[GHC] #10114: Kind mismatches with AnyK in rank-2 types

[GHC]

#10114: Kind mismatches with AnyK in rank-2 types
-------------------------------------+-------------------------------------
        Reporter:  cam               |                   Owner:
            Type:  bug               |                  Status:  new
        Priority:  normal            |               Milestone:
       Component:  Compiler          |                 Version:  7.8.4
      Resolution:                    |                Keywords:
Operating System:  Unknown/Multiple  |            Architecture:
 Type of failure:  GHC rejects       |  Unknown/Multiple
  valid program                      |               Test Case:
      Blocked By:                    |                Blocking:
 Related Tickets:                    |  Differential Revisions:
-------------------------------------+-------------------------------------

Comment (by goldfire):

 I'm not sure I understand comment:3. What are you worried about?

 I will also say that it's possible the right answer is

 {{{
 type T = forall k1. (f :: k1 -> *) (a :: k1). f a
 }}}

 so that `T :: *`. My understanding is that it's generally more common to
 restrict the body of a forall to have kind `*`. GHC doesn't do this, but
 all the papers about System FC do.

 In other work, I changed the typing rule for forall accordingly, and
 something broke horribly (eta-reducing newtypes, maybe?), but perhaps we
 don't have to infer a type that has a non-`*` kind as the body of a
 forall.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10114#comment:4>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler