RE: Making the namespace resource "real"

• From: Julian Reschke <reschke@medicaldataservice.de>
• Date: Thu, 8 Jun 2000 21:46:32 +0200
• To: <keshlam@us.ibm.com>, "Julian Reschke" <reschke@muenster.de>
• Cc: "Clark C. Evans" <cce@clarkevans.com>, <xml-uri@w3.org>
• Message-ID: <NCBBIPMOPKLLGKJPBINCCEPCEFAA.reschke@medicaldataservice.de>

Joe Kesselman wrote:
> >I think one issue with your proposal is that the namespace in general is
> an
> >infinite set of names.
>
> Sorry, but that is exactly what the namespace spec defines. A namespace

OK, I agree. I wasn't making my point very clear.

> defines a _space_ which names can be drawn from, not the content of that
> space, its meaning or anything else.  If you want a finite dictionary, or
> semantic bindings, or anything else constraining what the names
> are and how
> they're interpreted, you need to build some higher-level concept on top of
> Namespaces.
>
> Namespaces really are pretty close to being pure syntax. Adding semantics
> to them is someone else's problem, and is NOT part of the question
> currently on the table.

I agree again. However it seems that others disagree.

Received on Thursday, 8 June 2000 15:46:32 UTC