RE: vendor neutral XSL extension namespace ?

> Would it be possible for XSL implementors to agree on a 
> common extension namespace for some of these (that is explicitly an
_experimental_
> extension and does not imply later acceptance by W3C)

Excellent idea. James Clark and I have been unashamedly copying each other's
good ideas for a while now, which I hope is to the general benefit of the
community. I think that once XSLT 1.0 freezes there's a good opportunity for
a third party (yourself?) to define a library of these "standard extensions"
that implementors can incorporate into products.

There is a problem in making such standard extensions totally portable,
which is that products (certainly both xt and SAXON) are currently using a
namespace to identify both the interface to extension functions and the
implementation. If we have multiple implementations of extension functions
with the same interface we need the namespace to identify the interface and
some separate mechanism to identify the implementation, in such a way that
you can say for a single stylesheet "The XT implementation of
http://xslt.standard-extension.org/node-set is at com.jclark.xt.extensions
and the SAXON implementation is at com.icl.saxon.extensions"

Mike Kay


 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list