Re: Namespace: what's the correct usage?

[Martin Gudgin <marting@develop.com>]

----- Original Message -----
From: "David Carlisle" <davidc@nag.co.uk>
To: <marting@develop.com>
Cc: <xml-dev@lists.xml.org>
Sent: Saturday, May 19, 2001 7:30 PM
Subject: Re: Namespace: what's the correct usage?


>
> > I wouldn't 'complicate it with different namespaces for each element in
the
> > structure'. I would qualify the top-level element ( person ) and not
qualify
> > the children.
>
> but that means that the children and the top level element are
> potentially in different namespaces. One is in the namespace bound to
> the prefix, the children are in the default namespace.

When I say 'unqualified' I mean just that. i.e. In no namespace.

I never use default namespace decls. They're evil. To 'protect' myself I'd
put xmlns='' on the qualified element.

>
> It may be that you have xmlns="xxx" and xmlns:prefix="xxx" so these are
> the same namespace, but in that case it would seem clearer to either
> prefix all of them or none.

I have the top-level element qualified and children unqualified. Exactly
what the mapping is between namespace names and prefixes is a serialization
detail.

>
> You could add the neamespace declarations automatically in a DTD or
> schema but the namespace REC explicitly warns against doing that
> as it means that the namespace of the elements depends on the processor
> finding and reading the DTD.

I think it's safe to assume I'll get the namespace decls right.

Martin