Re: [Haskell-cafe] Existential quantification of config data types