- Haskell-Cafe - Haskell.org

Deducing a type signature
by R J 19 May '10

19 May '10

Bird 1.6.3 requires deducing type signatures for the functions "strange" and "stranger." Are my solutions below correct? (i) strange f g = g (f g) Assume g :: a -> b. Then f :: (a -> b) -> c. But since g :: a -> b,f g :: a, so c = a. Therefore, f :: (a -> b) -> a, and g (f g) :: a.Therefore, strange :: ((a -> b) -> a) -> (a -> b) -> a. (ii) stranger f = f f Assume f :: a -> b. Since "f f" is well-typed, type unification requiresa = b. Therefore, f :: a -> a, and stranger :: (a -> a) -> a. _________________________________________________________________ Hotmail is redefining busy with tools for the New Busy. Get more from your inbox. http://www.windowslive.com/campaign/thenewbusy?ocid=PID28326::T:WLMTAGL:ON:…

3 2

Extensible Records and Functional References
by Günther Schmidt 19 May '10

19 May '10

Hi all, I just read "Functional References are a cheap and cheerful technique for working with the existing (non-extensible) record system, and may be of interest to extensible record implementers. A good implementation can be found on ..." on http://hackage.haskell.org/trac/ghc/wiki/ExtensibleRecords Does this mean that there is a way to implement extensible records via functional references? Günther

1 0

[reactive] A pong and integrate
by Limestraël 19 May '10

19 May '10

Hello, I know there is a special mailing list for reactive, but I got no answers when sending to it. Also, maybe my problem is not totally reactive-centered, it concerns integration of an object speed to obtain its position with FRP. As a newbie with reactive, I'm trying to make a simple example, namely a pong. But I'm stuck with the paddle animation. Here is a how I proceed : A paddle: data Paddle = Paddle { paddleCol :: Double, paddleCenter :: Behavior Double } The paddle moves along a column, so only it's Y coordinate changes. paddleCenter is the Y coordinate of its center. This is how I create a paddle: data Direction = PaddleUp | PaddleDown | PaddleStill deriving (Eq, Show) newPaddle :: Behavior Direction -> Double -> Paddle newPaddle direction column = let position = fmap ((initY ^+^)) $ integral *beat* $ fmap ((speed *) . factor) direction factor PaddleUp = -1 factor PaddleDown = 1 factor _ = 0 speed = 50 -- pixels per second initY = 400 -- suppose we have a 800x600 window: paddle is right at the middle *beat = atTimes [0, 0.01..]* in Paddle column position Basically, a paddle is driven by a behavior which tells him to go up, down or to stand still. My problem is with the tuning of the beat, a ticker event which is used for the integration. How do I have to think ? - My speed is in pixels/second, so it has to tick every second? (atTime [0, 1..]) - My app will run at 60fps, so it has to tick 60 times per second? (atTime [0, 1/60..]) The main program (I'm not using any graphics library, just getting keys ('u' for up, 'd' for down) and displaying the current paddle pos) makeKeyboardEvent = do hSetBuffering stdin NoBuffering hSetBuffering stdout NoBuffering (buttonPressedSink, event) <- makeEvent =<< makeClock forkIO . forever $ getChar >>= buttonPressedSink return event main = do keyEvt <- makeKeyboardEvent let dirEvt = fmap key2dir keyEvt paddle = newPaddle (stepper PaddleStill dirEvt) 50 adaptE $ fmap print $ snapshot_ (paddleCenter paddle) $ atTimes [0, 0.01..] key2dir k = maybe PaddleStill snd $ find (($ k) . fst) [((== 'u'), PaddleUp), ((== 'd'), PaddleDown)] When I launch the program, I hit 'u' or 'd' and I see the position increase/decrease. If the beat ticks every second, I see the position being increased by 50 every second, which is normal, but it's a too low refresh rate for an app that is to run at 60fps. But when I set my beat to tick every 60 times per second, the position is well updated, but I clearly see that the display dramatically slows down after a few seconds of execution. Too heavy rate for integrate? That's why I'm asking: how should I use integrate?

6 10

cabal-install
by Serguey Zefirov 19 May '10

19 May '10

Why there is no switch to turn off any use of proxy in cabal-install? Or to supply username/password pair in command line. I have a strange situation: wget works like charm ignoring proxy (I downloaded Cabal and cabal-install to investigate the problem using wget), Firefox works like charm igoring proxy, and only cabal cannot access anything, because it follows proper proxy protocol and our proxy requires username and password. And I don't know of proper way to supply them through http_proxy environment variable. The most annoying part is that it is too much work to download and build cabal using wget and scripts - I once tried it for some other utility and didn't succeed. PS And while I am on it, I'd like to suggest that cabal-install should read common options like credentials from environment or configuration file.

