RE: URNs for 'naming authority assignment', not 'permanent'

> -----Original Message-----
> From: ext uri-request@w3.org [mailto:uri-request@w3.org]
> Sent: 17 September, 2003 18:13
> To: uri@w3.org
> Subject: RE: URNs for 'naming authority assignment', not 'permanent'
> 
> 
> 
> 
> 
> 
> On 17 Sep 2003 at 12:30, Patrick.Stickler@nokia.com wrote:
> 
>  > The HTTP protocol seems only to be telling you how to obtain
>  > representations of whatever is denoted by the URI, not how to
>  > obtain the actual semantics/denotation of the URI.
> 
> ...
> I'd say that, by definition, the URI http://foo.example/some/junk/
> denotes whatever can be retrieved by the HTTP protocol using this
> host and path.  

I've understood this as a common misconception. One cannot reliably determine
what a URI denotes based on what a server returns in response to
a GET request.

One can guess. But such guesses are not ratified in any way by the
formally defined web architecture. Thus, such a view is certainly not
"by definition" (at least not anywhere I can find in any standards
document).

Representations accessible via a given URI might change over time,
but its denotation should remain constant.

Thus, a URI denotes some resource. The HTTP protocol allows one to 
obtain representations of that resource, if the URI is of a scheme
that is meaningful to the HTTP protocol. Other protocols, such as
DDDS offer similar functionality. But at present, there is no 
standardized means of obtaining an authoritative definition of the
semantics of a URI.

That has been the primary motivation for developing URIQA, as a
standardized extension to HTTP allowing one to obtain an authoritative
and formally defined (RDF) description of the resource denoted by
a given URI.

Patrick

Received on Thursday, 18 September 2003 02:56:21 UTC