November 2007 - Haskell-Cafe

Properties of monads
by Radosław Grzanka 19 Nov '07

19 Nov '07

Hello, I am writing some toys programs to learn and try to apply Monads properties (without success, I must say). Although I spent half a day on this code: http://hpaste.org/3957 I couldn't simplify (shorten) getStrip function. After reading "Doing it with class" ( http://www.haskell.org/all_about_monads/html/class.html ) I had an impression that I could "collapse" cases using some Monads properties. But maybe I misunderstood something. Can anyone look at it and give me a pointers?? (I don't mind if the code becomes unreadable a bit.) Thank you, Radek.

2 4

HDBC-ODBC crashes on ghc 6.8
by Thomas Hartman 19 Nov '07

19 Nov '07

This minimal program below fails in 6.8.1, works for 6.6.1, for up to date HDBC-ODBC-1.1.3. (I tried installing both from package and from darcs head) Installed and running from cygwin. When I run it as runghc fail.hs I get a windows popup error "ghc.exe has encountered a problem and needs to close, sorry for the inconvenience". ghci, same behavior. When I run it as ghc -e 'main' fail.hs, I don't get a popup window, and no error message is output, but it fails all the same. (I know this because more complicated, non-minimal programs, fail.) So this would seem to be a problem with ghc 6.8.1. $ cat fail.hs import Database.HDBC import Database.HDBC.ODBC main = connectODBC "some valid connect string, works when run from 6.6.1, not from 6.8.1" --- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.

1 0

ANN: The Monad.Reader Issue 9: SoC special
by Wouter Swierstra 19 Nov '07

19 Nov '07

I am pleased to announce that a new issue of The Monad.Reader is now available: http://www.haskell.org/haskellwiki/The_Monad.Reader Issue 9 is a "Summer of Code Special" - it consists of three articles from student participants of Google's Summer of Code, describing the projects they worked on: * Thomas Schilling Cabal Configurations * Jason Dagit Darcs Patch Theory * Mathieu Boespflug TaiChi: how to check your types with serenity The Monad.Reader is a quarterly magazine about functional programming. It is less formal than a journal, but somehow more enduring than a wiki page or blog post. If you'd like to write something for the next issue of The Monad.Reader, please get in touch. I haven't fixed the deadline for the next issue just yet. I hope to send out an official call for copy fairly soon. Enjoy! Wouter This message has been checked for viruses but the contents of an attachment may still contain software viruses, which could damage your computer system: you are advised to perform your own checks. Email communications with the University of Nottingham may be monitored as permitted by UK legislation.

3 2

Performance problems with parallelizing QuickCheck using channels (Control.Concurrent.Chan)
by Gwern Branwen 18 Nov '07

