[GHC] #12784: Typechecker regression in GHC 8.0.2 involving DefaultSignatures

[GHC]

#12784: Typechecker regression in GHC 8.0.2 involving DefaultSignatures
-------------------------------------+-------------------------------------
        Reporter:  RyanGlScott       |                Owner:
            Type:  bug               |               Status:  new
        Priority:  highest           |            Milestone:  8.0.2
       Component:  Compiler          |              Version:  8.1
      Resolution:                    |             Keywords:
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 rwbarton):

 I just mean to say that `monad-logger` should just fix its code, since the
 fix will surely be needed anyway.

 RyanGlScott, it's more wrong than that. Let's charitably assume that `t m`
 refers to the instance we are deriving, like `IdentityT m`. Then how can
 the body of the default declaration type check?

 {{{#!hs
 askLoggerIO = Trans.lift askLoggerIO
 }}}

 `askLoggerIO` is a method of `MonadLoggerIO` and we need it at type `m`.
 But we only have the constraints `MonadLogger (t m), MonadIO (t m)`, which
 are insufficient (and useless). The only way GHC could think this type
 checks is if BOTH `t m` and `m` refer to the instance being derived, which
 is terribly wrong.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12784#comment:4>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler