[GHC] #11712: Very confusing error message with -fprint-explicit-kinds

[GHC]

#11712: Very confusing error message with -fprint-explicit-kinds
-------------------------------------+-------------------------------------
        Reporter:  bgamari           |                Owner:
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.0.1
       Component:  Compiler (Type    |              Version:  7.10.3
  checker)                           |
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Incorrect         |  Unknown/Multiple
  warning at compile-time            |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by goldfire):

 The reason for the inconsistency is because, without `-fprint-equality-
 relations`, all of `(~)`, `(~~)`, and `(~#)` render as `(~)`. In a program
 that doesn't mention equality (or mentions only `(~)`), any of these may
 appear. In order not to scare users with three equality relations, I
 taught GHC to print just one. But I'm very uncertain about this decision.
 I asked on devs a while ago, and the best suggestion there was to render
 the heterogeneous ones as heterogeneous when, in fact, the kinds on their
 sides differ -- but only then.

 Perhaps I'll implement that alongside the other suggestion here.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11712#comment:3>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler