W3C home > Mailing lists > Public > www-tag@w3.org > September 2002

Re: My action item on Moby Dec, issue 14, etc

From: Michael Mealling <michael@neonym.net>
Date: Fri, 20 Sep 2002 11:38:00 -0400
To: Bill de hÓra <dehora@eircom.net>
Cc: "'Michael Mealling'" <michael@neonym.net>, "'Jonathan Borden'" <jonathan@openhealth.org>, "'David Orchard'" <dorchard@bea.com>, "'Tim Bray'" <tbray@textuality.com>, "'Norman Walsh'" <Norman.Walsh@Sun.COM>, www-tag@w3.org
Message-ID: <20020920113800.B640@bailey.dscga.com>

On Fri, Sep 20, 2002 at 04:21:46PM +0100, Bill de hÓra wrote:
> > From: Michael Mealling [mailto:michael@neonym.net] 
> > 
> > And, depending on what RDF expects the definition of 
> > 'resource' to be, using another process to do that work may 
> > have been a mistake...
> 
> Interpretations are standard stuff in a model theory. Perhaps 'process'
> was a poor choice of word. RDF uses 'mapping'; the one that maps URIs to
> resources (or labels to the world) is called 'IS'.

Makes sense, but only if RDF is using the same definition for 'resource'
that URIs use. And I'm not sure it is....

> > What is RDFs definition of a 'resource'?
> 
> Not usefully different to HTTP's afaict; a resource is whatever a URI
> names. 

Correct. But is RDF expecting more out of that definition than it actually
states. Since a URI is the name of a resource and a resource is that thing
that is named by a URI, those are the only two things you can know at that
level. Semantics of resources don't exist. Is RDF assuming that other
statements about resources can be made other than their name?

> Nonetheless if RDF held an 1-1 mappings as an axiomatic, the MT
> wouldn't require the IS mapping. 

Ok, here's the issue: how can you not deductively conclude that mappings
are 1:1 exclusive given the definition of a resource as that thing that
is named by a URI?

> I suppose one could argue that IS is a
> redundant artefact from model theoretic semantics in the large, or the
> Web architecture provides IS 'for free' due to authoritative naming;
> neither seems an entirely satisfactory way to show the seamntic and the
> actual Web tee up axiomatically.

And that may mean you need some layer in between RDF and URIs that makes
it satisfactory. URIs only give you a unique string denoting one and only
one Resource. They don't give you anything else. Not equivalence, meaning,
resolvability, persistence, useability, available representations, etc...
Nothing....

-MM

-- 
--------------------------------------------------------------------------------
Michael Mealling	|      Vote Libertarian!       | urn:pin:1
michael@neonym.net      |                              | http://www.neonym.net
--------------------------------------------------------------------------------
!! The Trailblazer spacecraft is going to the Moon! And for just $2500 a gram !!
!! you can send something along with it! Business cards, momentos, cremains,  !!|| anything! See http://www.transorbital.net for details!                     !!
Received on Friday, 20 September 2002 11:43:25 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:11 GMT