Re: [Haskell-cafe] Type inference doesn't always calculate the most generic type?