RE: rdfms-resource-semantics

On Wed, 6 Jun 2001, Lee Jonas wrote:

>I didn't fully understand what you meant by this para.  My point is: the
>RDF spec should say RDF describes things via URI references as outlined in
>RFC2396, full stop.  I.e. stop trying to define a special notion of
>'resource'/'thing'/'entity' in the RDF spec - it *is* whatever can be
>referred to by a URI reference.  The RFC2396 definitions of Resource,
>Entity, URI, URI reference, etc are all normative and sufficient.

It was already pointed out that the fragment identifier is not part of the
URI, i.e. the identity of the *resource* pointed to by a URI reference.
RFC2396 uses the work 'resource' to mean whatever is being pointed to by a
URI, not a URI *reference*. The thing pointed to by a URI reference isn't
given a name, and people too often use the word 'resource' to refer to it,
even when this isn't the case. In this sense, the resource in the M&S sense
is very different from the RFC2396 one, and the distinction should be
emphasized.

Sampo Syreeni, aka decoy, mailto:decoy@iki.fi, gsm: +358-50-5756111
student/math+cs/helsinki university, http://www.iki.fi/~decoy/front

Received on Thursday, 7 June 2001 05:25:24 UTC