[GHC] #14386: GHC doesn't allow Coercion between partly-saturated type constructors

GHC ghc-devs at haskell.org
Wed Oct 25 02:12:17 UTC 2017


#14386: GHC doesn't allow Coercion between partly-saturated type constructors
-------------------------------------+-------------------------------------
        Reporter:  Iceland_jack      |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler          |              Version:  8.3
      Resolution:                    |             Keywords:  roles
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 nomeata):

 No, it would not be wrong. We just don't have the machinery yet.

 Interestingly, had you had

 {{{
 newtype NotOp cat a b = NotOp (cat b a)
 }}}
 then `:t Coercion :: Coercion (NotOp (NotOp cat)) cat` would work. Why?
 Because the axiom you get out of a `newtype` is eta-contracted as far as
 possible; in this case.
 {{{
 NotOp cat ~ cat
 }}}


 The thing with `Either` is a completely different story. Note that you are
 coercing _an argument of `Either`_ (still `Either` on both sides) whereas
 originally you are coercing between `Op` and it’s representation.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14386#comment:3>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler


More information about the ghc-tickets mailing list