[GHC] #14147: Confusing error messages with PolyKinds and superclasses

[GHC]

#14147: Confusing error messages with PolyKinds and superclasses
-------------------------------------+-------------------------------------
        Reporter:  enolan            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler          |              Version:  8.2.1
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
                                     |  Unknown/Multiple
 Type of failure:  None/Unknown      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by enolan):

 OK, I'm willing to accept that code that compiles without `PolyKinds`
 shouldn't necessarily compile with it turned on.

 I still think this is a bad error message though. Yes, `k` is a variable
 that exists, but it's not one I wrote. From the user's perspective this
 message is very unhelpful. Perhaps something along these lines could be
 appended to the message when an error like this is emmited: "where k is an
 inferred kind parameter in Tagged's kind: `forall {k}. k -> * -> *`". Then
 I'd know where `k` comes from and can figure out that I can add a
 `Typeable` constraint for it or fix it at `*`.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14147#comment:2>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler