Re: typo/bug in the namespace spec? [was: Mechanism, not policy...]

> Argument by assertion is no fun... Here's the
> argument by which I conclude that it's a typo:

there does seem to be some consensus that relative URIs got let in by
mistake when fragment ids were allowed (which is due no doubt to the
appalling terminology in the RFC which doesn't give a name to the most
specific identifier (absolute URI + fragment id) and calls something
"URI reference" when it isn't a reference to a URI, it's not
surprising these things happen.

However it isn't a typo that the resource identified by the namespace
name as a URI reference is not the namespace.

> I disagree. If you have any justification for this claim,
> I'm interested to see it.

Having posted I don't know how many messages to this list giving
such justification I don't know what else I can say.


> That's what "identifier" and "name" mean, no?
no

You use the URI of a resource to name your namespace so almost always
the resource identified by the namespace name is not the namespace.

> Hang on... do you mean URI reference,
I mean URI reference. That is what the REC says. Even if you argue
that it wasn't supposed to say that even you can't argue that the REC
currently says that a namespace name is a URI reference. It does not
say the namespace name is a URI obtained by combining a URI reference
with a base.

> URI references denote URIs

They are also used as namespace names.

> I agree that namespace processing per se is done just with namespace
> names, and that the namespaces themselves are formally
> irrelevant at that level

It is that level which is the subject of this list. Schema validation
while a good thing is a topic for another day.

David

Received on Thursday, 8 June 2000 12:44:28 UTC