RE: What is at the end of the namespace?

> -----Original Message-----
> From: ext Mark Baker [mailto:distobj@acm.org]
> Sent: 19 November, 2001 22:23
> To: Stickler Patrick (NRC/Tampere)
> Cc: www-talk@w3.org; uri@w3.org
> Subject: Re: What is at the end of the namespace?
> 
> 
> > > Earlier you suggested that "brilliance" was abstract, yet I happen
> > > to have a URI for it here;
> > > 
> > > http://www.markbaker.ca/2001/11/Brilliance/
> > 
> > No, that's a URL to retrieve a definition of the abstract
> > concept 'BRILLIANCE' from a particular web location.
> 
> No, it's what I say it is.  How can you, or anybody else, tell me what
> I'm naming?

Try expressing two RDF statements using the above URI: one that 
describes the abstract concept "BRILLIANCE" and one that describes
the definition obtainable by a HTTP GET request.

You can't.

> > The
> > fact that you have a clever redirect does not make the URL
> > any more suitable a representation of the abstract concept
> > than http://www.dictionary.com/cgi-bin/dict.pl?term=brilliance.
> 
> The redirect communicates that I'm currently (it's a temporary
> redirect) delegating the definition of my concept of brilliance to
> dictionary.com.

Right. The *definition* of your concept of brilliance. Not the
concept of brilliance itself. What if you wish to provide three
definitions and five examples of brilliance. Do all of those 
eight URIs all denote the abstract concept "BRILLIANCE"? They
may be expressions of it. They may be related to it. But they
do not denote *it*.

Sorry.

> > I.e.
> > 
> >    <rdf:Description
> > rdf:about="voc://patrick.stickler@nokia.com/concepts/brilliance">
> >       <foo:definitionOf
> > 
> rdf:resource="http://www.dictionary.com/cgi-bin/dict.pl?term=b
> rilliance"/>
> >    </rdf:Description>
> > 
> > could be one way I could point to a definition of
> > "BRILLIANCE" as *I* conceive it, though I may choose some
> > other definition, or multiple definitions, as I see fit.
> 
> Using an HTTP temporary redirect on a GET invocation means 
> more or less
> the same thing as what you have written there.  

Hardly.

> It's "just" another
> syntax that also happens to be a protocol, that allows you to automate
> the dissemination and manipulation of that information.
> 
> Without a protocol that does that, any other URI scheme is 
> automatically
> playing with a handicap.

On the contrary, I consider a URI scheme that has its resolution
agency named explicitly in its structure to be *FAR* more handicapped
as a persistent, context independent identifier of a resource!

The reason why URNs are needed is because we need identifiers that
can be resolved in a context free manner. An HTTP URI is bound to
a specific *host* for resolution, and just because you might be
able to either (a) redirect to some other host or (b) resolve in
some other manner than HTTP does not lessen that "handicap".

Patrick

Received on Wednesday, 21 November 2001 03:07:38 UTC