[Haskell-cafe] type families and type signatures

[Manuel M T Chakravarty]

Lennart Augustsson:
>
> On Thu, Apr 10, 2008 at 4:20 AM, Manuel M T Chakravarty <chak at cse.unsw.edu.au 
> > wrote:
> the five signatures
>
>  forall a. S a
>  forall b. S b
>  forall a b. S (a, b)
>  Int
>  S Int
>
> By alpha-convertible I mean the usual thing from lambda calculus,  
> they are identical modulo the names of bound variables.
> So only the first two are alpha-convertible to each other.
>
> If you involve any kind of reduction the equivalence test is no  
> longer trivial.
>
> All I'm asking for really, is to be able to paste in the type that  
> was inferred as the signature, without the compiler complaining.

I think a trivial, purely syntactic test, such as the one proposed,  
would generate at least as much puzzlement as the current situation  
does.  It would mean, you could not have String in your signature if  
the inferred signature has [Char].  I don't think that this is a  
solution at all.

Manuel