Re: URIs quack like a duck

Paul W. Abrahams scripsit:

> Sorry, I wasn't clear.  I was talking about comparing URI1 with URI2.
> Does RDF ever do that?

Well, indirectly.  It wouldn't make much sense to have RDF if you could
not determine whether two statements were about the same resource or
not.

> When you refer to RFC2396 comparison, are you talking
> about comparing one URI with another or about something else?
> I don't see anything in RFC2396 that talks about comparing one
> URI with another, but perhaps I just missed it.

My error.  I should have spoken of "comparison after RFC 2396 resolution"
(aka "absolutization").

> Let's see if I understand it now.  A namespace is a resource, so we
> want to have some way of using a URI to refer to it.  The "data:,"
> provides such a way.    So if we have 
>      ... xmlns:foo="http://www.sushi.edu/octopi" ...
> 
> then "data:,http://www.sushi.edu/octopi" provides a way to refer to
> that namespace in other contexts.  Right?

Just so.

I agree with TimBL that this is a subtle, even hair-splitting, approach,
but I see nothing else likely to get us away from the current standoff.  It's
a trick application of URIs to reach a compromise position.

It preserves the Namespace Rec principle that namespace names are
to be compared literally, but it also preserves the URI principle
that namespaces (like everything else) are resources and that resources
need a URI to identify them.

-- 
John Cowan                                   cowan@ccil.org
	Yes, I know the message date is bogus.  I can't help it.
		--me, on far too many occasions

Received on Saturday, 3 June 2000 00:03:38 UTC