> >
> 
> [http://www.google.com/search?hl=en&lr=&q=%22Haskell+programmi
> ng%22&btnG=Search]
> 
> gives us 36,100 hits
> 
> [http://www.google.com/search?hl=en&lr=&q=%22XSLT+programming%
> 22&btnG=Search]
> 
> gives us 17,800.
> 
> i.e. 1/2 that of Haskell 

You put the search terms in quotes, e.g. "XSLT programming". 

If "XSLT" scores 3*"Haskell", while "XSLT programming" scores 0.5*"Haskell
programming", that is indeed interesting: it shows that most people talking
about XSLT don't think of it as programming. Which means that the site's
assumptions are basically flawed.

After all, the words "program" and "programming" don't appear anywhere in
the XSLT 1.0 specification, whereas they appear more than 20 times in the
two-page introduction to the Haskell spec...

I guess you can use statistics to prove anything you want to prove, and
since you seem to have a very negative attitude to most XML technologies,
it's not surprising that you should choose data that's biased against them.

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