W3C home > Mailing lists > Public > xml-dist-app@w3.org > March 2001

RE: Finalised Glossary Definitions

From: Andrew Layman <andrewl@microsoft.com>
Date: Wed, 21 Mar 2001 15:13:42 -0800
Message-ID: <C3729BBB6099B344834634EC67DE4AE13490A4@red-msg-01.redmond.corp.microsoft.com>
To: <xml-dist-app@w3.org>
I have not followed this thread, so I may be missing some context, but
the suggestion to use namespaces to target anything appears to go
against the intention of the Namespaces specification.  A namespace is
simply a means to disambiguate names.  One could, for some namespaces,
infer some semantics associated with the namespace as a whole, but that
would certainly not be true of all namespaces (and would also, in my
opinion, be bad design).

One might very reasonably say that an element name or type causes
association with a particular handler.  It is also true that element
names and types are often associated with namespaces.  But, the
"targetting" is better a function of the element name or type than the
namespace lending that name or type universal identification.

-----Original Message-----
From: Mark Nottingham [mailto:mnot@akamai.com]
Sent: Wednesday, March 21, 2001 12:24 PM
To: Jean-Jacques Moreau
Cc: Mark Jones; skw@hplb.hpl.hp.com; Frystyk; xml-dist-app@w3.org
Subject: Re: Finalised Glossary Definitions

On Wed, Mar 21, 2001 at 03:44:54PM +0100, Jean-Jacques Moreau wrote:
> Ah, yes! but this is precisely why, in my opinion, we should be
> targetting handlers, not processors or nodes. The actor attribute
> (or the <namespace:tag>, if we follow that route), would be used
> for that purpose.

I think that the idea is to use the namespace to target handlers,
actors to target processors/nodes.


Mark Nottingham, Research Scientist
Akamai Technologies (San Mateo, CA USA)
Received on Wednesday, 21 March 2001 18:15:14 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 22:01:12 UTC