3 5

cabal: problem building ffi shared library and significance of __stginit
by Anthony LODI 19 May '10

19 May '10

Hello, I'm trying to build some haskell code as a .so/.dll so that it can ultimately be used by msvc. I have it working when I compile by hand (listed below) but I can't get the exact same thing built/linked with cabal. On linux everything builds fine, but when I try to link the resulting .so file, I get an error about a missing '__stginit_CInterface' reference. Indeed I couldn't find that name in any of the cabal-generated .dyn_o files. I checked the output of 'cabal build -v' and it seems to be executing about the same thing that I'm executing manually so I'm not sure what could be going wrong. On windows cabal won't even configure since '--enable-shared' seems to imply '-dynamic' (right?), and that's not currently supported. Also, when I remove the line 'hs_add_root(__stginit_CInterface);', and the corresponding forward declaration, the program runs fine! Does ghc no longer need this call or are my toy programs just being lucky sofar? Cheers, - Anthony LODI ========================================= (archlinux32 -- 2.6.32) > hsc2hs CInterface.hsc > ghc --make CInterface > cat dll_init.h #pragma once extern void __stginit_CInterface(void); extern void sp_init(int*, char***); extern void sp_exit(void); > cat dll_init.c #include "HsFFI.h" #include "dll_init.h" void sp_init(int* p_argc, char*** p_argv) { hs_init(p_argc, p_argv); hs_add_root(__stginit_CInterface); } void sp_exit(void) { hs_exit(); } > ghc -c dll_init.c > ghc -o libfoo.so -shared -dynamic CInterface.o CInterface_stub.o dll_init.o > ghc -c ctest.c > ghc -threaded -o ctest -L. -lfoo ctest.o -optl-Wl,-rpath,'$ORIGIN' > ./ctest ========================================= (winxp is similar except without '-threaded' or '-dynamic') ========================================= > cat test.cabal ... library build-depends: base >=4 && <5 exposed-modules: CInterface c-source: dll_init.c ghc_options: -threaded > cabal clean > cabal configure --enable-shared > cabal build

2 2

(no subject)
by R J 19 May '10

19 May '10

This is another proof-layout question, this time from Bird 1.4.7. We're asked to define the functions curry2 and uncurry2 for currying and uncurrying functions with two arguments. Simple enough: curry2 :: ((a, b) -> c) -> (a -> (b -> c))curry2 f x y = f (x, y) uncurry2 :: (a -> (b -> c)) -> ((a, b) -> c)uncurry2 f (x, y) = f x y The following two assertions are obviously true theorems, but how are the formal proofs laid out? 1. curry2 (uncurry2 f) x y = f x y 2. uncurry2 (curry 2 f) (x, y) = f (x, y) _________________________________________________________________ The New Busy is not the too busy. Combine all your e-mail accounts with Hotmail. http://www.windowslive.com/campaign/thenewbusy?tile=multiaccount&ocid=PID28…

2 1

Proof format
by R J 19 May '10

19 May '10

Is this how a rigorous Haskeller would lay out the proofs of the following theorems? This is Bird 1.4.6. (i) Theorem: (*) x = (* x) Proof: (*) x = {definition of partial application} \y -> x * y = {commutativity of "*"} \y -> y * x = {definition of "(* x)"} (* x) (ii) Theorem: (+) x = (x +) Proof: (+) x = {definition of partial application} \y -> x + y = {definition of "(x +)"} (x +) (iii) Theorem: (-) x /= (- x) Proof: (-) x = {definition of partial application} \y -> x - y /= {definition of prefix negation, which is not a section} (- x) _________________________________________________________________ Hotmail is redefining busy with tools for the New Busy. Get more from your inbox. http://www.windowslive.com/campaign/thenewbusy?ocid=PID28326::T:WLMTAGL:ON:…

