"Peter Rushforth" <peter.rushforth@xxxxxxxxx> writes:

> Are we likely to see improved performance of XSLT (2!)  in the (near) future,
> given the declarative nature of the language as a result of threaded
> execution of a single stylesheet?  Once a system has devoted the
> memory resources to parsing an xml document, it would seem a shame to
> limit processing of that static structure to one core if many could be
> used to speed the processing up.  Or does one have to have a language
> which syntactically supports multithreading?

As you say, xslt is a functional language and so these sorts of things
look very possible.

I am really interested in the possibilities of lazy evaluation. Lazy
evaluation means delaying the computation of things so that you might
not need to do it at all.



> Given the ability to model just about anything in
> XML, XSLT can be used to process just about any kind of data.  It is
> limited by memory and speed, from what I've seen.  Could the speed
> factor be addressed by threading ?

Yes.

-- 
Nic Ferrier
http://www.tapsellferrier.co.uk   for all your tapsell ferrier needs