#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 goldfire): I'm happy with Ryan's patch, too. But I'm still think my new approach would be simpler than the existing one. (a) is already done. That's !TcGenDeriv. (b) is already done: `tcInstDcls` (or similar) (c) will be the set of unsolved wanted constraints left over after solving. These will have to be checked for exoticness, but that's it! (d) is simply taking the result of (c) and sticking it in the instance decl. There is a matter that (c) outputs `Type` and (d) will want `HsContext` -- but I think we already have a way of squirreling a `Type` inside of an `HsType`, so this shouldn't be bad. (e) would be necessary, and it would be repeated work. So this new approach is probably slower than the existing one. In exchange, we get to delete the first several hundred lines of !TcDerivInfer (`inferConstraints`). In any case, if it's not broken, perhaps we shouldn't fix it. I just find that special-casing the 0-method case to be a bit odd. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/12814#comment:9 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler