Frankly, I was surprised that GND assumes any class constraint at all. Why should it be different than other forms of `deriving`? Just infer the class constraint based on the constraints that arise when type-checking
#12814: Should GND infer an instance context when deriving method-free classes? -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: Type: bug | Status: new Priority: normal | Milestone: 8.2.1 Component: Compiler (Type | Version: 8.0.1 checker) | Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: #11369, #12810 | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by RyanGlScott): Replying to [comment:2 goldfire]: the generated definition. I'm not sure how the thing you're imagining is different from what actually happens. GND must infer a constraint (applied to the representation type for the newtype) in order to typecheck any calls to `coerce` that GND generates. This is quite similar to stock deriving. For example, in: {{{#!hs data Foo a b = Foo b deriving Show }}} You must infer `Show b` in order to typecheck the generated `showsPrec` function. The only difference in this scenario is that there are no stock derivable classes without methods, but with GND, such a thing is possible. So it's quite a different beast. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/12814#comment:3 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler