RE: rdfms-resource-semantics

: Lee:
:
: Of course, making it concise is only the first step.  Further changes 
: are necessary.  From your previous mail [1]:
: 
: [[
: "RDF expressions are used to describe resources as discussed in [RFC2396].
: In RDF, resources MUST be identified by a URI plus optional anchor ids (see
: [URI]). The current state of an RDF resource MAY be machine accessible to an
: RDF application."
: ]]
: 
: This still refers to resources "identified by URI plus optional 
: anchor id", which I would like to see reworded to remove the 
: ambiguity of whether an optional anchor id identifies a resource - 
: according to RFC2396 it doesn't.

That's a good catch. But rewriting that _would_ change the meaning, which 
I tried not to do. Or at least not that much.


: In addition, this para (and the original RDF M&S spec) still says 
: that a 'resource' (in the RDF sense) MUST be identified - this is 
: contradicted elsewhere in the spec where it discusses anonymous resources.

I shall look over this anon (sorry, had to :).


: In my previous mail [2] I mentioned that an unambigious term (e.g. 
: 'Reference') should be used instead of 'resource' (in the RDF sense), 
: and that it should be defined as the union of the disjoint sets 'URI 
: reference' and 'anonymous reference'.

The actual term used is only important if it's 'resource'. RDF describes via 
URIs (plus...) and many people believe that the use of URIs are fundamental to 
the SW effort. It makes sense to me to move RDF toward rfc2396 understanding 
of the term 'resource'. I've considered proposing 'entity', but synthesis is 
preferable unless someone can persuade me that the resources in question are 
fundamentally different. 
 
regards,
Bill

----
Bill de hOra : InterX : bdehora@interx.com