RE: optimization for very large, flat documents

> With the number of times this issue comes up Mike,
> I'm surprised the WG haven't considered it, or some variant 
> such as the STX approach?
> I.e. some means of XSLT-like processing for large document,
> even given a strict list of restrictions.
> 
> It would appear that (most of) the constraints are known;
> The use case should (or could be) readily determined;
> All that's missing is the impetus to move on this.
> 

I think it would be wrong for W3C to attempt to define a language suitable
for serial processing without more implementation experience.

There's been a lot of talk about this, there are a lot of research papers on
the subject, but I don't think there's a solid consensus on the state of the
art that makes it ripe for standardisation.

I would actually expect to see more progress made in this area on the XQuery
front - XQuery, being a much smaller and more rigid language, is rather more
amenable to the kind of static analysis required, and (perhaps for the same
reason) attracts more academic interest.

Michael Kay
http://www.saxonica.com/