> But last time I read the data model closely, it said that it was not
> necessary to implement namespace nodes, if you didn't support the namespace
> axis.

I think that the situation is that you can implement a conformant
implementation of XDM that does not have namespace  nodes, but you could
only it use to support XQuery and (restricted) XPath. In order to
support XSLT you need an XDM implemention that does have namespace
nodes.

> my suggestion that the wording might be better as saying xsl:namespace
> produces a namespace binding in the tree), then only two items would be
> present in the sequence.

But that's less useful and is too big a change to suggest at this stage
of the process, I think. I think it would have to push XSLT back to last
call.

David

In fact, my failed attempt to reply to Florent's post last night
(Florent, please, do not use special French apostrophes (or whatever
they are called) in your signature, because any reply to such a
message is encoded by GMail in bin64 and is rejected by this list's
software) was this:

It doesn't seem to me that whatever the answer to your question may be
('yes' or ''no'), this will affect the way one writes XSLT code or any
already existing XSLT application.

--
Cheers,
Dimitre Novatchev
---------------------------------------
Truly great madness cannot be achieved without significant intelligence.
---------------------------------------
To invent, you need a good imagination and a pile of junk
-------------------------------------
You've achieved success in your field when you don't know whether what
you're doing is work or play