Re: The range of the HTTP dereference function

On Sunday 31 March 2002 12:50 am, Larry Masinter wrote:
> # You could argue that the implied contract here is outside the
> scope of # the architecture, but at some point there is a *need* to
> represent a # given resource with a particular set of bits, if only
> for reification # so you can discuss the properties of that
> representation.
>
> HTTP uses ETag for this, specifically so that caching can be
> managed.

Right... the point is that there's no way, given *just* a URI, to 
guarantee that a resource has but a single given representation. There 
are cases where this is obviously less than perfect. Ideally, there'd 
be something akin to Etag standardized for use in URI's. 

> URIs weren't designed to be the basis for knowledge representation,
> so the attempts to use them for such a purpose aren't very
> successful.

That's an orthoganal issue to the issue of identity raised above.