December 2008 - Haskell-Cafe

Cabal Install & Links to Source from Haddock Docs
by R Hayes 23 Dec '08

23 Dec '08

Is there a way I can get Haddock Docs WITH links to source (local) from modules installed with "cabal install xxx"? Getting the docs themselves is pretty easy by changing either ~/.cabal/ config or using --enable-documentation. Automatically generating the source (colourised or not) and integrated links eludes me. -r

3 3

Re: understanding enumerator/iteratee
by oleg＠okmij.org 23 Dec '08

23 Dec '08

Artyom Shalkhakov wrote > I would say that it [iteratee] just tells us how to react to various forms of > input. :) This is much like the function you pass to foldr. Precisely. To sum up all elements of Data.Map, we do Map.fold (+) 0 mp to sum up all elements of a set we do Set.fold (+) 0 st ditto for any other foldable data structure. Clearly the function that sums the current element with the accumulator, (+), doesn't know or care from which collection the elements are coming from. The initial seed, 0, is again unaware of the collection. Iteratee is indeed the function that you pass to fold (combined with the seed for practical reasons). One may conceptually consider iteratee to be a pair of the function to feed to fold, and the initial seed (the accumulator in the above example). That achieves the separation of concerns: fold (aka, enumerator) has the intimate knowledge of the collection and how to get to the next element; iteratee knows what to do with the current element.

1 0

Takusen
by Günther Schmidt 22 Dec '08

22 Dec '08

Hi all, where can I find some sample code or other examples to familiarize me with Takusen? Does Takusen use safe ccall only, or does it also use ccall unsafe? Günther

2 1

A hell of a question
by Andrzej Jaworski 22 Dec '08

22 Dec '08

I just want to say Hello to let you know that there are some serious entities watching you besides monads and FBI:-) There has been a hell of a discussion recently about logos, languages and religion and I want to add to this. First let me disassociate Haskell from Taoism which to may taste has left us in an unhealthy climate. It suffices to say that Taoism is a school of clever trics and cute aphorisms but without the slightest attempt to explain or generalize let alone produce an abstract idea or a system. That is why its wisdom is non transferable in spite of majority of humans desending from it. Haskell on the contrary is a minority school that implements abstract ideas for problem solving in the most transferable way to date, so that other languages look into it for their share. But don't worry, thay will choke becouse it is them who practice Taoizm. Playing too many tricks will eventually trick them, even if some are powerful enough to brainwash dicent professors to preach interoperability or the like. Every viable complexity needs a single underlying concept to survive, including you and the universe. Microsoft and the like excluding;-) Haskell has all that: consistency, transparency and self-contained concept. However Haskell is also somehow asynchronic with the Bible as it is condemned to perpetual purity only to be saved from it via monads, which according to Spinoza "are arranged by God in a perfect order which ascends to God, the supreme monad!!!". Thus we can look at Haskell's purity as a kind of harmless attempt to play God by means of incapsulating the world only to ignor it. This could somehow put it in the same boat with the devil. They both prefer your brain to Turing machin:-) I would leave with the eternal question "what the hell Haskell is?" and with my Santa Close emphasising Haskell's purity and my simplicity. Here: http://haskell.org/sitewiki/images/7/79/WCF.Andrzej.Jaworski.gif Have fun, -Andrzej Jaworski

4 5

OpenAL
by brian 22 Dec '08

22 Dec '08

Does the example code at http://articles.bluishcoder.co.nz/Haskell/OpenAL work for anyone? I added some putStrLns to see the device and context, and get $ ./Main2 Device (ALCdevice 0x08a744c8) Context (ALCcontext 0x08aabda0) AL lib: alSource.c:2291: alcDestroyContext(): 1 Source(s) NOT deleted AL lib: alBuffer.c:1097: exit() 2 Buffer(s) NOT deleted and no sound. I had more complex code using OpenAL working some time ago, but the sound was crackly because my distribution used the original OpenAL implementation. Now that people are switching to OpenAL Soft, Haskell's OpenAL seems to be acting strangely. I mailed Sven in September, but didn't hear back.

1 0

"Rewrite thunk action" rule?
by Peter Todd 22 Dec '08

22 Dec '08

I have a program with this data structure: data Element = Element { elementOrigin :: V, elementSubs :: [Element] } and this important bit of code that operates on it: transform :: T -> Element -> Element transform t e = Element { elementOrigin = tmulv t (elementOrigin e), elementSubs = map (transform t) (elementSubs e) } Where V is a vertex and T is a matrix transformation. The following is true: transform a (transform b e) == transform (a * b) e I have read about rewrite rules, such would efficiently rewrite the obvious transformations. However, given that the program will be applying *many* transformations to very deep and wide Element trees, and applying those transformations at many levels, I don't see how the optimizer would be able to handle all cases, for instance a thunk getting left over from perhaps evaluation prior to some IO function. FWIW here's an example "tree-builder" I'm using to test with: mkElems 0 _ = Element { elementOrigin = V 0 0, elementSubs = [] } mkElems depth width = Element { elementOrigin = V 0 0, elementSubs = replicate width $ transform (rotation (pi / (fromIntegral width))) $ transform (translation $ V 1 1) $ mkElems (depth - 1) width } with depth = width = 5 If I could somehow arrange for the transform function to detect when it is being applied to a unevaluated thunk, and then modify the thunk in place, that would basically be the behavior I need. Any suggestions? -- http://petertodd.org 'peter'[:-1]@petertodd.org

