Re: article on URIs, is this material that can be used by the SWEO IG?

Tim Berners-Lee writes:

> The TAG uses (I hope)  tag:representation only as a relationship 
> between a tag:InformationResource and a tag:Representation, the 
> latter being the class of (bits, metadata) pairs.  It is not 
> transitive.

Yes, indeed.  So, that's at least two of us :-).

> (I would say that its range and domain do not even overlap)

Um, need we have that discussion just now?  While I agree that in typical 
scenarios they don't, it's not clear to me that there aren't some edge 
case exceptions.  Consider, for example, a network debugging mechanism for 
the Web, and assume that the debugger provides a Web interface.  If you're 
debugging some particular GET request/response, then the representation in 
the response might itself be a useful resource, assigned a URI by the 
debugger.  So, in cases like this and some others, I could imagine 
identifying (with URI) and returning through HTTP a tag:representation of 
the tag:InformationResource that is itself a tag:representation.

Also, FWIW, I'd be a little happier calling the things we're discussing 
web:representation and web:InformationResource.  Surely 
web:representations existed prior to the formation of the TAG?

--------------------------------------
Noah Mendelsohn 
IBM Corporation
One Rogers Street
Cambridge, MA 02142
1-617-693-4036
--------------------------------------

Received on Monday, 11 June 2007 22:03:01 UTC