On Tue, 2003-09-02 at 16:32, Jan Grant wrote: > On Tue, 2 Sep 2003, Jeremy Carroll wrote: > > > Jan seems to argue with DanC for option (e) - that somehow or other > > datatyping subClassOf is extensional, whereas elsewhere it is intensional. > > Absolutely not. > > > I guess I find that really quite problematic. > > > > I seem to remember being isolated in defending the extensional semantics > > for subClassOf; I can't understand what (some of) the rest of you want - > > you got rdfs:subClassOf as intensional, but a least be consistent, and > > forget your extensional intuitions. > > My opinion is that the "best" solution is that a "datatyping > interpretation" would include additional axiomatic triples where > subClassOf relationships should exist. I'd thought that was option (e). Yes, that's what I meant by (e). > If it's not, make it (f) and I cast my vote for that: extensional > throughout, with the possibility of a DT interpretation including SCO > triples by fiat. -- Dan Connolly, W3C http://www.w3.org/People/Connolly/Received on Tuesday, 2 September 2003 21:54:00 EDT
This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:59:53 EDT