- Hugs-Bugs - Haskell.org

Lazy conflict reports
by Ralf Hinze 19 Oct '01

19 Oct '01

I am not sure whether this has been reported before. At least, it's not listed on http://www.cse.ogi.edu/PacSoft/projects/Hugs/pages/bugsandfeatures.htm --- Hugs doesn't implement the lazy conflict reports. Thus, if you define module A where head = ":-)" then hugs complains Definition of variable "head" clashes with import The clash shouldn't be reported as there is no ambiguous *use* of `head'. If we change module `A' to module A where import Prelude hiding (head) head = ":-)" then everything works as expected. However, if we import `A' then the error shows up again module B where import A yo = 1 Entity "head" imported from module "A" already defined in module "Prelude" Several on my programs rely on this Haskell feature, so this is a continuous stumbling block ... Cheers, Ralf

1 0

i have a problem
by Syndromtx＠aol.com 15 Oct '01

15 Oct '01

dear worker of the huskell inc. im writing to you because i have a problem installing the winhugs on my computer, i have windows 98 and fallowing message before it gets down and i cant do anything else: prelude.hs does not define standart type "Word" after that i get a fatal error message and by clicking ok (the only choice) the program closes... i dont have office installed, is that why i have problems? i would be very very thankfull if you could help me, even though i know that this isnt the right email adress i wrote to. but please help me, i really need that help thx a lot for reading this already,

1 0

job for you!!!
by TZar 14 Oct '01

14 Oct '01

Здравствуйте! Я хочу поделиться с Вами информацией, как можно легально, быстро и сравнительно просто заработать несколько тысяч долларов. Вы улыбнулись? Улыбайтесь :-), но это РЕАЛЬНОСТЬ. Естественно, что Вам понадобится некоторое терпение, настойчивость в достижении этой цели, познания в психологии на уровне интуиции и, конечно же, желание. Еще обязательно нужен доступ к Интернету и электронной почте. Если Вас это заинтересовало, отправьте, пожалуйста, заявку на ОБА приведенных ниже адреса и Вам будет выслан файл с инструкциями. ЕСЛИ ВАС ЭТО НЕ ИНТЕРЕСУЕТ - ПРОСЬБА НЕ ОТВЕЧАТЬ НА ЭТО ПИСЬМО! Поэтому можете сразу нажать на Del,... но можете поступить как я - отвитил пустым письмом с темой "Заявка...". Просто был в некотором недоумении, что же именно можно написать в "заявке"? Да и не хотелось показаться смешным, если бы это оказалась шутка друзей. ВНИМАНИЕ!!! Пожалуйста, не направляйте Ваши заявки с помощью кнопки "ответить отправителю", "ответить на выбранное письмо", или "reply" на ящик, с которого Вы получили данное письмо (в таком случае оно не будет прочитано!), поэтому адресуйте их ТОЛЬКО сюда - tzar(a)iteranet.ru tzarishere(a)mtu-net.ru Знайте, это не спем, это реальное предложение. Спасибо, что прочитали это письмо. Всего доброго, удачи! С уважением, Виктор. P.S. Пожалуйста, укажите в теле письма адрес, по которому я должен выслать Вам инструкцию.

1 0

Hugs for Macintosh 2001/10/12
by Hans Aberg 12 Oct '01

12 Oct '01

I made an update to the version that I and Pablo Azero used to develop; it is at http://www.matematik.su.se/~haberg/Computer/Hugs98Feb2001MacBin20011012.sit The sources I used to compile these, using Metrowerks CodeWarrior Pro 5 are at http://www.matematik.su.se/~haberg/Computer/Hugs98Feb2001MacSrc20011012.sit If somebody is willing to put this stuff up elsewhere, please let me know. As hinted, I do not develop this one anymore, and I do not think that Pablo has time for that either. This is essentially an version I made the last fall (2000), updated to use the Hugs Feb 2001 kernel. It was just an experiment, so try it out as is (but I do not think there should be any problems). I think people would be better off to get hold of MacOS X, and use a version using its BSD shells. If somebody is interested in making a GUI or a version capable of handling AppleEvent's, that should be under MacOS X I think, and not building on the current version. The (main) differences from the previous versions are: - It uses the Feb 2001 Hugs kernel. - One can now drop files onto Hugs when it is up and running. -- I removed the DropUNIX package and implemented AppleEvent's directly instead. - Both MacOS '\r' and UNIX '\n' are accepted as newlines when reading files. (The DOS newline, which uses both control characters, will be read as two newlines.) - There is an experimental package for sending and receiving AppleEvent's. If you do not use this feature, it should not be a bother to you. Hans Aberg

1 0

WinHugs crashes reversing infinite lists
by carlos.scheidegger＠via-rs.net 02 Oct '01

