Re: [Haskell-cafe] Deducing a type signature