Congratulations to Andy and his team on what sounds like an
impressive ICFP entry.
> * Speed
> - We ran the sample inputs in a few seconds per picture.
> For example, fib.gml ran in less than 7 seconds on my
> 650 Mz laptop.
The only other entry I'm really tracking is the OCaml group's,
which indicate a 0.58 second run time, although they do not
indicate the machine's specifications.
This (of course) leads me to wonder about the age-old question
of Haskell performance. Does the roughly one order of magnitude
difference in performance here indicate that lazy evaluation
is inherently less efficient than "early" binding? Or is this
just an artificat of the relative youth of the GHC compiler?
I'm not very familiar with the literature on this subject. Can
anyone provide the "Idiot's Guide" [tm] for me?
Thanks,
-Brent