02 Oct '01

Hello all. I think I found a bug on Hugs for Windows 32 (both on the GUI version and the console). If you try to reverse an infinite list, for example: reverse (repeat 0) reverse [1..] This will make Hugs segfault. Of course, reversing an infinite list is not something you should do, but so isn't segfaulting :) GHC for W2K just freezes (as would be expected). Hope I helped, Carlos

1 0

incomplete overloading resolution
by Ralf Laemmel 30 Sep '01

30 Sep '01

Dear Hugs-Bugsers, what a funny bug ... import Monad class Monad m => C1 m x class (Monad m, C1 m x) => C2 m x where c2 :: x -> m x instance Monad m => C1 m Bool instance C2 Maybe Bool where c2 = return test :: Maybe Bool test = c2 True 1. If you evaluate "test", then you get as expected "Just True". 2. Now remove the "Monad M" constraint in the class C2. This is of course legal and semantics-preserving since the monad constraints is implied by C1 anyway. 3. Now evaluate "test" again. Oops, it diverges. I did this Hugs Version February 2001 under Solaris and Linux. Command line option -98 is needed of course. Funny enough, if CTRL-C the evaluation of "test" and you try it the second time (without reloading anything etc.), then you see "Program error: {_Gc Black Hole}". Of course, there is no such black hole. I extracted the above fragment from a huge problem. The scheme is the following. It seems have to do with multi-parameter classes. It definitely has to do with multi-layered class hierarchies where one class has a class-wide superclass, as C2 has C1 in the example. It seems that the superclass is properly propagated during type-inference/checking but it is not properly propagated, as for as code determination in overloading resolution. Please, let me know if I can be of further assistance. I am actually amazed how general this scheme is (3 classes, 2 parameters). How does it come that it was not detected much earlier. Anyway, good luck, Ralf P.S.: I have more problems in a larger application. Certain overloaded expressions diverge. I can recover from that usually by using explicit function types for top-level functions, where again -- as in C1 above -- I repeat some constraints which are derivable. I was not able to extract a simple program for that problem. But I would be glad to help you by checking if the problem is gone after you did the bug fix. -- Dr.-Ing. Ralf Laemmel CWI & VU, Amsterdam, The Netherlands http://www.cwi.nl/~ralf/ http://www.cs.vu.nl/~ralf/

1 0

Working with MAC OSX
by Robin Harper 28 Sep '01

28 Sep '01

Hi, Just to let you know I have just finished compiling the feb distribution under the unix shell in MAC OSX. Brand new to Haskell so can only do a few tests, but it seems to work fine. The only problem I had was that configure seems to set the HAVE_LABELS_AS_VALUES wrong, changing it back to 0 allowed it to compile. Robin

1 0

bugs
by Guilherme Fernandes 24 Sep '01

24 Sep '01

the library not funcional of hugs98 i don´t import SOEGraphics ________________________________________________ Don't E-Mail, ZipMail! http://www.zipmail.com/

1 0

RE: Prelude and (:) and []((:), []) bugs?
by Simon Peyton-Jones 21 Sep '01

21 Sep '01

It's hard to distinguish a constructor from a class in an export list. S | -----Original Message----- | From: Alastair David Reid [mailto:reid@cs.utah.edu] | Sent: 20 September 2001 19:38 | To: Simon Marlow | Cc: Ian Lynagh; haskell(a)haskell.org; hugs-bugs(a)haskell.org | Subject: Re: Prelude and (:) and []((:), []) bugs? | | | | > Ah, I forgot that you can't export a constructor on its own. | | You can't? | | I probably knew this once but looking at it now, it seems | kinda surprising. Haskell's module system is supposed to be | just namespace control --nothing more-- so why is it | preventing me from doing something which is perfectly safe | and well-defined? | | I'll readily agree that there's no strong motivation for | exporting a constructor on its own (I think the only reason | Hugs allows it is just so we can export (:) from the Prelude) | but what is the motivation for disallowing it? | | -- | Alastair Reid reid(a)cs.utah.edu | http://www.cs.utah.edu/~reid/ | | | _______________________________________________ | Hugs-Bugs mailing list | Hugs-Bugs(a)haskell.org | http://www.haskell.org/mailman/listinfo/hugs-bugs |

1 0

RE: Prelude and (:) and []((:), []) bugs?
by Simon Marlow 20 Sep '01

20 Sep '01

Ian Lynagh writes: > On Thu, Sep 20, 2001 at 01:32:54PM +0100, Simon Marlow wrote: > > (:) is allowed in an export list; it is just a normal operator. > > An export is > > export -> qvar > | qtycon [(..) | ( qcname1 , ... , qcnamen )] (n>=0) > | qtycls [(..) | ( qvar1 , ... , qvarn )] (n>=0) > | module modid > > and we need to match (:). We can clearly reduce this to [ proof that this isn't possible deleted ] Ah, I forgot that you can't export a constructor on its own. In which case, I would like to submit a bug report for Hugs: Hugs allows constructors to be named in an export list, and it shouldn't (see Section 5.2 in The Report, point 2 in the enumerated list). Cheers, Simon

2 1