Re: Language = Namespace. was: How namespace names might be used

Julian Reschke wrote:
> 
> Dan Conolly wrote:
> 
> > > 3) Given an RDF document referring to that URI, does it make statements
> > > about a namespace, a language or an HTML document???
> >
> > about a namespace/language.
> 
> OK.
> 
> Let's take the first RDF example I can find right now
> (http://www.w3.org/TR/REC-rdf-syntax/, section 2.1.1 Examples):
> 
> <quote>
> Consider as a simple example the sentence:
> 
> Ora Lassila is the creator of the resource http://www.w3.org/Home/Lassila.
> This sentence has the following parts:
> 
>  Subject (Resource)   http://www.w3.org/Home/Lassila
>  Predicate (Property)   Creator
>  Object (literal)   "Ora Lassila"
> </quote>
> 
> I now create a namespace name which happens to be
> "http://www.w3.org/Home/Lassila". Which means that I have re-used an
> existing URI as namespace name, which is perfectly legal according to the
> namespace recommendation. Does the fact that the same URI is now in use as
> namespace name as well affect any of the RDF properties of this URI (for
> instance the "Creator" property)?

No.

It does allow us to conclude that the Creator of the namespace
is (called) "Ora Lassila", though.

> > > 4) If all of them are *the same thing*, what happens if the
> > HTML document is
> > > replaced by an XML schema document? (I think this happened
> > recently @ the
> > > namespace RECs own URI).
> >
> > What happens is that clients get a different representation of the
> > resource (i.e. the namespace/language) from what they previously got.
> > Just like when a GIF image is replace by a PNG.
> 
> To make this consistent, the namespace REC at least would need to require to
> use an URI ref that actually does more than "naming" the namespace.

Yes and no... yes, in that when we choose URIs to name things, we
import the fact that they name resources in the context
of the Web. But no, I don't see why we need to say anything
more than that.

> Right
> now it only requires a URI ref, but doesn't preclude the namespace creator
> from just re-using something that is already there, but happens not to be
> used as a namespace name yet.

Where is the inconsistency? All of the specs that use URIs (and
URI references) have this property: they allow you to say
nonsensical things about URIs, such as that RFC822 is an image:

	<img src="http://www.ietf.org/rfc/rfc0822.txt"/>

Such a claim might turn out to conflict with other evidence
(such as what you get back when you access that resource,
or such as what other documents at the IETF web site
say that resource is),
but that doesn't make the document illegal. The mechanisms
allow anyone to say anything about anything. It's bad
policy to lie and to say nonsensical things, but it's
not forbidden by the specs.


-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Tuesday, 20 June 2000 09:52:09 UTC