Hello, No, that is not the case, all imported modules are from packages which are already compiled and functional.
I found that removing the .o and .hi files and recompiling the source made it work. However, should either of these things be necessary? I think users oughtn't have to memorize and execute a list of haphazard tricks, for instance like reading the modules they import and visually checking that they start with 'module' (can I do this with sed? awk?), or removing output files which shouldn't be considered part of the compiler's input in the first place, to get valid code to compile. That doesn't lead to a very efficient workflow, and as such it is inconsistent with the rest of the Haskell experience. Furthermore, the error message doesn't even make sense - how can GHC claim that 'main' is not defined, when it is checking the type of a function called 'main'? I'm not saying that bugs like this are not to be expected, but I think that fixing them should be a high priority. Error messages that give the user nothing to go by have the effect of causing them to lose concentration and forget what they were doing as they puzzle over how to proceed. The mailing list (thank you Jeremy) is responsive and helpful, but using it still introduces a delay of a couple of hours in the software production cycle. Just my $0.02. Best regards, Frederik On Sat, May 12, 2007 at 05:55:37PM -0700, Jeremy Shaw wrote: > Hello, > > If one of the modules you are importing (for example, Fu.NewLogging), > does *not* contain a line at the top like: > > > module Fu.NewLogging where > > then I think you will get an error that looks something like that. I > am not sure if this is your problem, but if it is, it is an easy fix > :) > > hope this helps, > j. > > At Sat, 12 May 2007 22:13:36 +0100, > Frederik Eaton wrote: > > > > [1 <text/plain; us-ascii (7bit)>] > > Hello, > > > > I'm suddenly getting the following error when I compile a program: > > > > bayesian-sets.hs:1:0: > > The main function `main' is not defined in module `Main' > > When checking the type of the main function `main' > > > > I don't know what that means. Perhaps the error could be made more > > human-friendly? Here is the output of 'ghc -v': > > > > [1]$ ghc -v --make $A.hs -package vectro -lstdc++ > > -fallow-incoherent-instances > > Glasgow Haskell Compiler, Version 6.6.20070420, for Haskell 98, compiled by > > GHC version 6.6.20070420 > > Using package config file: > > /home/frederik/arch/i386//lib/ghc-6.6.20070420/package.conf > > Using package config file: > > /home/frederik/.ghc/i386-linux-6.6.20070420/package.conf > > wired-in package base mapped to base-2.0 > > wired-in package rts mapped to rts-1.0 > > wired-in package haskell98 mapped to haskell98-1.0 > > wired-in package template-haskell mapped to template-haskell-2.0 > > Hsc static flags: -static > > *** Chasing dependencies: > > Chasing modules from: bayesian-sets.hs > > Stable obj: [] > > Stable BCO: [] > > compile: input file bayesian-sets.hs > > Created temporary directory: /tmp/ghc8785_0 > > *** Checking old interface for main:Main: > > [1 of 1] Compiling Main ( bayesian-sets.hs, bayesian-sets.o ) > > *** Parser: > > *** Renamer/typechecker: > > > > bayesian-sets.hs:1:0: > > The main function `main' is not defined in module `Main' > > When checking the type of the main function `main' > > *** Deleting temp files: > > Deleting: /tmp/ghc8785_0/ghc8785_0.s > > Warning: deleting non-existent /tmp/ghc8785_0/ghc8785_0.s > > Upsweep partially successful. > > *** Deleting temp files: > > Deleting: > > link(batch): upsweep (partially) failed OR > > Main.main not exported; not linking. > > *** Deleting temp files: > > Deleting: > > *** Deleting temp dirs: > > Deleting: /tmp/ghc8785_0 > > > > Thanks, > > > > Frederik > > > > -- > > http://ofb.net/~frederik/ > > [2 bayesian-sets.hs <text/x-haskell; us-ascii (7bit)>] > > {-# OPTIONS_GHC -fglasgow-exts -fth #-} > > module Main where > > > > import Foreign.C.Types > > import Control.Arrow > > import System.Environment > > import Data.Typeable > > > > import Fu.GenUtil > > import Fu.NewLogging > > import Fu.Prepose > > import Fu.Domain > > import Fu.HList > > > > import Vector.Sparse > > import Vector.TypeVector > > import Vector.Sparse.Ops > > import Vector.Base > > import Vector.FileIO > > import qualified Vector.Sparse.Wrappers as W > > import Vector.Sparse.Raw > > import Vector.Sparse.RawTypes > > import Vector.Sparse.ArrowUtil > > > > --type I = CInt > > --type I = CUInt > > --type I = CInt > > --type E = Bool > > > > type I = CULLong > > type E = Double > > type V = Sparse I E > > > > type RV = RSV E I > > > > type SV a = Sparse I E E a > > > > e0 :: E > > e0 = 1.0 > > > > c :: E > > c = 2 > > > > main :: IO () > > main = do > > withLogLevel Normal $ do > > qsR::RV <- load "data/query-sess-2e6" > > $(typeVec (__::RV) ("query","session")) qsR $ \ (x_QS::(SV (L query, L > > session))) -> do > > putStrLn $ "sum="++show(vsum x_QS) > > let x_SQ = trans x_QS > > let mask_Q = mask $ vsum $ vcurry x_QS > > let qCount = vsum mask_Q > > let m_S = (vsum $ vcurry x_SQ) /. qCount > > let alpha_S = c .* m_S > > let beta_S = c .* (mask m_S - m_S) > > -- let v_S = vsum $ vcurry d_SQ > > putStrLn $ "(vsum alpha_S, vsum beta_S)="++show (vsum alpha_S, vsum > > beta_S) > > putStrLn $ "(nnz alpha_S, nnz beta_S)="++show (nnz alpha_S, nnz beta_S) > > putStrLn $ "qCount="++show qCount > > vseq m_S $ vseq x_QS $ vseq alpha_S $ do > > t <- profile $ > > putStrLn $ "vsum (m_S >@ x_QS)"++show(vsum (m_S >@ x_QS)) > > putStrLn $ "t="++show t > > -- test about how long computation of q_S will take: > > t <- profile $ > > putStrLn $ "vsum (m_S + alpha_S)"++show(vsum (m_S + alpha_S)) > > putStrLn $ "t="++show t > > return () > > [3 <text/plain; us-ascii (7bit)>] > > _______________________________________________ > > Glasgow-haskell-bugs mailing list > > Glasgow-haskell-bugs@haskell.org > > http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs > _______________________________________________ > Glasgow-haskell-bugs mailing list > Glasgow-haskell-bugs@haskell.org > http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs > -- http://ofb.net/~frederik/ _______________________________________________ Glasgow-haskell-bugs mailing list Glasgow-haskell-bugs@haskell.org http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
