Short term: GHC should not crash. Maybe it should just reject the
#14498: GHC internal error: "not in scope during TC but it passed the renamer" -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler (Type | Version: 8.2.1 checker) | Keywords: Resolution: | PatternSynonyms Operating System: Unknown/Multiple | Architecture: Type of failure: Compile-time | Unknown/Multiple crash or panic | Test Case: Blocked By: | Blocking: Related Tickets: #14288 | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by RyanGlScott): Replying to [comment:3 simonpj]: program. I agree that until we implement #14288, this shouldn't crash. In fact, given that `kk` doesn't technically scope over the pattern at the moment, I'm surprised that GHC doesn't come up with a fresh `kk` type variable in the pattern `(App (Typeable :: TypeRep (a::kk -> k')) n)` (in fact, if you change it to `(App (Typeable :: TypeRep (a::kk2 -> k')) n)`, then it compiles). This makes me suspect that the code which renamed pattern synonyms is doing something suspicious at the moment.
Slightly longer term: we need to decide if we want to adopt the proposal in commment:15 of #14288. Would this be worth a short GHC proposal? It's certainly a change in the surface language.
To answer your two questions: * Yes. I've been lacking the free time to look at comment:15 of #14288, but perhaps I can take a look this week. * I suppose we could put this forth as a proposal. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/14498#comment:4 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler