Re: [Haskell-cafe] restricted existential datatypes