- Haskell-Cafe - Haskell.org

Haskell and html input elements
by PR Stanley 15 Nov '07

15 Nov '07

Hi back again! How easy/hard is it to control a haskell program through a web browser? Cheers Paul

3 3

Why are OCaml and Haskell being used at these companies?
by Galchin Vasili 15 Nov '07

15 Nov '07

Hello, Using http://www.indeed.com with the search phrase "Haskell functional", I found several Haskell positions inclusing Quantitative Functional Programmer (Analyst - Level)<http://www.indeed.com/rc/clk?jk=950afd03471e43cb&from=ja&cd=0B84_stXHsQIi--…> Credit Suisse - New York, NY Specific *Language* (embedded in *Haskell*) that will be... of competence in either *Haskell* or another statically-typed functional *language*. The ability to build... >From Credit Suisse - November 11, 3:32 PM I am looking for (objective.. i.e. not juts FPL cheerleading) opinions as to why Wall Street ( http://www.janestcapital.com/) and banking are now using OCaml and Haskell. I really want to understand what industrial markets are adopting FPLs and why in order to help push FPLs penetration into industry. Another company is http://www.galois.com Kind regards, Vasya

16 27

why PArr slower than list ?
by Albert Lee 15 Nov '07

15 Nov '07

I read the GHC/Data Parallel Haskell/GHC.PArr page http://haskell.org/haskellwiki/Data_Parallel_Haskell/GHC.PArr and make a simple test to compare the speed of PArr against List: {-# OPTIONS -fparr -fglasgow-exts #-} module Main where import GHC.PArr import System.CPUTime dotp :: Num a => [:a:] -> [:a:] -> a dotp xs ys = sumP [:x * y | x <- xs | y <- ys:] main = do t1 <- getCPUTime print $ sum [x*y|x<-[1..90000]|y<-[1..90000]] t2 <- getCPUTime print $ dotp [:1..90000:] [:1..90000:] t3 <- getCPUTime print $ t2 - t1 print $ t3 - t2 and I get the result: *Main> main 243004050015000 243004050015000 384540000000 1701841000000 My laptop is macbook macosx 10.4.8 and ghc-6.8 , anything wrong or I missed ?

3 2

some links broken in 6.8.1 documentation
by Daniil Elovkov 15 Nov '07

15 Nov '07

Hello folks A quick look at the 6.8.1 user's guide reveals some broken links: 1) Obtaining code coverage, pointing to http://haskell.org/ghc/docs/6.8.1/html/users_guide/hpc.html redirects to http://projects.unsafeperformio.com/hpc.html which tells that it's not found but guesses at another location, where it is found. 2) Concurrent and parallel haskell http://haskell.org/ghc/docs/6.8.1/html/users_guide/lang-parallel.html (and possibly other pages) have links to libraries according to the pre-base-split structure, like .../base/.. instead of .../parallel/..., which results in those links being broken. Thank you -- Daniil Elovkov

3 4

Re: [Haskell-cafe] Brazilian Haskellers ?
by PR Stanley 14 Nov '07

14 Nov '07

> > Hi brazilian haskellers, > >Wow! I knew the Haskell community has been growing... but there's a >brazillian of us? > >-- >Rich A mini UN, that's us.