3 4

Re: [Haskell-cafe] A hell of a question
by Andrzej Jaworski 22 Dec '08

22 Dec '08

Arnaud Bailly kindly exposed my mistake: I cited Leibniz Monadology attributing it to Spinoza. Call me idiot but it wasn't thoughtless mistake, I really mixed up Spinoza's concept of 'modes' with Leibniz's concept of 'monad'. But it is because of my laziness rather than foolishness (at least this time). My argument is that in his final writings Leibniz's 'monads' encompass reality as perceptions and emanate from God as thought emanates from the mind, which is exactly what Spinoza's 'modes' are about. So monads and modes being equal I prefer Spinoza as a patron for Haskell for his purity of cause-effect treatment. Arnold adds another Haskell-Spinoza nicety, I hope he won't mind my including his all letter. Thank you Arnold, it is encouraging to find guys like you. Philosophy is not my thing but I believe every man should tackle it as exercise. Hello Andrzej, Thanks a lot for your bit of rationalism among all these devotions :-) I think however that you are confusing Spinoza with Leibniz in the following assertion. Monads are a concept invented by the latter, together with the famous "the best possible world ever" mocked by Voltaire in "Candide". The comparison between Haskell and spinozism is rather interesting though, especially when one considers that Spinoza's "Ethic" is based on the idea that the ultimate goal of one self is to increase its power to live, an affect which it calls "Joy". Thinking of Haskell as a way to increase one's joy and one's power is a nice thought. Regards, Arnaud Bailly

1 0

Re: [Haskell] ANN: HLint 1.0
by Gwern Branwen 22 Dec '08

22 Dec '08

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 So has anyone figured out how to integrate hlint with ghci? I've tried several approaches. Using hlint as a preprocessor ( http://www.haskell.org/ghc/docs/latest/html/users_guide/options-phases.html… ) doesn't work because the pre-processor needs to take 3 args. Even if one writes a wrapper shell script along the lines of 'cp $2 $3 && hlint $2', this still won't work. It apparently breaks when you load a file which requires another file. I am not sure why. I then thought to redefine :load and :reload in ghci; this would work very well with Emacs and Yi, since they tell ghci to :load and :reload buffers. Turns out, you're not allowed to overwrite :load and :reload. You can do :def reload whatever to your heart's content, but the new version never takes. (Not even if defined in .ghci.) Alright. A :hlint would probably be enough. I can always edit Yi/Emacs to follow a :reload/:load with a :hlint, after all. And aren't there shell command primitives already? So I set out, and I realize: I can't simply settle for some version of :! "hlint .", because the working directory rarely changes - if one opens ghci in ~/, and loads a file elsewhere, :pwd reveals one to still be in ~/. And a 'hlint .' could take a very long time indeed in ~/. Well, alright. I just need the file name of the module I'm working with. Surely there is an easy way to do that. And indeed, :show modules gives me what I'm looking for: *Recorder Control.Monad Data.Char Data.List> :show modules Util ( Util.hs, interpreted ) Recorder ( Recorder.hs, interpreted ) Game ( Game.hs, interpreted ) Monadius ( Monadius.hs, interpreted ) Demo ( Demo.hs, interpreted ) Well, sort of. Ok, we can parse that. Let's assume a variable x holds the output of :show modules as a String. We call lines on it, then map words on it, do a !! 2 on it, and we get ["Util.hs,", "Recorder.hs,", "Game.hs,", "Monadius.hs,", "Demo.hs,"]. Chuck in a map (filter (\= ',')), and we get a good list. We can turn the list into a string suitable for hlint with a quick unwords. So our long sought after command becomes ':def hoogle (\_ -> return $ ":! " ++ (unwords $ map (filter (\= ',')) $ (map words $ lines x) !! 2))'. But wait, how do we get 'x'? How do we call :show modules inside a Haskell expression? I have carefully looked over http://haskell.org/haskellwiki/GHC/GHCi#Using_GHCi and http://www.haskell.org/ghc/docs/latest/html/users_guide/ghci-commands.html and my conclusion is that you can't. You can't do a let x = :show modules, there is no function which will take ":show modules", and so on. :functions can accept Haskell output, but it's a one-way barrier. It's no good writing Haskell functions which need information from the :functions. Which all leaves me where I started - no ghci integration with hlint. Any ideas? - -- gwern -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) iEYEAREKAAYFAklNNSIACgkQvpDo5Pfl1oJwWgCfZJ7nwkHfsZymIRass0ewmhFE en4AnjmQam3V8Go6pGoLTRsp5zZEzAms =nmJm -----END PGP SIGNATURE-----

2 1

A hell of a question
by Andrzej Jaworski 22 Dec '08

22 Dec '08

Hi Luke, When neurosurgeons split the brain into left and right hemisphere cutting it along corpus callosum the patient will talk to you with his left half and right half independently - each time unaware what his other half was talking about a moment earlier. I believe Zen emulates such split on a microscopic scale. We all go thru similar state when we conceive an original idea - if we do not write this down or formulate it immediately the discovery will loose its punch and sense of depth. The best documented and most prominent example is Hegel's discovery of dialectical logic - after extensive writing about that at some point he honestly admits loosing the original concept. But if not for his western oververbosity we might not today use the term 'naive' set theory. Every serious mathematician touching foundations is today perfectly aware of the price we pay for formulating things. But I am unaware of any virtue of trying to clap with one hand;-) Regards, -Andrzej

1 0

haskell for windows mobile 6 on Intel PXA270 processor
by Constantine Potapov 22 Dec '08

22 Dec '08

Hi All. I'm looking for haskell interpreter/compiler for windows mobile 6 (I have an Asus A696 PDA with intel PXA270 processor). Somebody head about any haskell implementation for this platform? -- With best regards, Constantine Potapov.

1 0