Re: XSLT 1.1 comments

Michael Kay wrote:
> 
> > Can you be a bit more specific about which XSLT 2.0 style "type system
> > ramifications" a <saxon:function> feature would raise that aren't
> > already raised in <xsl:script>?
> 
> The aspiration is that XSLT 2.0 should be schema-aware in some sense, though
> I don't think anyone knows exactly what this means. But one possible
> implication is that functions should be in some sense polymorphic: there are
> already core functions whose behavior depends on the data types of the
> arguments, and it is desirable that user-written functions should also have
> this property. This isn't the case for the current <saxon:function>.
> 

It's a very interesting problem, and worthy of a well-thought-out
solution, but doesn't it apply equally to xsl:script and the bindings to
ecmascript and java?

While XML Schema has reached Candidate Release status, it is a complex
spec and is still evolving. Apparently - from current comments - this
will then be followed by a lengthy period of design ruminations about
how best to implement XML Schema type support in XSLT. I am really
concerned about the length of time during which platform-dependent
extensions will be more "standard" and - god help us - more "portable"
than extension functions in XSLT itself.

Can't the XSLT community be given *some* loophole, some "experimental"
or "optional" or "transitional" feature, for it to demonstrate how much
it does or doesn't care about having an XML transformation language that
is as portable as XML itself?

Francis.

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