On Wed, May 22, 2013 at 6:22 PM, Ross Paterson
Wasn't there also mention of making Applicative a superclass of Monad?
That might actually have less impact on beginners than generalizing the types of foldr and sum, and doing it first might allow us to rationalize the situation with some of the monad combinators that are clashing.
That would create rather more significant breakage though, would it not? The change discussed here should be relatively painless in comparison. Even if it's a good idea, which it is, and even if we want to do it, which we should, it warrants a separate proposal. As demonstrated here, the community is overwhelmingly in favor of generalizing mapM &c. anyway, so there's no need to make it dependent on another improvement. - C.