Re: [Haskell-cafe] forLoop + strict State monad is much faster than foldl'

Em 29-04-2014 13:35, Niklas Hambüchen escreveu:

(http://htmlpreview.github.io/?https://github.com/nh2/loop/blob/master/result...),

Is it just or is criterion broken in some way? It considers a 1140% effect on the mean "moderate", so I would guess that the number is being incorrectly printed. One of the tests even has a "severe" 7280% effect! Also, for "sum32/sumIntStrictState" is reports microseconds on the charts but seconds on the table. Cheers, =) -- Felipe.