- From: Seth Russell <seth@robustai.net>
- Date: Mon, 4 Jun 2001 07:54:41 -0700
- To: "Brian McBride" <bwm@hplb.hpl.hp.com>, "Peter Crowther" <peter.crowther@networkinference.com>
- Cc: <www-rdf-logic@w3.org>
From: "Brian McBride" <bwm@hplb.hpl.hp.com> > Peter Crowther wrote: > > > > LL of course, could be encoded in RDF, but that is not the > > > same thing as it 'being' RDF. > > > > Brian, can I ask how you would encode such a language in RDF yet prevent it > > from being RDF --- or, at least, being mis-interpreted as RDF by a > > non-LL-aware agent? I guess I'm taking issue with the 'of course' in the > > above paragraph; I think that encoding process needs clarification. > > Now I wish I hadn't included that paragraph. I included it only to try > to head off any confusion between extension and encoding. > > I'd like to focus on my main question though. What are the relative > merits of 'extending' RDF v designing a new language for > expressing rules which operate on ground facts expressed in RDF. If you look at the axioms of how the internet works: { [1] Anybody can say anything about anything at any particular time => [2] Any agent can read and collect anybody's statements at any particular time } then I don't think that there would be any practical difference between RDF and LL. The statements that transform a graph that is supposidly RDF to one that is LL are the statements that define the arc labels. But any agent will encounter and collect such statements at no particular time and in no particular order and such a collection is in no way guaranteed to be complete. In fact were LL to be flexiable and evolvable, then the completeness of LL is undefined, and it's distinctness from RDF is completely fuzzy. Me thinks you attempt to draw a distinction that doesn't make any sense on the Internet. Attempting to make such a distinction is to attempt to climb a slippery slope. Why bother? Seth Russell
Received on Monday, 4 June 2001 11:01:38 UTC