Re: [Haskell-cafe] How to bring existentially quantified type variables into scope