RE: Documents, Cars, Hills, and Valleys

Sean B. Palmer wrote,
> If, however, the range of HTTP can be anything, then we're simply 
> not sure. In that case, we have to go back to one of:-
>
> * Using the indirection predicates
> * Using the test case domain idea
> * Possibly looking at the headers for Resource-Type headers:-

I think I'm largely in agreement (and I particularly like EARL as an
example of the phenomenon I've been blathering on about). My only
real quibbles are that I don't see your list of mechanisms as being
exhaustive (what about Accept:? or mechanism-less implicit context
and convention?), and I'm not convinced that we either need or could
constuct an exhaustive list. Piecemeal, ad hoc, case by case solutions
might be the best we can hope for and all we need ... after all, 
that's what we have now, and they more or less work most of the time.

This is, I guess, hard-lines for RDF. Or rather, I guess it's tough
for a view of the semantic web which understands semantics in a formal
sense ... if it's primary designators are infected with the same
kinds of ambiguity and vagueness that affect natural language then
formal methods are liable to lead to disaster. But I think there
room for an alternate view which understands semantics in a much
looser and informal sense, and which might be able to accommodate
the messiness of the actually existing web.

That'd be tougher to reason about mechanically, but then I suppose 
this is just the age old dilemma of choosing between saying very 
precise things about nothing very much, or saying imprecise things
about quite a lot.

Cheers,


Miles

Received on Thursday, 11 April 2002 08:00:08 UTC