- From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
- Date: Tue, 11 Dec 2001 10:51:04 -0000
- To: "Pat Hayes" <phayes@ai.uwf.edu>
- Cc: "Jeff Heflin" <heflin@cse.lehigh.edu>, "Deborah McGuinness" <dlm@ksl.stanford.edu>, <ned.smith@intel.com>, <jeremy_carroll@hp.com>, <connolly@w3.org>, <jos.deroo.jd@belgium.agfa.com>, <herman.ter.horst@philips.com>, <hendler@cs.umd.edu>, <www-archive@w3.org>
Jeremy: > >Justifications of Ns & ?s > >------------------------- > >N> data persistence > >N> security > >N> xml interfaces > >?> internationalization > > > >are part of the metadata layer of semantic web architecture > (RDF), not part > >of the ontology layer. Pat: > I fail to follow this distinction. (I know it is taken from the > 'layer cake' slide. I fail to follow it there as well.) RDF, for > example, is surely a (simple) ontology language. Do y'all have some > idea that metadata is somehow distinct from , and more primitive > than, ontology? (?? In what sense?) Pat: > Maybe xml and internationalization (maybe), but I don't see any way > to isolate security and data persistence issues from the ontology > 'layer', wherever that happens to be. > I see RDF as giving primitives for describing any data, and the ontology layer as describing constraints on how to use those primitives. These constraints are themselves data, described in RDF, and subject to a set of constraints! Not exactly layers; & I certainly don't see TBL's layered picture as Gospel Truth,; but it is a fair enough starting point. Aspects of security that appear in database schemata include access control. If we can include access control info in our ontology language that will be good, but it is not a MUST HAVE for me. It seems that in order to define such a thing we need to have a much better model of web security than we currently have, so I think it is more profitable to duck it for now. I understand persistence as about storing our data on a disk, or similar. This seems to be a general problem for any RDF data and not ontology specific. Unless we wish to annotate ontologies to say that this property should be stored persistently, and this property not. (Again not a feature I would die for). I am suggesting a meta criteria that we should address issues that do not appear for RDF; and not address issues for which an RDF solution would also be a WebOnt solution. But I am not dogmatic, and having expressed my opinion don't wish to get in the way of people coming up with good security or data persistence solutions! Jeremy
Received on Tuesday, 11 December 2001 05:51:15 UTC