- Haskell - Haskell.org

LDTA 2004 - Call for Participation
by jas＠di.uminho.pt 09 Feb '04

09 Feb '04

Apologies if you receive multiple copies of this message. ********************************************************** *** Fourth Workshop on *** *** Language Descriptions, Tools and Applications *** *** LDTA 2004 *** *** *** *** Satellite event of ETAPS 2004 *** *** *** *** In Co-Operation with ACM SIGPLAN *** *** *** *** APRIL, 3, 2004 *** *** BARCELONA, SPAIN *** *** *** *** http://www.di.uminho.pt/LDTA04 *** ********************************************************** --> Deadline for early registration: February, 15 The ETAPS online registration form is available at http://www.lsi.upc.es/etaps04/Registration/registration-frame.html - The LDTA 2004 Preliminary Program is available at http://www.cs.lth.se/Research/LDTA2004/ - Invited speaker: Jim Cordy, Queen's University, Canada Title of the talk: TXL - A Language for Programming Language Tools and Applications

1 0

behavioral difference between GHC and GHCi
by S. Alexander Jacobson 09 Feb '04

09 Feb '04

This code works echos lines properly under GHCi, but just sucks in input when running the GHC compiled executable. import IO main= do x<-hGetLine stdin putStr x main Am I doing something wrong? -Alex- _________________________________________________________________ S. Alexander Jacobson mailto:me@alexjacobson.com tel:917-770-6565 http://alexjacobson.com

3 2

Making Haskell web development easier
by S. Alexander Jacobson 08 Feb '04

08 Feb '04

I decided to try developing a web app using Haskell/HaskellWebServer/HaXML and found the barriers both very high and perhaps easily lowered. Could someone out there: -- 1. Provide a link to the Haskell web server CODE on <http://www.haskell.org/practice.html> (or, if it has been end of lifed, remove it from the CVS repository) 2. Document anon CVS access to haskell.org in some obvious location? 3. Add a README for the package saying * this the package doesn't build with enclosed Makefile * this the package doesn't buld on win32 because it depends on the posix package * STRONGLY suggest that people use the sourceforge package rather than this one --- Also, what does it take to get posix/HWS working on win32? I suspect that much of posix is easily portable, but that a few functions are not. I don't know if HWS is dependent on these functions? If HWS (and other packages) actually use the portable subset of posix, perhaps it makes sense to create a separate posix_LowestCommonDenominator package. If not, how much effort is involved in making the posix package work on win32? -Alex- _________________________________________________________________ S. Alexander Jacobson mailto:me@alexjacobson.com tel:917-770-6565 http://alexjacobson.com

1 0

[Fwd: Midland Graduate School 2004 / APPSEM Spring School 2004]
by Conor McBride 05 Feb '04

05 Feb '04

Hello all I'm forwarding this announcement on behalf of Thorsten Altenkirch. Conor

1 0

What are possible causes of "C stack overflow" in Hugs?
by Graham Klyne 05 Feb '04

05 Feb '04

I'm getting a "C stack overflow" error from Hugs when processing a moderately large dataset, but not doing anything that looks unreasonably complex or involved. As far as I can tell, the code causing the problem is about here: [[ -- Return list of distinct labels used in a graph graphLabels :: (Label lb) => [Arc lb] -> [lb] graphLabels gs = graphLabels1 gs [] graphLabels1 (t:gs) ls = graphLabels1 gs $ foldl (flip addSetElem) ls (arcLabels t) graphLabels1 [] ls = ls ]] I think the function graphLabels1 is set to call itself recursively to a depth of a little over 1000. Can this kind of recursion blow the Hugs stack? This code has been working quite happily for some time. The called function arcLabels is quite trivial, and addSetElem is pretty simple too: [[ addSetElem :: (Eq a) => a -> [a] -> [a] addSetElem e es = if e `elem` es then es else e:es ]] ... I changed the function 'graphLabels' thus: [[ graphLabels :: (Label lb) => [Arc lb] -> [lb] graphLabels gs = nub $ concat $ map arcLabels gs ]] and the C stack overflow goes away. But I've been burned in the past with serious performance problems using nub, so I'm wary of this. #g ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact

2 2

partition in Data.List bug
by Koji Nakahara 04 Feb '04

04 Feb '04

Hi, I've encountered a bug of partition in Data.List shipped with ghc-6.2 and hugs(Nov 2002) It cannot work with an infinite list. Example: > partition (==0) (cycle [0,1]) does not return and so not conform to the specification in the Haskell 98 and the library document. I found some discussion on this bug in the haskell mailing-list archive, e.g. http://www.mail-archive.com/haskell@haskell.org/msg07789.html Is there some good reason not to change this behaviour? (And if it is, I think that the library document should be fixed.) Patch: --- List.hs.orig Thu Feb 5 04:15:59 2004 +++ List.hs Thu Feb 5 04:16:31 2004 @@ -240,8 +240,8 @@ {-# INLINE partition #-} partition p xs = foldr (select p) ([],[]) xs -select p x (ts,fs) | p x = (x:ts,fs) - | otherwise = (ts, x:fs) +select p x ~(ts,fs) | p x = (x:ts,fs) + | otherwise = (ts, x:fs) -- @mapAccumL@ behaves like a combination -- of @map@ and @foldl@; --- Koji Nakahara

