[GHC] #10141: Kind inference regression in closed type families

[GHC]

#10141: Kind inference regression in closed type families
-------------------------------------+-------------------------------------
        Reporter:  goldfire          |                   Owner:  goldfire
            Type:  bug               |                  Status:  new
        Priority:  highest           |               Milestone:  7.10.1
       Component:  Compiler          |                 Version:  7.10.1-rc2
      Resolution:                    |                Keywords:
Operating System:  Unknown/Multiple  |            Architecture:
 Type of failure:  None/Unknown      |  Unknown/Multiple
      Blocked By:                    |               Test Case:
 Related Tickets:                    |                Blocking:
                                     |  Differential Revisions:
-------------------------------------+-------------------------------------

Comment (by goldfire):

 Responding to comment:3:

 No, the return kind could well be `*`. I'll rewrite with explicit kinds:

 {{{
 type family G (a :: k) :: * where
   G * Int  = Bool
   G * Bool = Int
   G * a    = a
 }}}

 This `G` doesn't really use its kind-polymorphism, but the definition kind
 checks.

 However, I think you're right about the suggestion for a CUSK, here and in
 other cases. And I think it's possible for GHC to have at least some
 heuristics for when a CUSK is the answer. Specifically, it could include a
 note about CUSKs in error messages that arise from kind mismatches in non-
 CUSK right-hand sides.

 And, yes, this would make for a decent test case. Thanks for the
 suggestion.

 I'm very unsure I can get to this by Friday (the next RC release), though!

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10141#comment:6>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler