[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):

 What of this:

 * nominal by default when constructors aren't all exported
 * representational by default when all constructors are exported

 I understand the concern with (C) and the burden, I'm just not sure how
 high it is with some heuristics applied as suggested above. I like this
 approach (although I may be missing some corner cases) as it matches what
 I feel is a good high level starting point that GND and surrounding
 infrastructure (coercion) shouldn't allow you to write code that you
 couldn't write by hand unless explicitly opted-in by the library author.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8827#comment:23>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler