On Jun 10, 2010, at 05:59 , Philip K.F. Hölzenspies wrote:
[holzensp@ewi1043:work/FPPrac]% ghci BugDemo.hs GHCi, version 6.12.1: http://www.haskell.org/ghc/ :? for help Loading package ghc-prim ... linking ... done. Loading package integer-gmp ... linking ... done. Loading package base ... linking ... done. Loading package ffi-1.0 ... linking ... done. [1 of 1] Compiling BugDemo ( BugDemo.hs, interpreted ) Ok, modules loaded: BugDemo. *BugDemo> 5 5 *BugDemo> :t 5 5 :: (P.Num t) => t *BugDemo> :q Leaving GHCi. [holzensp@ewi1043:work/FPPrac]% ghci -fno-implicit-prelude BugDemo.hs GHCi, version 6.12.1: http://www.haskell.org/ghc/ :? for help Loading package ghc-prim ... linking ... done. Loading package integer-gmp ... linking ... done. Loading package base ... linking ... done. Loading package ffi-1.0 ... linking ... done. [1 of 1] Compiling BugDemo ( BugDemo.hs, interpreted )
on the commandline: Warning: -fno-implicit-prelude is deprecated: use -XNoImplicitPrelude or pragma {-# LANGUAGE NoImplicitPrelude #-} instead Ok, modules loaded: BugDemo. *BugDemo> 5
<interactive>:1:0: Not in scope: `>>'
This doesn't surprise me; when putting it in the module, it affects only that module. When using either command line version, it affects *everything*... and what's breaking is not your definition of Number, but the ghci expression printer (which, being in IO, is doing something like (print it >> putStr "\n"). Since the command line option has global effect, the Prelude's (>>) isn't defined for ghci's guts either. -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allbery@kf8nh.com system administrator [openafs,heimdal,too many hats] allbery@ece.cmu.edu electrical and computer engineering, carnegie mellon university KF8NH