- From: Graham Klyne <GK@Dial.pipex.com>
- Date: Tue, 02 May 2000 17:20:52 +0100
- To: "McBride, Brian" <bwm@hplb.hpl.hp.com>
- Cc: "RDF Interest (E-mail)" <www-rdf-interest@w3.org>
At 02:25 PM 5/2/00 +0100, McBride, Brian wrote: > >As Brian has already demonstrated, the XOR aggregator is not > >sufficient > >even for case (a). The immediately evident problem here is > >that (X XOR Y > >XOR Y) == X, for all Y. So how about using simple addition, > >modulo 2^n, as > >an aggregator? > > > > >There is a small subtlety here. An RDF model is a set of statements. >Sets have no duplicates, so no statement can appear twice. >Cancellation is therefore less of a risk. It was my understanding that while there could be no duplicate resource names, there could be duplicate instances of a given property between the same subject/object pair. I think it would certainly be possible to describe this using multiple reified statements with different statement resource IDs. #g ------------ Graham Klyne (GK@ACM.ORG)
Received on Tuesday, 2 May 2000 13:32:45 UTC