Kindness of strangers (or strangeness of Kinds)

[AntC]

Edward Kmett <ekmett <at> gmail.com> writes:

> On Mon, Jun 11, 2012 at 9:58 PM, AntC <anthony_clayden <at> clear.net.nz> 
wrote:
>>
>> [snip ...]
>>
>> Could we have :k (->) :: OpenKind -> * -> *  -- why not?
> 
> I don't quite understand why you would want arbitrary kinded arguments, but 
only in negative position. 
>

Thanks Edward, oops I've used the wrong terminology, sorry for the confusion. 
I didn't mean OpenKind but AnyKind. I put that only in a negative position 
more to sharpen the question, but also because I assumed the result from (->) 
would have to be grounded in Kind *; and then at least one of its arguments 
would also have to be grounded in Kind *.


I think perhaps(?) more PolyKindness is on the horizon: 
http://hackage.haskell.org/trac/ghc/wiki/GhcKinds  (section on GADKs, and sub-
pages on KindPolymorphism and ExplicitTypeApplication). I guess GHC is getting 
there by small steps, and doesn't yet have powerful enough Kind refinement nor 
Kind equality constraints, nor interleaving of Type and Kind inference.