#9117: Coercible constraint solver misses one -------------------------------------+------------------------------------ Reporter: goldfire | Owner: nomeata Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.2 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by goldfire): Fair enough. I believe any of these problems exist with non-pathologically recursive newtypes, such as my `Bar` example in comment:21. But, I'm also happy enough to say "caveat programmer" if they use recursive newtypes with `Coercible`. We should make sure to say that somewhere, though. If we neglect recursive newtypes (which is reasonable, I would say) then the reordered solver (as it is in HEAD, as I understand) seems better than the one released with 7.8.2. But, how are we faring on the original request at the top? PS: Just realized that I never contributed to the wiki page. Will do shortly. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/9117#comment:27 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler