- From: Jonathan Borden <jonathan@openhealth.org>
- Date: Thu, 30 May 2002 17:33:30 -0400
- To: <connolly@w3.org>, "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>
- Cc: <phayes@ai.uwf.edu>, <www-webont-wg@w3.org>
Peter F. Patel-Schneider wrote: > > From: Dan Connolly <connolly@w3.org> > Subject: Re: layering (5.3,5.10): a same-syntax model theory > Date: 30 May 2002 15:16:19 -0500 > > > On Thu, 2002-05-30 at 14:37, patrick hayes wrote: > > [...] > > > > Whoa. If this really is a same-syntax extension then in order for > > > this to be meaningful you need to show how recursion can be expressed > > > in RDF (good luck). > > > > Huh? In the DAML+OIL model theory, the prose appeals > > to all sorts of traditional set theoretic notions > > when expressing constraints on interpretations... > > it uses stuff like |{...}| to denote the > > cardinality of sets and such. > > Yes, but the DAML+OIL model theory conditions work on the n-triples, i.e., > on the syntax, where things are much nicer. You may remember that the > axiomatic definition had to be modified a few times to take care of these > subtle points. > Someone had explained to me that the DAML+OIL model theory works _as if_ every triple is "dark", and that since, at the time it was written, there was no RDF MT, that this was an entirely reasonable interpretation. Jonathan
Received on Thursday, 30 May 2002 18:21:58 UTC