On Fri, 28 Nov 2003, Stuart Brown wrote:

> > > element name in SGML?  I'm assuming the use of @xmlns on the TEI
> > > element is designed to allow this kind of versioning, so: <TEI
> > > xmlns="http://www.tei-c.org/P5/"> for P5 and heaven forfend if P6 ever
> > > comes along then <TEI  xmlns="http://www.tei-c.org/P6/"> etc.?
>
> > precisely, in my view
>
> Bad idea, in my view. I think namespaces should not change as part of a
> versioning system. Just because P5 revels in breaking old documents doesn't
> mean that all subsequent versions will. The namespace must be declared in
> the XML files themselves, so you are forcing into the XML itself a statement
> of version (no better, really, than the joyously removed <TEI.2>). If I want
> to (at some hypothetical point in the future) test files against P6 to see
> if they are valid or need tweaking, I need to go in and fix all the
> namespace declarations to point to /P6/. Of course, if I use
> xsi:schemaLocation I'll have to fix that anyway, but it is optional and
> applications have other ways of aligning schemas and documents which do not
> require an explicit declaration in the XML file (and this is a good thing,
> isn't it?); whereas the namespace declaration will be mandatory. Let's just
> have a namespace of http://www.tei-c.org/namespaces/TEI/ or similar.
>
> Follow the example of the chaps who invented the thing in the first place.
> XSLT 2 has the same namespace as XSLT; MathML 2 has the same namespace as
> MathML, etc., etc...

Ok, I can understand why that might be a problem in some ways.  I'm sure
I'm just being really dense here, but why not just <TEI version="5">
(in addition to any namespaces desired).  Can't this attribute be
constrained only to allow 1|2|3|4|5 as values?  This sounds much too
simple so I'm obviously missing something.  But XML and XSLT do this
don't they?  As Stuart points out XSLT2 has the same namespace as XSLT,
but it does have a different version number, and I believe the spec
requires this number.

I'm sure there are good reasons not to do this, I just don't know
what they are.

-James

---
Dr James Cummings, Oxford Text Archive, [log in to unmask]