Re: [Haskell-cafe] Unique functor instance

Are identity and composition sufficient to guarantee that the mapped function is actually applied?

eek - f :: a -> b, not f :: a -> a, so that example doesn't work !! Sorry for the noise, Claus