Mark Carroll mark at
Thu Jan 1 01:47:39 EST 2004

On Wed, 31 Dec 2003, Ken Shan wrote:

> Don't you need a (s2 -> s1) function as well, to translate the final
> state back into StateT s1?

Yes, you're right: the thing actually running the stateful computation
presumably expects to start it with a state of type s1 and to be able to
extract from it a state of type s1 when it's done.

-- Mark

More information about the Haskell-Cafe mailing list