Re: [Haskell-cafe] Could FDs help usurp an ATs syntactic restriction?

2008-12-07 Thread Manuel M T Chakravarty
Nicolas Frisby: From the error below, I'm inferring that the RHS of the associated type definition can only contain type variables from the instance head, not the instance context. I didn't explicitly see this restriction when reading the GHC/Type_families entry. Could perhaps the "a b -> bn" f

[Haskell-cafe] Could FDs help usurp an ATs syntactic restriction?

2008-12-04 Thread Nicolas Frisby
>From the error below, I'm inferring that the RHS of the associated type definition can only contain type variables from the instance head, not the instance context. I didn't explicitly see this restriction when reading the GHC/Type_families entry. Could perhaps the "a b -> bn" functional dependen