RE: RFC2518 (WebDAV) / RFC2396 (URI) inconsistency

> -----Original Message-----
> From: ext Dan Brickley [mailto:danbri@w3.org]
> Sent: 21 November, 2001 14:45
> To: Stickler Patrick (NRC/Tampere)
> Cc: julian.reschke@gmx.de; ejw@cse.ucsc.edu; w3c-dist-auth@w3.org;
> uri@w3.org
> Subject: RE: RFC2518 (WebDAV) / RFC2396 (URI) inconsistency
> 
> 
> 
> 
> On Wed, 21 Nov 2001 Patrick.Stickler@nokia.com wrote:
> 
> > > > As a result, I recommend that the XML namespace
> > > recommendation be modified
> > > > to allow the use of just the URI scheme name as a namespace
> > > identifier,
> >
> > The interpretation of a URI scheme prefix as a valid
> > URI also makes sense from the viewpoint of RDF (at least
> > to me ;-) in that otherwise, one has no way to make
> > statements about the URI scheme itself.
> 
> Not so. 

I stand (somewhat) corrected.

> It is possible to describe things without knowing 
> their URI. 

Just because something is possible, does not mean it is optimal.

> We
> do it all the time. Here, you can describe the scheme indirectly, eg:
> 
> <rdf:RDF blah blah...>
> <Scheme>
>  <prefix>dav</prefix>
>  <name>The DAV URI scheme</name>
>  <seeAlso rdf:resource="http://www.webdav.org/"/>
>  <blurb>A URI schema for ... </blurb>
> </Scheme>
> </rdf:RDF>

But this opens up the question of what the canonical identity
of the URI scheme is and how equivalence operations (graph
tidying, etc.) apply to such complex structures.  A single
URI provides a single, consistent point of reference,
and one that is not bound to RDF, but is acceptable to
any web application that is "URI aware".

> We can further tighten things by saying that the RDF property sketched
> here, 'prefix', is a UniqueProperty (using DAML+OIL or another Web
> ontology language). That tells us that an individual Scheme 
> is picked out
> by each value of 'prefix'.

But that's the whole point of a URI, no? Why introduce yet another
mechanism (and one that is less general) when a reliable and
fully unambiguous interpretation can be ascribed to URI prefixes
and URN Namespace prefixes? 

Surely the more general and economical solution would be better?

And BTW, most RDF parsers are already quite happy interpreting 
"http:" and similar URI prefixes as a URI ;-)

Patrick

--
               
Patrick Stickler              Phone: +358 50 483 9453
Senior Research Scientist     Fax:   +358 7180 35409
Nokia Research Center         Email: patrick.stickler@nokia.com

Received on Wednesday, 21 November 2001 07:56:47 UTC