Re: [xsl] What is exciting in XSLT 2.0 (Was: Re: [xsl] parameters in XSLT 2.0)

[James Fuller <jim.fuller@xxxxxxxxxxxxxx>]

> OK, but what kind of practical problems can it help solve for XSLT
> programmers?

most problems can be represented by a composition of functions....with
added benefit of being side effect free (or more dryly /referential
transparency)/

also using a functional approach means thinking in terms of data flow
instead of programmatic flow, which is of course familiar to XSLT.

a good real world example is FXSL, Dimitre has shown us that using
functional methods it is possible to create a  bunch of math functions,
perform numerical differentation and intergration using pure XSLT
1.0.....so clearly, yes you can use FP approach to *solve* math problems
if you have them.

an fp approach, set in a language (xslt) which is not a full blown
computer language means that we should be talking about processing of
text or xml....for example I have in my emacs (which has lisp) quite a
large set of functions I have built up over the years for doing things
like searching, deleting, formatting, finding and replacing (in some
unique weird manner)...all of which I could reimplement in XSLT 1.0
using FXSL. I am not saying that there are difficulties in porting LISP
or Haskell (which FXSL heavily borrows from). Whenever anyeone asks me
why use an fp approach, I point them to emacs and using built in lisp
processor...past that there are many resources online which show how fp
approach solves real world problems outside of math domain.

cheers, Jim Fuller