I wrote: >> A more basic issue is that fn is in the IO monad, >> but its use inside the mapM will need it to be in the ST >> monad.
Daniel Fischer wrote: > No, > return (fn particle) :: ST s (IO ()) > , so that's fine. Ah, true. But I doubt that Andrew really meant to do the calculation in ST s (IO ()). > Indeed, filling in a few dummies, the code compiles > with the 6.8 branch. So then the only way to know why it didn't compile for Andrew is if we see more of the code. But that's only a technical issue. I hope that now Andrew has a better idea about what's wrong. If not - please let us know. Regards, Yitz _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe