> object in a triple with an earl:outcome predicate is of type >>> earl:Outcome. However, for clarity, I would make "#cannotTell" et al. >>> explicitly an earl:Outcome, as it is in the version CarlosV sent. >> >> I don't know what you mean. > > In Carlos' current schema, we have: > > <earl:Outcome rdf:ID="cannotTell"> > ... > </earl:Outcome> > > > which means > > a resource of type earl:Outcome. > > However each resource that is used as an object in a triple with > predicate earl:outcome is _by the range of the earl:outcome property_ of > type earl:Outcome. So it's not _necessary_ to give the predefined > outcomes a type earl:Outcome. So it might also be > > <owl:Thing rdf:ID="cannotTell"> > ... > </owl:Thing> > > or > > <rdf:Description rdf:ID="cannotTell"> > ... > </rdf:Description> > > But I think, that it's clearer to be explicit. > -- Shadi Abou-Zahra - http://www.w3.org/People/shadi/ | WAI International Program Office Activity Lead | W3C Evaluation & Repair Tools Working Group Chair |
Received on Monday, 17 November 2008 13:02:17 UTC