18 Nov '07

(My previous email showed up as mangled in <http://www.haskell.org/pipermail/haskell-cafe/2007-November/034717.html>; I think the PGP signature was the problem; I removed it and sent it again, but it was put into moderation for size, and hasn't been released yet - so I'm sending this a third time but I'm removing attachments so it'll be under the annoying 20k size limit) So I've been trying to get my QuickCheck tests to run in parallel. I did take a look at Don's Parallel QuickCheck library <http://www.cse.unsw.edu.au/~dons/pqc.html>, but I didn't like how much code it had in it and I figured it'd be a good exercise to try to do myself. After quite a lot of help from the good folk of #haskell, I eventually came up with this: module Pcheck (parTest, parCheck) where import Control.Monad (replicateM_, liftM) import Control.Concurrent.Chan (newChan, writeChan, getChanContents) import Control.Concurrent (forkIO) import Test.QuickCheck (quickCheck', Testable()) -- | Takes a list of functions using parCheck, and returns True iff all return -- True. Evaluates them in parallel. parTest :: [IO Bool] -> IO Bool parTest = andTest . parList where andTest :: IO [Bool] -> IO Bool andTest = liftM and {- | Test in parallel. Forks off a QuickCheck 'n' times; QuickCheck tests using the proposition 't'. Returns True if all tests were passed, else False. Should be run with parallelizing options like with +RTS -N4 -RTS &etc. -} parCheck :: (Testable prop) => prop -> Int -> IO Bool parCheck t n = do chan <- newChan replicateM_ n $ forkIO $ (writeChan chan) =<< (quickCheck' t) liftM (and . take n) $ getChanContents chan -- | Takes a list of functions (presumably using parCheck) and evaluates all in parallel. parList :: [IO a] -> IO [a] parList fs = do chan <- newChan mapM_ (\m -> forkIO $ m >>= writeChan chan) fs liftM (take n) $ getChanContents chan where n = length fs I liked how simple the Channels library <http://haskell.org/ghc/docs/latest/html/libraries/base-3.0.0.0/Control-Conc…> seemed to be; I could just pass the channel as an argument and have every forkIO'd test simply chuck its Boolean result into it when done - which seem'd much simpler than using MVars and the other techniques for returning stuff from forkIO threads. And so it compiles, it runs tests correctly, and so on. But the problem is that it does so slowly. I have another module of equations about nuclear bombs called nuke.hs, which has a number of QuickCheck properties defined. Here's what happens when main is defined as 'parTest [the various tests..]': ./nuke +RTS -N7 -sstderr -RTS 40.57s user 46.55s system 116% cpu 1:14.61 total ./nuke +RTS -N6 -sstderr -RTS 40.72s user 47.66s system 117% cpu 1:15.50 total ./nuke +RTS -N5 -sstderr -RTS 42.33s user 49.08s system 116% cpu 1:18.67 total ./nuke +RTS -N4 -sstderr -RTS 43.71s user 48.41s system 117% cpu 1:18.48 total ./nuke +RTS -N3 -sstderr -RTS 41.51s user 48.25s system 114% cpu 1:18.10 total ./nuke +RTS -N2 -sstderr -RTS 42.28s user 47.18s system 115% cpu 1:17.39 total ./nuke +RTS -N1 -sstderr -RTS 27.87s user 18.40s system 99% cpu 46.498 total (From <http://hpaste.org/3886#a6>; compiled as "=ghc -v --make -threaded -O2 ./nuke.hs".) For some reason, running the parallel tests with a single thread is faster than running with 4 threads (I have a quad-core Intel processor)? I find this counter-intuitive to say the least. the par* functions are indeed operating in parallel, as evidenced by it using more than 100% CPU time, or, running on multiple cores, and all the tests are passed as True in both -N1 and -N[2-7] versions, so -N1 can't be bailing out early due to "and"'s laziness, and in general everything seems to be written correctly. I am perplexed by this. Is Chan simply a very inefficient way of parallelizing things? Is it not as parallel as I think? Or am I missing something else entirely? -- gwern .45 GIGN jya. wire ISI SADCC JPL embassy Recon World

2 1

Network.HTTP problem
by Radosław Grzanka 18 Nov '07

18 Nov '07

Hello, I have a problem with Network.HTTP module (http://www.haskell.org/http/) version 3001.0.0 . I have already mailed Bjorn Bringert about it but I didn't get answer yet so maybe someone here can help me. GHC v. 6.6.1 Ubuntu 7.10 x86_64 . I have turned on debug flag. Using get example (http://darcs.haskell.org/http/test/get.hs) I can download pages like this: $ ./get http://www.haskell.org/http/ <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> <html xmlns="http://www.w3.org/1999/xhtml"> <head> <title>Haskell HTTP package</title> <link href="style.css" rel="stylesheet" type="text/css" /> </head> <body> .... SNIP rest of the content SNIP .... Also the log contain content of this file. However, some links misbehaves like: $ ./get http://www.podshow.com/feeds/gbtv.xml ... no-output ... however I see content of this xml in debug file and wget downloads almost 250 kB of data. Also: $ ./get http://digg.com/rss/indexvideos_animation.xml ... hangs ... and debug file has size 0, but wget downloads the file I could suspect this is xml problem but: $ ./get http://planet.haskell.org/rss20.xml <?xml version="1.0"?> <rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/"> <channel> <title>Planet Haskell</title> <link>http://planet.haskell.org/</link> <language>en</language> <description>Planet Haskell - http://planet.haskell.org/</description> .... SNIP rest of the content SNIP .... so it works. Do you have any idea what is going on here? What goes wrong? What other (high level) modules could I use to download files through http? Cheers, Radek. -- Codeside: http://codeside.org/ Przedszkole Miejskie nr 86 w Lodzi: http://www.pm86.pl/

3 7

[Haskell-cafe] Letting the darcs test fail, if QuickCheck tests fail
by Henning Thielemann 18 Nov '07

18 Nov '07

When following the description on http://www.haskell.org/haskellwiki/How_to_write_a_Haskell_program#Add_some_… then darcs will run the QuickCheck tests on each 'darcs record', but the new patch is also accepted by darcs if one of the tests fail. What is the most simple way to let 'darcs record' fail, when a QuickCheck test fails?

3 3

Haskell performance question
by Dan Piponi 17 Nov '07

17 Nov '07

I see lots of shootout examples where Haskell programs seem to perform comparably with C programs, but I find it hard to reproduce anything like those figures when testing with my own code. So here's a simple case: I have this C program: #include <stdio.h> #define n 100000000 double a[n]; int main() { int i; for (i = 0; i<n; ++i) { a[i] = 1.0f; } for (i = 0; i<n-1; ++i) { a[i+1] = a[i]+a[i+1]; } printf("%f\n",a[n-1]); } And this Haskell program: import Data.Array.IO import Control.Monad n = 10000000 main = do a <- newArray (0,n-1) 1.0 :: IO (IOUArray Int Double) forM_ [0..n-2] $ \i -> do { x <- readArray a i; y <- readArray a (i+1); writeArray a (i+1) (x+y) } x <- readArray a (n-1) print x Even though 'n' is 10 times bigger in the C program it runs much faster than the Haskell program on my MacBook Pro with Haskell 6.6.1. I've tried lots of different combinations of flags that I've found in various postings to haskell-cafe but to no avail. What flags do I need to get at least within a factor of 2 or 3 of C? Am I using the wrong kind of array? Or is Haskell always going to be 15-20 times slower for this kind of numerical work? -- Dan

17 63

Re: [Haskell] AmeroHaskell
by Tim Chevalier 17 Nov '07

17 Nov '07

On 11/17/07, Derek Elkins <derek.a.elkins(a)gmail.com> wrote: > However, to put things in motion for something concrete at all, we're > hoping to put together a meeting taking place in the Portland area as > that seems most convenient to the most people who had registered > interest in AmeroHaskell and an easy place to find a venue. In that > vein, while no plans have been made, something aimed for the > January/February time-frame and probably hosted by Galois, if they are > willing, has been discussed. > In Portland, we've already had the first pdxfunc meeting: http://groups.google.com/group/pdxfunc While billed as a functional programming user group rather than a Haskell-specific group, we shouldn't duplicate effort. Cheers, Tim -- Tim Chevalier * catamorphism.org * Often in error, never in doubt "I don't know whether I believe in me, but I still believe in my friends" -- Nerissa Nields

2 3

Small optimisation question
by Andrew Coppin 17 Nov '07

17 Nov '07

Suppose I write something like this: foo :: [Int] foo = concat (replicate 4 [4,7,2,9]) The value of "foo" is completely determined at compile-time. So, will the compiler generate calls to concat and replicate, or will it just insert a large list constant here? Obviously, once somebody has completely examined the contents of "foo", after that point it won't matter either way. I'm just curiose. Concatinating some strings is cheap; I sometimes write constructs like the above using much more expensive operations. (Expensive in time; the space taken up by the result isn't that great.)

7 12

Showing Data.Ratio - different on GHC vs Hugs/Yhc
by Neil Mitchell 16 Nov '07

16 Nov '07

Hi Under Hugs and Yhc, showing a Ratio 1%2 gives "1 % 2". Under GHC showing 1%2 gives "1%2". Does the standard say anything about this? Is someone wrong? And how do Yhc/nhc/Hugs pass Bernouilli in the Nofib suite given that the output doesn't match? Thanks Neil

3 2