[GHC] #14132: Report an error for a missing class instance before an error for type family instances of an associated type.

[GHC]

#14132: Report an error for a missing class instance before an error for type
family instances of an associated type.
-------------------------------------+-------------------------------------
        Reporter:  duog              |                Owner:  duog
            Type:  feature request   |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler (Type    |              Version:  8.2.1
  checker)                           |
      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 simonpj):

 > This means moving ReporterSpecs between report1 and report2 doesn't work
 as is

 I don't agree.  I believe you can move "non-tv-eqs", "homo-eqs" and
 "otehr-eqs"; they are not "truly insoluble".

 Ah: I see that `trulyInsoluble` is a poorly-named predicate.  It is only
 applied to constraints in `wc_insol`, and they are things like `Int ~
 Bool`, but NOT things like `a ~ Int` (which are no necessarily insoluble).
 It'd be much clearer if `trulyInsoluble` really did check for truly-
 insoluble constraints.

 Separately I'd been thinking that separating `wc_insol` from `wc_simple`
 is a waste of effort; it'd be better to combine them.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14132#comment:4>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler