Re: [Haskell-cafe] 2-level type analysis of introducing naming into data types