Re: [Haskell-cafe] re-definition of '.'