[topicmapmail] Meaning of URIs - ongoing debate on new W3C forum

[Jan Algermissen]

Bernard Vatant wrote:

> Maybe some of TM gurus would like to jump in that debate to support that
> viewpoint. What is at stake is critical - making W3C specifications endorse
> TBL's theory of URIs "unique meaning" - which is obviously a dead end, or
> have them move to something making more sense, along the lines that TM have
> explored ...

Bernard--

I agree with your desire to 'fix addressing in the semantic web' and I also
agree with Murrays 'beating my head against the wall' feeling ;-)

The problem is that Web architecture in general is (among other things)
grounded in the idea of a 'single unique addressing context' (what you
quote as "unique meaning"). The useful consequence is that when I am
given a URI, I *allways* know what I can do with it (apply GET,POST,PUT
and DELETE methods to it). For a technology' to be "on the Web" it must
conform to this, uhm.., axiom. RDF essentially is intended to be "on
the Web" so there is not really much that can be done about it.

Of course this does not prevent technologies to make use of Web technology
while also using different addressing contexts. Topic Maps do exactly this
by using the 'subjectIndicators'- *and* 'subjectAddress'-addressing contexts.
The consequence is that Topic Maps are *not* "on the Web": you cannot use
a URI (found in a TM), enter it in your browser's location bar and be sure
that you get back a representation of what the URI identified. You also need
the addressing context in which the URI has been used.

This is not neccesserily a problem, it just creates significant difficulties in
bridging RDF and Topic Maps.

Some time ago I had the idea that every RDF namespace would *implcitly* carry
around the addressing context that would beneeded for a conversion to TMs (e.g.
that Dublin Core allways uses the rdf:about attribute in the subjectAddress
context) only to find out some weeks ago that this is wrong. (dc:description
is for example also used to provide a description about cars, dogs and the like).

Thoughts?


Jan 


> 
> Thanks for your attention
> 
> Bernard Vatant
> Senior Consultant
> Knowledge Engineering
> Mondeca - www.mondeca.com
> bernard.vatant@mondeca.com
> 
> _______________________________________________
> topicmapmail mailing list
> topicmapmail@infoloom.com
> http://www.infoloom.com/mailman/listinfo/topicmapmail

-- 
Jan Algermissen                           http://www.topicmapping.com
Consultant & Programmer	                  http://www.gooseworks.org