RE: Information resources?

> | http://xmlns.com/wordnet/1.6/Hoary_Marmot is_a "web page"
> |
> | is wrong.  The URI identifies a resource; and the "web page" is a 
> | _representation_ of the resource.  It is not _the_ _resource_.
> 
> No, web pages are a class of resource and I can assert that 
> http://xmlns.com/wordnet/1.6/Hoary_Marmot is a member of that class.

Good point; there is nothing preventing someone from minting new URIs
that denote the actual stream of bits returned by an HTTP call.  

However, I think it's very important to be consistent -- so if people
use http: URIs to denote the "resource" rather than the
"representation", then they should always do so -- and pick a different
convention for denoting the representation.  

Larry Masinter once proposed the "tdb" scheme to do the converse --
tdb:http://xmlns.com/wordnet/1.6/Hoary_Marmot would denote the actual
resource (in this case, a member of class "word"), while
http://xmlns.com/wordnet/1.6/Hoary_Marmot would denote the "web page".

If I were suggesting conventions, I personally would always say that an
http: URI should denote a "resource representation dispenser".  IMO this
is the only consistent long-term view.

But overall, I just think it should be consistent.  We should not allow
people to use the same URI to alternately denote representation,
resource, and representation dispenser.  This would be a disaster.

Received on Thursday, 9 September 2004 21:21:07 UTC