It's because there's an application of a strict function, the dictionary selector 'a'. In general, if f is strict then f (g x) is turned into case g x of y { DEFAULT -> f y } But it is rather inscrutable. It didn't arise from a strictness annotation put in by the user. The right thing to do is to fix the bytecode generator. Simon | -----Original Message----- | From: Julian Seward (Intl Vendor) | Sent: 08 February 2002 13:55 | To: 'Dean Herington'; glasgow-haskell-users@haskell.org | Cc: Simon Peyton-Jones | Subject: RE: inscrutable warning | | | | The warning says that the interpreter has ignored a | polymorphic case, since it's hard to implement. Your program | should still behave the same, although possibly less strictly | than you intended. | | What really mystifies us is why there is such a thing in the | program in the first place. Using -ddump-sat we can see that | there is indeed a | case ... of { DEFAULT -> ... } | but how it came to be, I do not know. | | We may investigate further. | | J | | | Can someone explain the following warning? | | | | ------------------------------- | | module Warning where | | | | type A = IO () | | | | class CA t | | where a :: t -> A | | instance CA () | | where a = return | | instance (CA t) => CA (IO t) | | where a = (>>= a) | | | | data B = B A | | | | class CB t | | where b :: t -> B | | instance (CA t) => CB (IO t) | | where b = B . a | | ------------------------------- | | | | % ghci Warning.hs | | ___ ___ _ | | / _ \ /\ /\/ __(_) | | / /_\// /_/ / / | | GHC Interactive, version 5.02.2, | | for Haskell | | 98. | | / /_\\/ __ / /___| | http://www.haskell.org/ghc/ | | \____/\/ /_/\____/|_| Type :? for help. | | | | Loading package std ... linking ... done. | | Compiling Warning ( Warning.hs, interpreted ) | | WARNING: ignoring polymorphic case in interpreted mode. | | Possibly due to strict polymorphic/functional constructor args. | | Your program may leak space unexpectedly. | | | | Ok, modules loaded: Warning. | | Warning> | | | | _______________________________________________ | | Glasgow-haskell-users mailing list | | Glasgow-haskell-users@haskell.org | | http://www.haskell.org/mailman/listinfo/glasgow-haskell-users | | |