[uf-discuss] URIs vs URLs (reference)

[Danny Ayers]

I know this isn't really the right forum (what is? or even WHAT is?) ,
but I hope I may be forgiven a couple of direct questions here -

On 10/25/05, Dr. Ernie Prabhakar <drernie at opendarwin.org> wrote:

> Hmm, I read that as agreeing with my understanding, which matches the
> "contemporary" view:

> > The W3C and IETF should jointly develop and endorse a model for
> > URIs, URLs and URNs consistent with the '"Contemporary View"

Sorry, I don't actually know what you mean by the contemporary view.
As a spec-tart I'd go with what the RFCs say. What do you mean?
(Direct question #1)

> Thus, my earlier statement -- I prefer URL over (non-URL) URIs -- is
> at least meaningful, whether or not it is a good idea. :-)

The spec definition works for me. There are identifiers (names,
constrained-syntax strings with mini-semantics,  URIs, IRIs) of
conceptual resources, which can be used with the HTTP protocol to
access concrete representations of those resources. It's a bit
convoluted, but does a very important (IMHO) separation of concerns,
the identification and the access (location). Which means that URIs
can be freely and meaningfully used outside of the the HTTP context
(e.g.  XML namespaces, WS-*,  XMPP messaging, RDF). In principle I
guess it also means that HTTP could be used with other forms of GUID
as Ian referred to earlier.

So, direct question #2: how do you see URL being advantageous?

Cheers,
Danny.

--

http://dannyayers.com