#8634: Code valid in GHC 7.6 is impossible to move over GHC 7.7 (because of liberal coverage condition) -----------------------------------+--------------------------------------- Reporter: danilo2 | Owner: Type: bug | Status: new Priority: high | Milestone: Component: Compiler | Version: 7.7 Resolution: | Keywords: Operating System: | Architecture: Unknown/Multiple Unknown/Multiple | Difficulty: Unknown Type of failure: None/Unknown | Blocked By: Test Case: | Related Tickets: #1241, #2247, #8356 Blocking: | -----------------------------------+--------------------------------------- Comment (by danilo2): Replying to [comment:4 rwbarton]:
That doesn't seem to be a problem, as those will be instances of different classes `Method2`, `Method3`. Ouch, of course, you are right. I wanted to tell about the problem you've discovered later:
(...) With these changes your second program runs for me with GHC HEAD as long as I comment out `testx`--which is not used in the program.
And this is the problem I was talking about on StackOverflow and my original question - I marked it as "the problematic function" in GHC 7.7. It is now impossible, to make this function an "Vectora associated method" `method3`. We are generating all type class instances with template Haskell reifying needed functions to get their type signatures. If such function does not compile, reify would also fail. Even writing type class instances manually, it is impossible to make instances of `Method3` and `Call` for it, because it does not compile. By the way, this function will make more sense if it will be written as: {{{#!haskell testx v x = call (method1 x) (OneTuple "test") }}} , which means, calling associated "method1" of "x" with one argument of "test". It is still working in GHC 7.6. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8634#comment:5 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler