Re: Exposing newtype coercions to Haskell