- From: Sergey Melnik <melnik@db.stanford.edu>
- Date: Thu, 12 Jul 2001 15:01:14 -0700
- To: Brian McBride <bwm@hplb.hpl.hp.com>
- CC: RDFCore WG <w3c-rdfcore-wg@w3.org>
Brian McBride wrote: > > Thanks for completing the action. Now how does this affect our work plan? > [...] > Are there other issues which should belong in that group? For example, > Sergey has raised the question of whether namespaces are part of the model. > [...] > We also have a host of semantic questions regarding the model, and > they are all highly inter-related. How best to tackle those? > Here is a suggestion. > > We cannot tackle them piecemeal, one at a time, they are too inter- > related for that. In agreement with Frank Manola's > recent comments about discussing issues in the light of a specific > text, how about we begin with a strawman model theory > and discuss the issues in the light of that. Maybe one of our > logician's, Pat - you've mentioned model theory several times - could > produce a strawman. Brian, I completely agree that many of the modeling issues are interdependent. The approach I would suggest is to try to explore different, but self-contained candidates for the RDF model/a.s., and to see how much of the existing spec and applications would be affected by favoring one or the other approach. Each approach should explicitly address the issues to be resolved. I slightly revised the proposal that I made some time ago and made it available at http://www-db.stanford.edu/~melnik/rdf/formal-model.txt As far as I remember, Graham posted another proposal last month on the list. We could definitely use a couple of more starting points. > Would this be a good way forward with the technical discussions? Think so. Sergey
Received on Thursday, 12 July 2001 17:48:01 UTC