[GHC] #8678: Derivin `Functor` complains about existential type

[GHC]

#8678: Derivin `Functor` complains about existential type
-------------------------------------+------------------------------------
        Reporter:  heisenbug         |            Owner:
            Type:  bug               |           Status:  new
        Priority:  normal            |        Milestone:
       Component:  Compiler          |          Version:  7.7
      Resolution:                    |         Keywords:
Operating System:  Unknown/Multiple  |     Architecture:  Unknown/Multiple
 Type of failure:  None/Unknown      |       Difficulty:  Unknown
       Test Case:                    |       Blocked By:
        Blocking:                    |  Related Tickets:
-------------------------------------+------------------------------------

Comment (by heisenbug):

 Replying to [comment:3 monoidal]:
 > My guess is that the type of `Standard` is rewritten to `m ~ S n => a ->
 NonStandard m a` (which is existential in `n`). Simpler test case: `data A
 t a where A :: A Int a deriving Functor`.

 Yeah, this is Henry Ford polymorphism ("you can have your car in any color
 as long as it is black"). But still, `deriving Functor` should only
 concern itself with the last parameter, which is manifestly not
 existential.
 Also, the parameter referred to has kind `Nat`, and thus is irrelevant for
 endofunctors on `*`. Furthermore even your example establishes a
 //functional dependency// between `t ~ Int`, so it should not count as
 existential.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8678#comment:4>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler