October 2002 - Haskell

Ph.D. Scholarship in Computer Science in Norway
by Bjarte M. Oestvold 16 Oct '02

16 Oct '02

[ Possible research areas includes functional programming, see below. -Bjarte ] Ph.D. Scholarship in Computer Science in Norway ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Østfold College in cooperation with the University of Oslo (UiO), Norway, offers a three year Ph. D. scholarship. Østfold College has the biggest computer science department among the Norwegian colleges and is selected as the national center for college computer science education in Norway. In addition to office space and supervision at Østfold College, where the main supervision is expected to take place, the Ph.D. candidate will receive office space and supervision at the department of informatics, UiO, through the group of professor Olaf Owe. The candidate is required to take relevant courses at UiO. The university of Oslo has a long tradition of outstanding computer science research, starting with the invention of object-oriented programming in the 1960's. The Ph.D. degree, called Dr.Scient. in Norway, will be awarded by UiO. The research area should be functional languages, program transformation or rewrite systems and must be relevant for automatic programming. The topic is chosen by the candidate in cooperation with the supervisors. Salary is from NOK 256200 to NOK 333300 per annum. More information is available from Roland Olsson, tel: +47-69215369, roland.olsson(a)hiof.no or http://www-ia.hiof.no/~rolando. Closing date for applications is 20 November, 2002. -- Bjarte

1 0

new version of edison?
by Hal Daume III 16 Oct '02

16 Oct '02

Hi all, I was wondering if anyone has a brand-spaking-new version of edison or anything like it. The edison docs still refer to ghc 4.06, which can't be good. If not, is there an edison-like project out there? - Hal -- Hal Daume III "Computer science is no more about computers | hdaume(a)isi.edu than astronomy is about telescopes." -Dijkstra | www.isi.edu/~hdaume

2 1

forall and class contexts
by Wolfgang Jeltsch 15 Oct '02

15 Oct '02

Hello, I want to define something like class (Eq (forall a. T a), Monad T) => C T where the context shall mean that T has to be an instance of Monad and that for every type a the type T a has to be an instance of Eq. Is such a thing possible with some Haskell implementation? Wolfgang

1 0

Q: GHC, Parallel libraries
by Jan Kybic 15 Oct '02

15 Oct '02

Hello, I would like to parallelize my program in Haskell but I have difficulties getting started. I would therefore appreciate any help or pointers. I am using ghc-5.04 on i386 Linux. I installed from a binary bundle on the ghc website. I thought I would just compile with --parallel and I would get acess to the 'par' primitive. But no, it seems that the parallel libraries are missing on my system. If I understood correctly, they are files with _mp in their name, right? I looked on the web for a precompiled binary version of those libraries, in vain. I found the GUM web page but it seems to be much outdated, based on the ghc-4.06. Is there any more recent version available? Then, I tried to get the ghc-5.04.1 source distribution and compile it from source. I found the building guide on the ghc site but it talks only about fptools in general and does not mention the ghc specificities. The build (./configure --prefix=$HOME ; make) fails. First of all it could not find 'readline', which I solved by some config.mk hacking, now it fails with: Main.hs: can't locate import `Exception' apparently because the binary bundle directory hierarchy is different from what the Makefile expects. I also tried to build only the libraries. It first failed with not being able to find the ghc-pkg-inplace, after changing this to the installed version it fails with Reason: No such file or directory File: ../../ghc/driver/package.conf Any help appreciated. Thanks, Jan -- ------------------------------------------------------------------------- Jan Kybic <kybic(a)ieee.org> Odyssee, INRIA, Sophia-Antipolis, France or <Jan.Kybic(a)sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845 http://kybic.host.sk/

3 2

RE: GHC, Parallel libraries
by Simon Peyton-Jones 15 Oct '02

15 Oct '02

| I looked on the web for a precompiled binary version of those | libraries, in vain. I found the GUM web page but it seems to be much | outdated, based on the ghc-4.06. Is there any more recent version available? Not as far as I konw. The parallel version of GHC is maintained and developed by the GPH team, http://www.cee.hw.ac.uk/~dsg/gph/ They are working on a GHC5.x version I believe, but it certainly does not come with an ordinary GHC distribution. Building it from source is unlikely to work, but I'm sure Phil Trinder and Hans Loidl would love your help in making it work! | Then, I tried to get the ghc-5.04.1 source distribution and compile it | from source. I found the building guide on the ghc site but it talks | only about fptools in general and does not mention the ghc specificities. | The build (./configure --prefix=$HOME ; make) fails. First of all it | could not find 'readline', which I solved by some config.mk hacking, | now it fails with: Main.hs: can't locate import `Exception' | apparently because the binary bundle directory hierarchy is different | from what the Makefile expects. Building GHC from source (non-parallel) should certainly work. Try again now; there was a moment a couple of days ago when the imports were messed up. What version of GHC are you compiling with? Simon

1 0

efficiency of mfix
by Hal Daume III 14 Oct '02

14 Oct '02

Hi all, I've been experimenting in ghc with mfix (I noticed on the cvs logs that mdo is going to be supported shortly :P), so I decided to see how it would work for me. My simple task is to normalize an array. That is, given an array, replace each value with a new value such that the array sums to one and that the ratios between values before and after normalization are the same. There is of course a two pass solutions, using a fold and a map. This is implementing using IOArrays, so this looks like: norm a = do t <- foldM (\t i -> (t+) `liftM` readArray a i) 0 [1..sz] mapM_ (\i -> readArray a i >>= writeArray a i . (/t)) [1..sz] where (_,sz) = Data.Array.IO.bounds a There is a one pass solution using mfix. This looks like: mfix (\ ~s -> ioaA a 1 s) where ioaA is defined as: ioaA a i s | i > sz = return 0 | otherwise = do s' <- ioaA a (i+1) s v <- readArray a i writeArray a i (v/s) return (s' + v) where (_,sz) = Data.Array.IO.bounds a Now, in practice, the two-pass solution is about twice as fast as the one-pass solution. I've looked a bit at the core generated, but without knowing what's going on internally, it's hard to say where the inefficiencies come from. I'm hoping someone can shed some light on this. Some timing results based on different array sizes: 50000 elements, 32m heap fix: 0.79u 0.04s 0:00.89 93.2% 0.77u 0.07s 0:00.89 94.3% 0.75u 0.10s 0:00.94 90.4% normal: 0.58u 0.01s 0:00.60 98.3% 0.53u 0.03s 0:00.66 84.8% 0.53u 0.05s 0:00.61 95.0% 100000 elements, 32m heap fix: 2.40u 0.10s 0:02.73 91.5% 2.38u 0.10s 0:02.54 97.6% 2.37u 0.13s 0:02.70 92.5% normal: 1.50u 0.03s 0:01.55 98.7% 1.49u 0.05s 0:01.57 98.0% 1.52u 0.03s 0:01.59 97.4% 200000 elements, 32m heap fix: 7.90u 0.16s 0:08.09 99.6% 7.99u 0.17s 0:08.38 97.3% 7.96u 0.21s 0:08.40 97.2% normal: 4.43u 0.06s 0:04.69 95.7% 4.38u 0.14s 0:04.75 95.1% 4.49u 0.07s 0:04.79 95.1% 400000 elements, 32m heap fix: 29.09u 0.34s 0:30.95 95.0% normal: 14.51u 0.16s 0:15.02 97.6% Using unboxed arrays is not possible in the fixed point version (loop). On the normal version, it just about triples the speed across the board. - Hal -- Hal Daume III "Computer science is no more about computers | hdaume(a)isi.edu than astronomy is about telescopes." -Dijkstra | www.isi.edu/~hdaume

2 3

RE: Bug? 'Storable' and 'with'
by Simon Marlow 14 Oct '02

14 Oct '02

> I am using ghc-5.04 and a code like: > > with c ( \c' -> hPutBuf h c' (sizeOf c)) > > fails with "Fail: Prelude.undefined" when "c" is a user defined type, > such as a pair: > > instance (Storable at,Storable bt) => Storable (at,bt) where > sizeOf (a,b) = sizeOf a + sizeOf b > alignment (a,b) = max (alignment a) (alignment b) > peek p = do a <- peek ((castPtr p):: Ptr at) > b <- peekByteOff ((castPtr p):: Ptr bt) (sizeOf a) > return (a,b) > poke p (a,b) = do poke ((castPtr p):: Ptr at) a > pokeByteOff ((castPtr p):: Ptr bt) > (sizeOf a) b sizeOf and alignment are not supposed to evaluate their arguments. You might try making the definitions in your instance above a little lazier: sizeOf z = sizeOf a + sizeOf b where (a,b) = z alignment z = max (alignment a) (alignment b) where (a,b) = z (feel free to use '~' if you prefer). Cheers, Simon

1 0

ghc MacOS X 10.2 behaviour
by Markus.Schnell＠infineon.com 14 Oct '02

14 Oct '02

Hi all, I'm working with Haskell on MacOS X 10.2.1 using the binaries from www.uni-graz.at/imawww/haskell. Now, when I compile some code with ghc I get two errors: "/tmp/ghc617.hc:283: conflicting types for 'GHCziTopHandler_runIO_closure'" and "/usr/local/lib/ghc-5.04/include/RtsAPI.h:125: previous declaration of 'GHCziTopHandler_runIO_closure'" Using ghci or compiling with -O2 doesn't show this error. Any ideas what's wrong? Thanks, Markus By the way, if there is a list more suited for the subject, please tell me.

1 0

Q: Bug? 'Storable' and 'with'
by Jan Kybic 14 Oct '02

14 Oct '02

Hi, I am using ghc-5.04 and a code like: with c ( \c' -> hPutBuf h c' (sizeOf c)) fails with "Fail: Prelude.undefined" when "c" is a user defined type, such as a pair: instance (Storable at,Storable bt) => Storable (at,bt) where sizeOf (a,b) = sizeOf a + sizeOf b alignment (a,b) = max (alignment a) (alignment b) peek p = do a <- peek ((castPtr p):: Ptr at) b <- peekByteOff ((castPtr p):: Ptr bt) (sizeOf a) return (a,b) poke p (a,b) = do poke ((castPtr p):: Ptr at) a pokeByteOff ((castPtr p):: Ptr bt) (sizeOf a) b On the other hand, if I replace 'with' with 'with1': with1 :: (Storable a) => a -> ( Ptr a -> IO b ) -> IO b with1 x f = do p <- mallocBytes (sizeOf x) let p'=castPtr p :: Ptr a poke p' x y <- f p' free p return y it seems to work. I could not even get to work 'malloc :: IO (Ptr (Int,Int))', while 'malloc :: IO (Ptr Int)' works fine. Am I doing something wrong, or is there a bug in the library code? Yours, Jan -- ------------------------------------------------------------------------- Jan Kybic <kybic(a)ieee.org> Odyssee, INRIA, Sophia-Antipolis, France or <Jan.Kybic(a)sophia.inria.fr>,tel. work +33 492 38 7589, fax 7845 http://www-sop.inria.fr/odyssee/team/Jan.Kybic/index.en.html

1 0

FLAIR 2003 Last CFP: track on Constraint solving and programming
by Stefano Bistarelli 13 Oct '02

13 Oct '02

apologies for multiple copies ------------------------ C A L L F O R P A P E R S Special Track at The 16th International FLAIRS Conference (FLAIRS-03) on CONSTRAINT SOLVING AND PROGRAMMING Held in co-operation The American Association for Artificial Intelligence http://www.cs.ucc.ie/~osullb/flairs2003/ ____________________________________________________________________________ _ Aims and Scope -------------- Constraints have emerged as the basis of a representational and computational paradigm that draws from many disciplines and can be brought to bear on many problem domains. This special track is concerned with all aspects of computing with constraints including algorithms, applications, environments, languages, models, and systems. We also solicit papers from any domain employing constraints, including computational linguistics, configuration, decision support, design, diagnosis, graphics, hardware verification, molecular biology, planning, qualitative reasoning, real-time systems, resource allocation, robotics, scheduling, software engineering, temporal reasoning, vision, visualization, and user interfaces. Papers that bridge disciplines or combine theory and practice or discuss novel reasoning methods are especially welcome. Special attention is focused around the use of constraint technologies in the networking, wireless and internet fields. Submissions ----------- We invite authors to submit papers on research on constraint solving and programming, with particular emphasis on assessing the current state of the art and identifying future directions. All submissions must be original work. Complete manuscripts must be no more than 10 double-spaced pages (3400 words) in length in postscript or pdf format. Please submit papers directly to FLAIRS (http://www.flairs.com/flairs2003/) Detailed submission guidelines are available from the FLAIRS-2003 web-site (http://www.cs.uwp.edu/flairssubsys/index.php3) Papers, of sufficiently high quality, will be selected to appear in a special issue of the International Journal on Artificial Intelligence Tools (IJAIT). Important Dates --------------- Deadline paper submissions October 25, 2002 Notification of acceptance January 7, 2002 Camera-ready deadline February 17, 2003 Organising Committee -------------------- Barry O'Sullivan (Primary Contact) Cork Constraint Computation Centre Department of Computer Science University College Cork, Ireland Email: b.osullivan(a)cs.ucc.ie Web: http://www.cs.ucc.ie/~osullb/ Stefano Bistarelli C.N.R. Istituto di Informatica e Telematica, Pisa and Universita "G. D'Annunzio" Dipartimento di Scienze, Pescara Em ail: Stefano.Bistarelli(a)iit.cnr.it Tel: +39-050-315.3438 Fax: +39-050-315.2593 Web: http://www.sci.unich.it/~bista Program Committee ----------------- Christian Bessiere, LIRMM, France Philippe Codognet, University of Paris and INRIA, France Boi Faltings, EPFL, Switzerland Helene Fargier, Laboratoire IRIT, Universite Paul Sabatier, France Thom Fruehwirth, University of Ulm, Germany Carmen Gervet, IC Parc, Imperial College, UK Javier Larrosa, UPC, Spain Francesca Rossi, University of Padov, Italy Marius-Calin Silaghi, Florida Institute of Technology Mark Wallace, IC Parc, Imperial College, UK Toby Walsh, University College Cork, Ireland

1 0