- From: Graham Klyne <GK@ninebynine.org>
- Date: Wed, 29 Sep 2004 09:29:17 +0100
- To: "Hammond, Tony" <T.Hammond@nature.com>, www-rdf-interest@w3.org
At 11:11 28/09/04 +0100, Hammond, Tony wrote: >Yes, I may have shot the original mail off too quickly as I see Patrick's >MT qualifier now. However, I am still at a loss to know as to where in MT >there is any mention of URI opacity (or not) - I just checked. (BTW, why are >we talking about MT? Surely this doc has been superseded by the new crop of >recs earlier this year?) Tony, I assume by MT you mean the Model Theory document, aka "RDF Semantics": http://www.w3.org/TR/rdf-mt/ which *is* one of the "new crop of recs". Anyway, URI opacity is not mentioned in so many words, but it's taken for granted by the nature of the maths used. If I write a mathematical formula, say (a x^2+b x+c = 0), one implicitly treats all the variable names (a, b, c, x) as being opaque. There's no basis for doing anything else. >My point anyway is not that a generic RDF processor should be able to >interpret URI structure (apart from validating that it is a URI, of course) >but rather to say that certain URIs besides exposing the generic URI syntax >can also expose public data if there is a standard public specification for >the URI syntax. Applications - RDF or otherwise - are free to interpret this >data as they so choose. Well, you can of course do whatever you like with a URI to learn more about what it references, but the RDF specifications don't license such activity and any other RDF user/applications would be justified in disputing any conclusions obtained by such activity. #g ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact
Received on Wednesday, 29 September 2004 09:15:17 UTC