[GHC] #14991: GHC HEAD regression involving TYPEs in type families

[GHC]

#14991: GHC HEAD regression involving TYPEs in type families
-------------------------------------+-------------------------------------
        Reporter:  RyanGlScott       |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.6.1
       Component:  Compiler (Type    |              Version:  8.5
  checker)                           |             Keywords:  TypeInType,
      Resolution:                    |  TypeFamilies
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  GHC rejects       |  Unknown/Multiple
  valid program                      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by goldfire):

 This is conceivably correct behavior. In the last two lines, all we know
 about the kinds of `a` and `b` is that they are `TYPE r1` and `TYPE r2`.
 Because you don't have instances to promote/demote `TYPE`, GHC gets stuck.

 In function type signatures, GHC defaults `RuntimeRep`s to `LiftedRep`.
 Perhaps it should, also, in type family instances.

 I don't know why this behavior changed.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14991#comment:1>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler