Re: [Haskell-cafe] type inference futility