- From: Dan Connolly <connolly@w3.org>
- Date: Tue, 09 Oct 2001 14:38:35 -0500
- To: Pat Hayes <phayes@ai.uwf.edu>
- CC: w3c-rdfcore-wg@w3.org
Pat Hayes wrote: > > >Pat Hayes wrote: > > > >[...] > > > >>If I thought that 'literal' meant simply 'character string'. I > >>would agree with him (and I suspect, you), but I have never thought > >>that it did mean that. > > > > > >In RDF 1.0, there are no integer literals, if that's what you mean. > > > > My current problem is that in some extensions of RDF there *are* > things like integer literals, not unless they use parseType="integer" or some such. If you're speaking of DAML+OIL, I don't believe it has integer literals either. > though, so I want the RDF model theory > not to actually break when those kinds of things are added to it. On the contrary: the RDF model theory should be 100% clear that RDF 1.0 has no integer literals. > If > we incorporate very strong assumptions about what RDF literals must > be, extensions will have a lot of trouble. Yes, extensions will have a lot of trouble: RDF 1.0 software doesn't grok integer literals. I see no reason to hide this fact. > It would be better if we > could at least provide a kind of escape hatch for extended languages > to use richer collections of literals. I disagree; I think it's important to be clear about what RDF 1.0 is and what it isn't, and the sooner the better. Perhaps it's time to design RDF 1.x or RDF 2.0 that doesn't have these limitations. But we can't do that (inside a W3C WG) until we're done with RDF 1.0. -- Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Tuesday, 9 October 2001 15:38:38 UTC