1 0

Re: [Haskell] Weaving the Web with Haskell
by Graham Klyne 04 Feb '04

04 Feb '04

If I can get the POpen module working under Windows, we have a plausible route to achieving this using existing XSLT programs. See my message: http://haskell.org/pipermail/haskell/2004-February/013522.html Indeed, this could be a powerful tool for adding functionality quickly to Haskell on the major platforms (*nix and Windpows). Any help with this would be appreciated! #g -- At 20:55 03/02/04 -0700, Seth Kurtzberg wrote: >Don't forget about XSLT translation. In the commercial projects that I >have worked recently, using XSLT to translate to printed and screen form >representations. > >If I can help with this aspect, or any other aspect of the project, please >let me know. ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact

1 0

RE: [Haskell] RE: Hugs/GHC incompatibility
by Simon Peyton-Jones 04 Feb '04

04 Feb '04

Yes, that's it. We don't want an extra conditional test on every store, plus perhaps an array of bits to record which items have been filled. Simon | -----Original Message----- | From: haskell-bounces(a)haskell.org [mailto:haskell-bounces@haskell.org] On Behalf Of Iavor S. | Diatchki | Sent: 03 February 2004 18:23 | To: Wolfgang Jeltsch | Cc: The Haskell Mailing List | Subject: Re: [Haskell] RE: Hugs/GHC incompatibility | | hi, | i am speculating here, but i guess that if double initializations were to | result in _|_ one would have to check if a cell has already been | initialized, | and if so replace it with _|_. | currently ghc probably simply overwrites the array elements. | i personally prefer ghc's behavior. | -iavor | | Wolfgang Jeltsch wrote: | | >Am Dienstag, 3. Februar 2004 18:17 schrieb Simon Peyton-Jones: | > | > | >>>I think, I've found the problem. GHC seems to not implement | >>>Data.Array.array correctly. The implementation doesn't adhere to the | >>>following sentence from the Data.Array.IArray.array documentation: If any | >>>two associations in the list have the same index, the value at that index | >>>is undefined. | >>> | >>>Could this be fixed? | >>> | >>> | > | > | > | >>[...] | >> | >> | > | > | > | >>We don't intend to fix this because fixing it will have a significant | >>overhead on every array construction. | >> | >> | > | >Sorry, but I don't see where the overhead should come from. | > | > | > | >>[...] | >> | >> | > | > | > | >>Simon | >> | >> | > | >Wolfgang | > | >_______________________________________________ | >Haskell mailing list | >Haskell(a)haskell.org | >http://www.haskell.org/mailman/listinfo/haskell | > | > | | | -- | ================================================== | | Iavor S. Diatchki, Ph.D. student | | | Department of Computer Science and Engineering | | | School of OGI at OHSU | | | http://www.cse.ogi.edu/~diatchki | | ================================================== | | _______________________________________________ | Haskell mailing list | Haskell(a)haskell.org | http://www.haskell.org/mailman/listinfo/haskell

1 0

ANNOUNCE: HaskellDB 0.5
by Anders Höckersten 03 Feb '04

03 Feb '04

We are pleased to announce the first alpha release of our "remake" of HaskellDB. HaskellDB is a Haskell library for expressing database queries and operations in a type safe and declarative way. HaskellDB compiles a relational algebra-like syntax into SQL, submits the operations to the database for processing, and returns the results as ordinary Haskell values. HaskellDB uses HSQL to interface with standard database drivers, and is thus compatible with several standard databases (at the moment of writing, MySQL (only versions >=4.1), PostgreSQL and most ODBC databases). The original version of HaskellDB was written by Daan Leijen and is available at http://www.haskell.org/haskellDB/ This alpha release works pretty well so far, but has not been tested much, so we expect it to contain a lot of bugs. If you try it out and find something you suspect to be a bug, please contact us and we will try to help you as much as we are able. We also welcome all feature requests and suggestions for improvements. For more information and download see: http://haskelldb.sourceforge.net/ The developers can be reached at haskelldb-users(a)lists.sourceforge.net or on #haskelldb at irc.freenode.net. Thanks to Daan Leijen for his original version of HaskellDB. We would also like to thank Krasimir Angelov for all his help with HSQL, and for being nice enough to release a new version of it just in time for our release. Also, a special thank you to John Hughes for helping us out with everything regarding this project. Happy hacking, and don't try this on any production databases! :) / The HaskellDB development team

1 0

RE: Hugs/GHC incompatibility
by Simon Peyton-Jones 03 Feb '04

03 Feb '04

| I think, I've found the problem. GHC seems to not implement Data.Array.array | correctly. The implementation doesn't adhere to the following sentence from | the Data.Array.IArray.array documentation: | If any two associations in the list have the same index, the value at that | index is undefined. | Could this be fixed? It's a documented incompatibility with H98 http://www.haskell.org/ghc/docs/latest/html/users_guide/bugs-and-infelic ities.html#HASKELL98-DIVERGENCE We don't intend to fix this because fixing it will have a significant overhead on every array construction. Actually I rather think H98 should have left the behaviour of double indices undefined, rather than defining it to be bottom. But it's too late now. Simon

3 2