Re: RFC 3986 and RDF URI reference

Daniele Alessandrelli wrote:
> 
> Hello all,
> 
> I am a bit puzzled about the concept of URI in RDF models given the
> difference between the RFC 2396 , which is mentioned directly in the RDF
> Concepts document [1], and the defintion in RFC 3986.
> 
> given the new RFC is it correct to use the term URI in place of what was
> defined as URIREF in RDF Concepts?
> 
> Thanks
> Dany
> 
> [1] http://www.w3.org/TR/rdf-concepts/#section-Graph-URIref

RDF-concepts was drafted while RFC2386 was the current URI specification.
Specifically, the form of identifier used by RDF allows a fragment identifier,
which was not part of what RFC2396 allows in a "URI".  Since then, RFC3986 was
introduced, reflecting consensus that the term URI should include any fragment
identifier (but not relative URIs).  I guess that if we were drafting
RDF-concepts today, we would just say "URI" (or, as Damian suggests, IRI).

So, yes, I think it would be OK to say URI (or IRI) now, but that change can't
appear in the RDF spec until a new revision is published.  (As a practical
matter, when referring to RDF-concepts, it might be helpful to acknowledge the
other term.)

#g

-- 
Graham Klyne
For email:
http://www.ninebynine.org/#Contact

Received on Friday, 7 April 2006 16:38:58 UTC