Re: Namespaces, schemas, Simon's filters.

[Evan Lenz <elenz@xyzfind.com>]

Matthew Fuchs wrote:
> > >#2 was never true for me in the first place. In fact, I
> > would consider the
> > >use of namespaces to be orthogonal to the use of local element types.
> >
> > Yes!  Anyone disagree?
> >
>
> I'm not Evan means what you think he means, as he later agrees that local
> names should not be in a namespace, which you seem to vehemently disagree
> with.

What I said was: "I would consider the use of namespaces to be orthogonal to
the use of local element types." I don't think I claimed that the use of
local element types should be in a namespace or not (hence "orthogonal").
I'm not sure what you're referring to...?

> In any case, namespaces are just a mechanism which can be used in a
> variety of ways.  I believe the use of namespaces is related to what
you're
> using them for.  In particular, I believe they were intended to help
> disambiguate things, to uniquely identify things.  If you are working with
> Schemas, rather than in the well-formed world, then namespaces should be
> used to disambiguate the kinds of things you can define with a Schema.

I agree with all of this. In fact, my understanding of namespaces has been
heavily influenced by XSLT's use of namespaces, which further tears down the
notion that a name necessarily signifies a content model. QNames are used
for variable names, template names, mode names, etc. (Let's consider the
prefix-namespace-binding mechanism separately; mentioning this does not
necessitate a debate about the relative merit of using QNames in attribute
values.) Content models (or ty*es) are irrelevant. This reinforced for me
the idea that namespaces are just, like you said, "a mechanism which can be
used in a variety of ways".

Evan Lenz
XYZFind Corp.