1 0

cabal problem?
by Jens Blanck 14 Nov '07

14 Nov '07

I have problems building X11. I just installed ghc 6.8 but I got the same behaviour when asking it to use the old compiler. Jens > runghc Setup.hs configure Configuring X11-1.3.0.20071111... checking for gcc... gcc checking for C compiler default output file name... a.out checking whether the C compiler works... yes checking whether we are cross compiling... no checking for suffix of executables... checking for suffix of object files... o checking whether we are using the GNU C compiler... yes checking whether gcc accepts -g... yes checking for gcc option to accept ANSI C... none needed checking how to run the C preprocessor... gcc -E checking for X... libraries /usr/X11R6/lib64, headers /usr/X11R6/include checking for gethostbyname... yes checking for connect... yes checking for remove... yes checking for shmat... yes checking for IceConnectionNumber in -lICE... yes checking whether to build Xinerama... yes checking for egrep... grep -E checking for ANSI C header files... yes checking for sys/types.h... yes checking for sys/stat.h... yes checking for stdlib.h... yes checking for string.h... yes checking for memory.h... yes checking for strings.h... yes checking for inttypes.h... yes checking for stdint.h... yes checking for unistd.h... yes checking X11/extensions/Xinerama.h usability... yes checking X11/extensions/Xinerama.h presence... yes checking for X11/extensions/Xinerama.h... yes configure: creating ./config.status config.status: creating config.mk config.status: creating X11.buildinfo config.status: creating include/HsX11Config.h config.status: include/HsX11Config.h is unchanged config.status: creating include/X11_extras_config.h config.status: include/X11_extras_config.h is unchanged > runghc Setup.hs build Preprocessing library X11-1.3.0.20071111... Building X11-1.3.0.20071111... [ 1 of 17] Compiling Graphics.X11.Xlib.Types ( dist/build/Graphics/X11/Xlib/Types.hs, dist/build/Graphics/X11/Xlib/Types.o ) [ 2 of 17] Compiling Graphics.X11.Types ( dist/build/Graphics/X11/Types.hs, dist/build/Graphics/X11/Types.o ) [ 3 of 17] Compiling Graphics.X11.Xlib.Atom ( dist/build/Graphics/X11/Xlib/Atom.hs, dist/build/Graphics/X11/Xlib/Atom.o ) [ 4 of 17] Compiling Graphics.X11.Xlib.Color ( Graphics/X11/Xlib/Color.hs, dist/build/Graphics/X11/Xlib/Color.o ) [ 5 of 17] Compiling Graphics.X11.Xlib.Context ( Graphics/X11/Xlib/Context.hs, dist/build/Graphics/X11/Xlib/Context.o ) [ 6 of 17] Compiling Graphics.X11.Xlib.Display ( Graphics/X11/Xlib/Display.hs, dist/build/Graphics/X11/Xlib/Display.o ) [ 7 of 17] Compiling Graphics.X11.Xlib.Event ( dist/build/Graphics/X11/Xlib/Event.hs, dist/build/Graphics/X11/Xlib/Event.o ) [ 8 of 17] Compiling Graphics.X11.Xlib.Font ( dist/build/Graphics/X11/Xlib/Font.hs, dist/build/Graphics/X11/Xlib/Font.o ) [ 9 of 17] Compiling Graphics.X11.Xlib.Misc ( dist/build/Graphics/X11/Xlib/Misc.hs, dist/build/Graphics/X11/Xlib/Misc.o ) [10 of 17] Compiling Graphics.X11.Xlib.Region ( Graphics/X11/Xlib/Region.hs, dist/build/Graphics/X11/Xlib/Region.o ) [11 of 17] Compiling Graphics.X11.Xlib.Screen ( Graphics/X11/Xlib/Screen.hs, dist/build/Graphics/X11/Xlib/Screen.o ) [12 of 17] Compiling Graphics.X11.Xlib.Window ( Graphics/X11/Xlib/Window.hs, dist/build/Graphics/X11/Xlib/Window.o ) [13 of 17] Compiling Graphics.X11.Xlib.Image ( Graphics/X11/Xlib/Image.hs, dist/build/Graphics/X11/Xlib/Image.o ) [14 of 17] Compiling Graphics.X11.Xlib ( Graphics/X11/Xlib.hs, dist/build/Graphics/X11/Xlib.o ) [15 of 17] Compiling Graphics.X11.Xlib.Extras ( dist/build/Graphics/X11/Xlib/Extras.hs, dist/build/Graphics/X11/Xlib/Extras.o ) Warning: orphan instances: instance base:GHC.Read.Read [ X11-1.3.0.20071111:Graphics.X11.Xlib.Types.Rectangle] = $f1 [16 of 17] Compiling Graphics.X11.Xinerama ( dist/build/Graphics/X11/Xinerama.hs, dist/build/Graphics/X11/Xinerama.o ) [17 of 17] Compiling Graphics.X11 ( Graphics/X11.hs, dist/build/Graphics/X11.o ) /usr/bin/ar: creating dist/build/libHSX11-1.3.0.20071111.a > sudo runghc Setup.hs install root's password: Setup.hs: Warning: Unknown field 'build-type' Setup.hs: error reading ./.setup-config; run "setup configure" command?

3 2

Renaming constructors for readability
by Dougal Stanton 14 Nov '07

14 Nov '07

I wonder, is there an equivalent of the 'type' keyword for constructors? An example: -- create a pseudo-C pointer type -- which can point to a value or a -- null. type Pointer a = Maybe a -- int a = 3; -- int *pa = &a; ampersand :: t -> Pointer t ampersand a = Just a -- int b = *pa. star :: Pointer a -> a star (Just a) = a -- note this function behaves -- in an 'authentic' fashion ;-) To really complete the illusion it would be nice to replace the names Just and Nothing with PointerTo and Null. Then the constructors would really mean something. Is there a solution? -- Dougal Stanton dougal(a)dougalstanton.net // http://www.dougalstanton.net

5 7

Screen scraping with an interactive process: Buffering problems?
by Denis Bueno 14 Nov '07

14 Nov '07

Hi all, I'm writing some code to interact with an ACL2 [0] process. I'd like to be able to write a function test :: String -> IO Bool that will attempt to prove something by forking an ACL2 process and screen scraping its output, to see whether the conjecture was proved. The code below [1] is close: it launches the process, sends the conjecture to ACL2, and reads ACL2's output line-by-line until it finds "Proof succeeded" or "attempt has failed". This works well enough for the simple case, but there is a problem. In readACL2Answer, the code shouldn't have to check for success *and* failure, since the conditions are complementary. But if I change that code to check only for success, then it is possible (and indeed it happens every time) that if I send ACL2 a conjecture it can't prove, readACL2Answer blocks indefinitely -- and the print statements show that hWaitForInput has said there is more data on the Handle but hGetLine blocks trying to read it. The reason for this, as far as I can tell, is that there *is* more data -- a character, for example -- but no newline, so hGetLine blocks waiting for a newline. However, if I'm reading the BufferMode documentation [3] correctly, the fact that I've set my streams to LineBuffering should mitigate that scenario. Concretely, I know that ACL2 runs in a REPL and the last thing it prints is in fact an incomplete line -- "ACL2 !>". I'd rather not have to read every line character by character just to detect the uncommon case. Should I just do it character by character? Am I reading the BufferMode docs wrong, and if so, what is the correct interpretation? Is there better approach for the original problem I'm trying to solve? Thanks for reading this far. =] -- Denis [0] http://www.cs.utexas.edu/users/moore/acl2/ [1] import System.Process import System.IO type Proc = (Handle, Handle, Handle, ProcessHandle) testThm = test "(equal x x)" -- something acl2 can prove testBad = test "(equal x (not x))" -- something it can't -- | Pass a conjecture form to ACL2, attempt to prove it, and return a boolean -- indicating whether ACL2 was able to prove it. test :: String -> IO Bool test thm = do i@(stdin, _, _, pid) <- setUpACL2 success <- testProc i thm hPutStrLn stdin "(good-bye)" putStrLn "saying goodbye" exitCode <- waitForProcess pid case exitCode of ExitSuccess -> return success ExitFailure i -> error ("ACL2 bad exit code: " ++ show i) -- | Pass a conjecture vio the given process' input stream, and return whether -- it is proved. testProc :: Proc -> String -> IO Bool testProc i@(stdin, stdout, _, pid) thm = do hSetBuffering stdout LineBuffering hSetBuffering stdin LineBuffering success <- testInACL2 ("(thm " ++ thm ++ ")") i putStrLn "found success value" return success succeedRx = mkRegex "Proof succeeded[.]" failedRx = mkRegex "attempt has failed[.]" -- create acl2 process and return handles setUpACL2 :: IO Proc setUpACL2 = runInteractiveCommand "acl2" testInACL2 :: String -> Proc -> IO Bool testInACL2 str (pin, pout, perr, pid) = do putStrLn "printing to acl2..." hPutStrLn pin str putStrLn "getting acl2 result..." readACL2Answer pout -- | How long we should wait for ACL2 to start saying something. Five -- seconds, currently. readTimeout = 100 readACL2Answer :: Handle -> IO Bool -- I *should* be required *only* to check for *success* in the matchRegex -- statement below, instead of checking for success and failure. But if I do -- that, and the theorem to be proved fails, this function blocks indefinitely -- (at least I observe a much greater than 5 second wait period), even though -- it should only block for `readTimeout' milliseconds. I'm not sure why this -- is, but at least there is an acceptable workaround. readACL2Answer pout = hWaitForInput pout readTimeout >>= (\isReady -> if isReady then do putStrLn "getting line" line <- hGetLine pout putStrLn $ "read line: " ++ line -- Check for success or failure. case (matchRegex succeedRx line, matchRegex failedRx line) of (Nothing, Nothing) -> readACL2Answer pout (Nothing, Just _) -> return False (Just _, _) -> return True else do putStrLn "not ready" return False) [3] http://haskell.org/ghc/docs/latest/html/libraries/base-3.0.0.0/System-IO.ht…

2 4

More good performance results for ghc 6.8
by Don Stewart 14 Nov '07

14 Nov '07

Trying out some of the great language shootout programs with ghc 6.8 is producing nice results. For example, our "classic" cache-hammering, bitwise sieve benchmark is out of the box 10% faster with the new compiler. The (already rather good) benchmark is here (the same speed as the OCaml version under ghc 6.6): http://shootout.alioth.debian.org/gp4/benchmark.php?test=nsievebits&lang=all And timing with old and new ghc: ghc 6.6.1 Primes up to 81920000 4774995 Primes up to 40960000 2488465 Primes up to 20480000 1299069 ./A66 13 4.50s user 0.02s system 100% cpu 4.515 total ghc 6.8.1 Primes up to 81920000 4774995 Primes up to 40960000 2488465 Primes up to 20480000 1299069 ./A68 13 4.13s user 0.01s system 99% cpu 4.142 total Lovely work GHC HQ, when low level, highly tuned code like this gets magically faster! Once 6.8 is in Gentoo (or earlier...) we should see similar improvements across a range of shootout programs. -- Don

2 1

how does ghci and hugs handle redirecting of stdin/stdout?
by Galchin Vasili 14 Nov '07

14 Nov '07

Thanks, Vasya

1 0