Re: [Haskell-cafe] An issue with EDSLs in the ``finally tagless'' tradition