[GHC] #9200: Milner-Mycroft failure at the kind level

[GHC]

#9200: Milner-Mycroft failure at the kind level
----------------------------------------------+----------------------------
        Reporter:  ekmett                     |            Owner:
            Type:  bug                        |           Status:  new
        Priority:  normal                     |        Milestone:
       Component:  Compiler (Type checker)    |          Version:  7.8.2
      Resolution:                             |         Keywords:
Operating System:  Unknown/Multiple           |     Architecture:
 Type of failure:  GHC rejects valid program  |  Unknown/Multiple
       Test Case:                             |       Difficulty:  Unknown
        Blocking:                             |       Blocked By:
                                              |  Related Tickets:
----------------------------------------------+----------------------------

Comment (by simonpj):

 Richard, I have tried to address your questions on the wiki page.  Better
 now?

 I still don't know what the difference between bad and good polymorphism
 is, but maybe it doesn't matter now, since you say that it is
 "tangential".

 You say "Because open type families do not have a body, they would still
 need their own kind inference story, where unconstrained meta-variables
 default to *." but I don't understand that.  Add a subsection for it?
 (Open type families simply have their own, complete, kind signature, no?

 I'm happy with (PARGEN) if you are happy with the typing rule including
 the side condition.

 Simon

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9200#comment:14>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler