- Hugs-Bugs - Haskell.org

errors to stdout?
by Andrew Pimlott 23 Mar '03

23 Mar '03

Hugs dumps errors, traces, and other diagnostics to stdout. It even has the rather perverse #define errorStream stdout (except on windows, it appears) so I assume there is a reason for it. However, I can't imagine the reason, and it is quite annoying to me. I am running "December 2001" hugs on Debian GNU/Linux. I looked at the CVS sources, and trace now seems to go to stderr, but errors still seem to go to stdout (I haven't tried it to be sure). Would you accept a patch to change everything to stderr? Andrew

2 1

failure to check completeness in instance declarations
by Andrew Tolmach 21 Mar '03

21 Mar '03

The following file is accepted by hugs (Nov.02 version on linux) without complaint. Attempting to evaluate z causes a Segmentation fault (not unreasonably, perhaps!). As far as I can tell, completeness is never being checked statically. Sometimes it is successfully checked dynamically and causes an "Undefined member" error. ----------- data Foo a = Foo a instance Eq (Foo a) where z = Foo 1 == Foo 2 ---------------

2 1

existential types and functional dependencies
by Christian Maeder 20 Mar '03

20 Mar '03

The following module is accepted by ghc (with -fglasgow-exts) but rejected by hugs: module ParseAny where class Syntax tag as | tag -> as where parse :: tag -> String -> as data AnyAs = forall tag as. Syntax tag as => MkAs tag as data AnyTag = forall tag as. Syntax tag as => MkTag tag parseAny :: AnyTag -> String -> AnyAs parseAny (MkTag tag) str = MkAs tag $ parse tag str -- ------------------------------------------------------- Hugs mode: Restart with command line option +98 for Haskell 98 mode Reading file "/home/linux-bkb/hugs/lib/hugs/libraries/Hugs/Prelude.hs": Reading file "/home/linux-bkb/hugs/lib/hugs/libraries/Prelude.hs": Reading file "ParseAny.hs": Type checking ERROR "ParseAny.hs":13 - Cannot justify constraints in explicitly typed binding *** Expression : parseAny *** Type : AnyTag -> String -> AnyAs *** Given context : () *** Constraints : Syntax _0 a When I omit the functional dependency "| tag -> as", ghc yields a similar error Cheers Christian

1 0

Your message to Hugs-Users awaits moderator approval
by hugs-users-admin＠haskell.org 18 Mar '03

18 Mar '03

Your mail to 'Hugs-Users' with the subject A very new website Is being held until the list moderator can review it for approval. The reason it is being held: Message body is too big: 143966 bytes but there's a limit of 40 KB Either the message will get posted to the list, or you will receive notification of the moderator's decision.

1 0

MarshalUtils.hs
by Dennis Sidharta 18 Mar '03

18 Mar '03

I tried to run a program that imports SOEGraphics, and I got this error message when hugs tried to read MarshalUtils.hs: "Reading file ... ... Reading file "C:\Program Files\Hugs98\lib\exts\MarshalUtils.hs": Parsing ERROR "C:\Program Files\Hugs98\lib\exts\MarshalUtils.hs":2 - Syntax error in module definition (unexpected keyword "with"; possible cause: -W option not used.) " FYI, I am using WinXP. -- __________________________________________________________ Sign-up for your own FREE Personalized E-mail at Mail.com http://www.mail.com/?sr=signup

1 0

Hugs internal error, parsing
by Scott Turner 17 Mar '03

17 Mar '03

I'm using :version -- Hugs Version November 2002 If I make a .hs file with the two lines: import Maybe Need. And do a :load IE.hs, the output is Reading file "IE.hs": Reading file "/usr/local/lib/hugs/lib/Maybe.hs": Reading file "IE.hs": INTERNAL ERROR: parseInput

1 0

Direct use of if-then-else with function type cause "Program error"
by Mike Gunter 14 Mar '03

14 Mar '03

With the following code > run f = f 7 (-3) > programError wd initVal = (if initVal < 0 then showAsUnsiged else const show) wd initVal > noProblem wd initVal = f wd initVal where f = if initVal < 0 then showAsUnsiged else const show > showAsUnsiged width nIn = show n where n | nIn < 0 = 2^width + nIn > | otherwise = nIn I get: $ hugs HugsIfFuncBug.lhs __ __ __ __ ____ ___ _________________________________________ || || || || || || ||__ Hugs 98: Based on the Haskell 98 standard ||___|| ||__|| ||__|| __|| Copyright (c) 1994-2002 ||---|| ___|| World Wide Web: http://haskell.org/hugs || || Report bugs to: hugs-bugs(a)haskell.org || || Version: November 2002 _________________________________________ Haskell 98 mode: Restart with command line option -98 to enable extensions Reading file "/usr/lib/hugs/libraries/Hugs/Prelude.hs": Reading file "/usr/lib/hugs/libraries/Prelude.hs": Reading file "HugsIfFuncBug.lhs": Hugs session for: /usr/lib/hugs/libraries/Hugs/Prelude.hs /usr/lib/hugs/libraries/Prelude.hs HugsIfFuncBug.lhs Type :? for help Main> run noProblem "125" Main> run programError " Program error: {instShow_v14_v1224 (showAsUnsiged instIntegral_v34 instOrd_v27 instNum_v30 (Num_fromInt (Real__sc0_Real (Integral__sc0_Integral instIntegral_v34)) 7))} Main> :q [Leaving Hugs] Thanks. mike

1 0

Hugs documentation update
by Henk-Jan.van.Tuyl 06 Mar '03

06 Mar '03

L.S., The web page http://cvs.haskell.org/Hugs/pages/hugsman/libs.html states: The libraries Directory, Time, and CPUTime, are not currently supported. As there are files Directory.hs, Time.hs, and CPUTime.hs in the {Hugs}\lib directory, I suppose this is not true anymore. Greetings, Henk-Jan van Tuyl --

1 0

type inference bug
by Mark Tullsen 06 Mar '03

06 Mar '03

Here's a (serious) bug in the type-inferencer of the November 2002 Hugs (tested using hugs98-Nov2002-rh7.3-1.i386.rpm): data T a b = T {f1 :: a, f2 :: b} updateF1 x t = t{f1 = x} the type inferred for updateF1: observed behavior: updateF1 :: a -> T b c -> T a d expected behavior: updateF1 :: a -> T b c -> T a c FYI, this bug didn't exist in the Dec 2001 version. - Mark

2 1

Concurrent Haskell bug?
by Daniel McNeill 03 Mar '03

03 Mar '03

{- Hi, Possible Concurrent Haskell bug Here is an attempt at concurrent, data-driven assertions for Haskell. The code has been simplified to just work for type Char: -} import Concurrent import IOExts(unsafePerformIO) assert :: (Char -> Bool) -> Char -> Char assert p x = unsafePerformIO $ do mv <- newEmptyMVar forkIO $ check p (listen mv) return (demand mv x) check :: (a -> Bool) -> a -> IO () check p x | p x = return () | otherwise = putStrLn "Assertion failed!" demand :: MVar Char -> Char -> Char demand mv c = unsafePerformIO $ do putMVar mv c; yield; return c listen :: MVar Char -> Char listen mv = unsafePerformIO $ takeMVar mv main :: IO () main = putChar (assert (isLower) 'z') {- When main is evaluated, it causes evaluation of assert, which returns the data it was called on, but also forks a "check" IO action to check predicate p on the data. This check is synchronised using "listen" which waits on an MVar into which the data will be placed when available. The data is placed here by evaluation of "demand" (which is returned by assert). Demand yields after putting the data, to allow the check to be carried out before computation is continued. In this example, the data is immediately available so the result should simply be 'z'. However, the actual output is: Thread raised exception: no more threads (deadlock?) which must then be interrupted with Ctrl-C. If listen is replaced with the following version, the printout "takeMVar: 'z'" occurs after the message about the exception, but execution still halts. listen mv = unsafePerformIO $ do val <- takeMVar mv putStrLn ("takeMVar: " ++ show val) return val I've tried the code in GHC and it works fine. Any ideas? Cheers, Dan -}

1 0