2 1

Bug with [Double]
by Dmitry Olshansky 19 May '10

19 May '10

Hello all, It seems that I saw something like this in Cafe recevtly. But I am not sure... In GHC 6.12.1 (Platform 2010 on Windows Vista) I have Prelude> [1,1+2/3..10] [1.0,1.6666666666666665,2.333333333333333,2.9999999999999996,3.666666666666666,4.333333333333332,4.999999999999998,5.666666666666664,6.33333333333333,6.9999999999999964,7.6666666666666625,8.333333333333329,8.999999999999995,9.66666666666666,10.333333333333327] -- It is a bug! Prelude> [1,5/3..10] [1.0,1.6666666666666667,2.3333333333333335,3.0,3.6666666666666665,4.333333333333333,5.0,5.666666666666667,6.333333333333334,7.000000000000001,7.666666666666668,8.333333333333336,9.000000000000004,9.666666666666671] -- correct, but... Prelude> [1,5/3..4] [1.0,1.6666666666666667,2.3333333333333335,3.0,3.6666666666666665,4.333333333333333] -- ... wrong again Prelude> [1,1+2/3..10] :: [Float] [1.0,1.6666667,2.3333335,3.0000002,3.666667,4.333334,5.000001,5.666668,6.333335,7.000002,7.666669,8.333336,9.000003,9.66667] -- correct Prelude> [1,1+2/3..10] :: [Double] [1.0,1.6666666666666665,2.333333333333333,2.9999999999999996,3.666666666666666,4.333333333333332,4.999999999999998,5.666666666666664,6.33333333333333,6.9999999999999964,7.6666666666666625,8.333333333333329,8.999999999999995,9.66666666666666,10.333333333333327] -- wrong Any comments?

7 11

Numerical Analysis
by Pierre-Etienne Meunier 19 May '10

19 May '10

Hello Cafe, Being a complete beginner in the field of numerical analysis, but anyway needing it to solve "real problems", I wrote a few functions recently to solve systems of polynomial equations using the "projected polyhedron" method by Maekawa and Patrikakalis. This requires solving systems of linear equations precisely, thus the simple Gauss method was not enough, and I had to write also an algorithm for the "SVD decomposition". Upon discovering the algol / fortran specifications of these :-( , heavily published in important journals, I thought it would be nice to provide the world with fast reliable implementation of these numerical methods (i.e. not simply bindings to lapack). Moreover, writing numerical things in haskell is much more pleasant than I thought at first. Here are a few random thoughts on this : - The haskell 98 norm does not require enough about IEEE-754 compliance, thus C bindings are still needed to guess for instance the machine epsilons, or manipulating ULPs. Moreover, taking advantage of hardware rounding is not easy, even if the hardware is IEEE-compliant : calling a C function from haskell screws up the speed advantages of hardware rounding, for instance. Maybe the new LLVM backend will make this possible ? - The current Array library is definitely not adapted to production code. It makes debugging tricky, requires a heavy use of Debug.Trace to actually see what happens, and does not seem as fast as one could expect. It seems that each algebra library on hackage redefines part of it, but a unified version would be nice : a discussion within the haskell community seems to be needed... - A numerical analysis library should really take advantage of the parallelism in GHC, especially with the arrival of hardware such as fermi (anyway, I do not know how much haskell is compilable to fermi code). The love for loops and side-effects among this community is hard to understand, but that's more of a cultural problem. Finally, as stated by William Stein, the creator of SAGE, of course it would take thousands of man-years to rewrite these codes in python, but if a language like haskell, and a compiler do 90% of the work, how many man-years are left ? If anyone here has got the time, the team and the will to start such a project, I'd love to contribute ! Cheers, PE

7 13

Re: [Haskell-cafe] Bug with [Double]
by Dmitry Olshansky 19 May '10

19 May '10

But Prelude Data.List> [1,1+2/3..4] :: [Double] [1.0,1.6666666666666665,2.333333333333333,2.9999999999999996,3.666666666666666,4.333333333333332] Prelude Data.List> unfoldr (\n -> let n'=n+2/3 in if n' <= 4 then Just (n',n') else Nothing) 1 :: [Double] [1.6666666666666665,2.333333333333333,2.9999999999999996,3.666666666666666] Prelude Data.List> takeWhile (<=4) $ iterate (+2/3) 1 :: [Double] [1.0,1.6666666666666665,2.333333333333333,2.9999999999999996,3.666666666666666] How 'dodgy' it should be to produce different result? How [a,b..c] works in this case? > > > 2010/5/19 Ivan Lazar Miljenovic <ivan.miljenovic(a)gmail.com>: >> Dmitry Olshansky <olshanskydr(a)gmail.com> writes: >> >>> Hello all, >>> >>> It seems that I saw something like this in Cafe recevtly. But I am not sure... >>> In GHC 6.12.1 (Platform 2010 on Windows Vista) I have >>> >>> Prelude> [1,1+2/3..10] >>> [1.0,1.6666666666666665,2.333333333333333,2.9999999999999996,3.666666666666666,4.333333333333332,4.999999999999998,5.666666666666664,6.33333333333333,6.9999999999999964,7.6666666666666625,8.333333333333329,8.999999999999995,9.66666666666666,10.333333333333327] >>> >>> -- It is a bug! >> >> No it isn't, because of the dodgy Ord instance for Float and Double values. >> >> -- >> Ivan Lazar Miljenovic >> Ivan.Miljenovic(a)gmail.com >> IvanMiljenovic.wordpress.com >> >

1 0