[GHC] #8827: Inferring Safe mode with GeneralizedNewtypeDeriving is wrong

[GHC]

#8827: Inferring Safe mode with GeneralizedNewtypeDeriving is wrong
-------------------------------------+------------------------------------
        Reporter:  goldfire          |            Owner:
            Type:  bug               |           Status:  new
        Priority:  normal            |        Milestone:  7.8.1
       Component:  Compiler          |          Version:  7.8.1-rc2
      Resolution:                    |         Keywords:
Operating System:  Unknown/Multiple  |     Architecture:  Unknown/Multiple
 Type of failure:  None/Unknown      |       Difficulty:  Unknown
       Test Case:                    |       Blocked By:
        Blocking:                    |  Related Tickets:  8226, 8745
-------------------------------------+------------------------------------

Comment (by dterei):

 So given the late hour, perhaps (sadly) as others have suggested, we keep
 the status quo for 7.8 and leave GND and Coercion unsafe. Be nicer if we
 could have solved for 7.8 but that's on me for not be reachable easily or
 following the mailing list.

 Going forward, my suggestion or a similar proposal such as nomeata's in
 comment:27 sound the nicest to me. We could potentially argue for this
 with data as well by seeing how prevalent coercion and GND is on abstract
 types that don't have explicit roles.

 If everyone can quickly reach consensus though on the above proposal then
 I'm very happy with that as well :)

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8827#comment:28>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler