RE: [xsl] 2.1 Must Allow Matching on Default Namespace Without Explicit Prefix

Subject: RE: [xsl] 2.1 Must Allow Matching on Default Namespace Without Explicit Prefix
From: "Evan Lenz" <elenz@xxxxxxxxxxx>
Date: Wed, 21 Feb 2001 12:35:43 -0800
Steve Muench wrote:
> To someone like yourself who is an XSL veteran, it may
> never be a mistake that you make, or the workaround is
> so burnt into your brain that you don't think twice about it,
> but the WG perceived this issue as a major "learnability" hurdle.

Interesting... Now, it will be a learnability hurdle for those of use who
have already learned. But I certainly hope we'll be in the minority :)

Of course, I think they might be equally frustrated by the fact that, in
order to generate a result that's *not* in a namespace, they will have to
insert xmlns="" on their literal result elements. It's all a matter of
perspective, I suppose.

It might be worth noting that such confusion probably only arises when the
source document uses a default namespace, as in your example, which
technically is orthogonal to whether the stylesheet uses a default namespace
or not.

The root of this confusion is what many have called a design flaw in XSLT,
namely the mapping of namespace prefix mappings to the in-scope namespace
declarations. I believe that DOM implementations are allowed to arbitrarily
change prefixes, because that supposedly would not change the meaning of the
document for applications.....except for XSLT stylesheets. I can't come up
with a better solution, and that's probably not going to change. I'm not
sure that I'd want it to anyway.

But you still didn't answer my question. Will the default namespace be used
to evaluate XPath expressions? I know they're different specs, but I think
inconsistency there would cause even more confusion.

I like the rule in XPath that says anything in a namespace has to have a
prefix. Yeah it's different than the case with XML Namespaces, but it's a
handy thing to be able to immediately recognize whether a name test is in a
namespace or not. I could cope with this change, but it seems like it would
cause a fairly significant backward compatibility headache.

Evan Lenz
XYZFind Corp.

 XSL-List info and archive:

Current Thread