On Tue, 2005-07-05 at 12:39 +0100, Simon Marlow wrote:
I hope you're not going to conclude *anything* based on the performance of ackermann and tak! :-)
Well .. whatever a 'registered' build is it does this, the test is ackermann(3,n), the number in [] is the number of data points, and (min, average, max) is shown. xlator ghc-6.2.2 n 10 -[ 14]-> ( 0.68, 0.69, 0.72) n 11 -[ 18]-> ( 2.81, 2.90, 3.01) n 12 -[ 13]-> ( 11.72, 12.76, 16.95) xlator ghc-6.4.1 n 10 -[ 4]-> ( 0.20, 0.20, 0.20) n 11 -[ 6]-> ( 0.84, 0.84, 0.84) n 12 -[ 1]-> ( 3.53, 3.53, 3.53) Of course I won't conclude anything .. :)) -- John Skaller <skaller at users dot sourceforge dot net> Download Felix: http://felix.sf.net