Simon Marlow wrote:
Neil Mitchell wrote:
Would it not be possible to add a GHC rule like the following:
forall a b . sequence a >> b = sequence_ a >> b
I'm not sure if thats correct, a valid rule definition, or semantics preserving, but if it was it would be nice :)
Now there's a good idea!
Well this would work in the case of sequence vs sequence_ but it seems to me that what's really needed is a compiler that can do whole program optimization so that the list building inside sequence would be eliminated in this case (because the list is never used) without needing an ad-hoc re-write rule. Otherwise, a programmer must know about all the re-write rules as well as all the functions, and be able to apply them in their head to calculate what the space/time complexity of their program actually is, which imho is much more difficult than just writing sequence_ in the first place. Has anyone done work on an equivalent of MLton for Haskell? Thanks, Brian. -- Logic empowers us and Love gives us purpose. Yet still phantoms restless for eras long past, congealed in the present in unthought forms, strive mightily unseen to destroy us. http://www.metamilk.com