#14201: Implement ideas from "Compiling Pattern Matching to Good Decision Trees" -------------------------------------+------------------------------------- Reporter: AndreasK | Owner: AndreasK Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by bgamari): The assessment in comment:4 sounds pretty reasonable to me. I agree that there's little reason to lose sleep over the `f (raisesLaterCatchedException) (causesMatchFailure)` case; such code produce warnings under any production-worthy flag set and users will almost certainly understand that this sort of code is fragile. The `f (raisesException1) (raisesException2)` case is a tad more worrying since users may be mislead by the (change in the) resulting exception. However, as you point out, the Report does say that both exceptions are admissible and I don't think we should be shy about taking advantage of this freedom if there really is performance on the table. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/14201#comment:5 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler