compilation of pattern-matching?

[Claus Reinke]

>When you tried switching Nil and Cons, did you try it on many examples?
>For a single example a 2-3% could be easily attributed to random
>effects like different instruction cache hit patterns.  If you get it
>consistently over several programs then it seems likely to mean
>something, but I'm not sure what.

Agreed. There is a fairly consistent pattern over experiment/size
variations (eg, different times for different orderings), but the time
differences themselves are too small for useful interpretation. I
would need a way to see what is really going on on the processor
(I'm sure there are such things - are there any free ones?) to make
sure there are no other effects interfering, or I might as well 
measure a random generator.

Claus