March 2013 - Haskell-Cafe

OdHac: second call for participation
by Roman Cheplyaka 06 Mar '13

06 Mar '13

------------------------------------------------------------------------ OdHac International Haskell Hackathon May 3-5, 2013 Odessa, Ukraine http://haskell.org/haskellwiki/OdHac Co-organized with Provectus IT ------------------------------------------------------------------------ The international Haskell hackathon in Odessa is just two months away! To attend, please register at http://bit.ly/OdHacReg. ## Projects A big idea of this Hackathon is to concentrate the effort around a few important and established projects. Although participants are free to work on their own projects, it's more fun to work together! So far we've got two great projects: Hakyll and Fay http://www.haskell.org/haskellwiki/OdHac/Projects#Hakyll http://www.haskell.org/haskellwiki/OdHac/Projects#Fay Feel free to add new projects to the wiki! One project we'd love to see represented, because of its enormous value to the community, is Cabal. So far several core Cabal hackers have expressed willingness to offer help via IRC, but it would be great to have a Cabal expert on site. If you can coordinate the effort around Cabal and answer questions, and you're planning to come — please get in touch with me. ## Accommodation The Hackathon dates overlap with a series of holidays in Ukraine and neighbour countries, and Odessa is a popular tourist destination. So book early! We offer an accommodation program to make it simple for you: http://www.haskell.org/haskellwiki/OdHac/Accommodation

1 0

Library API design: functional objects VS type classes
by Rob Stewart 05 Mar '13

05 Mar '13

Hi, I have a question about API design for Haskell libraries. It is a simple one: functional object data structures encapsulating mutable state VS type classes encapsulating mutable state Here is a simple example. I present an API: using a type class `FooC`, and aso as a data structure `FooT`. Both are stateful, in the form of an MVar holding an Integer, with an operation `incrFoo` to increment this value by one, and another `readFoo` to read the Integer value. ----- import Control.Concurrent -- API approach 1: Using type classes class FooC a where mkFooC :: IO a readFooC :: a -> IO Int incrFooC :: a -> IO () newtype Bar = Bar (MVar Int) instance FooC Bar where mkFooC = newMVar 0 >>= \i -> return $ Bar i readFooC (Bar mv) = readMVar mv incrFooC (Bar mv) = modifyMVar_ mv $ \i -> return (i+1) -- API approach 2: Using direct field records data FooT a = FooT { readFooT :: IO a , incrFooT :: IO () } mkBar :: IO (FooT Int) mkBar = do mv <- newMVar 0 return FooT { readFooT = readMVar mv , incrFooT = modifyMVar_ mv $ \i -> return (i+1) } -- Tests the type class API testTypeClass :: IO () testTypeClass = do bar <- mkBar incrFooT bar incrFooT bar i <- readFooT bar print i -- prints 2 -- Tests the direct data structure API testDataStruct :: IO () testDataStruct = do bar <- (mkFooC :: IO Bar) incrFooC bar incrFooC bar i <- readFooC bar print i -- prints 2 ---- With that, I now ask: which is more common? Which is the better API design for a library? The APIs are almost identical. Under what conditions is the type classes preferred over the "mutable object" style data structure? There are two related resources that provides additional context here, that favour the functional objects approach: - Section 3.4 "Mutable Objects" in "Haskell's Overlooked Object System" http://goo.gl/gnZXL - A similar question (data structures vs type classes) in "Haskell Antipattern: Existential Typeclass" http://lukepalmer.wordpress.com/2010/01/24/haskell-antipattern-existential-… Thanks! -- Rob

4 3

"translating" recursively defined sequence
by Christopher Howard 05 Mar '13

05 Mar '13

Hi. My Haskell is (sadly) getting a bit rusty. I was wondering what would be the most straightforward and easily followed "procedure" for translating a recursively defined sequence into a Haskell function. For example, this one from a homework assignment. quote: -------- a_1 = 10 a_(k+1) = (1/5) * (a_k)**2 -------- (The underscore is meant to represent subscripting what follows it.) -- frigidcode.com

4 4

Concurrency performance problem
by Łukasz Dąbek 05 Mar '13

05 Mar '13

Hello Cafe! I have a problem with following code: http://hpaste.org/83460. It is a simple Monte Carlo integration. The problem is that when I run my program with +RTS -N1 I get: Multi 693204.039020917 8.620632s Single 693204.039020917 8.574839s End And with +RTS -N4 (I have four CPU cores): Multi 693204.0390209169 11.877143s Single 693204.039020917 11.399888s End I have two questions: 1) Why performance decreases when I add more cores for my program? 2) Why performance of single threaded integration also changes with number of cores? Thanks for all answers, Łukasz Dąbek.

6 11

Haskell + RankNTypes + (forall p. p Char -> p Bool) sound?
by Shachaf Ben-Kiki 05 Mar '13

05 Mar '13

I was trying to figure out a way to write absurd :: (forall p. p Char -> p Bool) -> Void using only rank-n types. Someone suggested that Haskell with RankNTypes and a magic primitive of type (forall p. p Char -> p Bool) might be sound (disregarding the normal ways to get ⊥, of course). Is that true? Given either TypeFamilies or GADTs, you can write absurd. But it doesn't seem like you can write it with just RankNTypes. (This is related to GeneralizedNewtypeDeriving, which is more or less a version of that magic primitive.) This seems like something that GADTs (/TypeFamilies) give you over Leibniz equality: You can write data Foo a where FooA :: Foo Char FooB :: Void -> Foo Bool foo :: Foo Bool -> Void foo (FooB x) = x Without any warnings. On the other hand data Bar a = BarA (Is a Char) | BarB (Is a Bool) Void bar :: Bar Bool -> Void bar (BarB _ x) = x bar (BarA w) = -- ??? Doesn't seem possible. If it's indeed impossible, what's the minimal extension you would need to add on top of RankNTypes to make it work? GADTs seems way too big. Shachaf

2 1

What pattern is this (Something.T -> IO a) in Sound.ALSA.Sequencer
by Martin Drautzburg 05 Mar '13

05 Mar '13

Hello all, this was previously posted on Haskell Beginners, but only partially answered. In Sound.ALSA.Sequencer, there are a number of functions which together set up a midi environement (client, port, queue). They all have a type, where the last argument has a type like: (something.T -> IO a) i.e. *Main> :t SndSeq.withDefault SndSeq.withDefault :: SndSeq.OpenMode mode => SndSeq.BlockMode -> (SndSeq.T mode -> IO a) -> IO a *Main> :t Port.withSimple Port.withSimple :: SndSeq.T mode -> String -> Port.Cap -> Port.Type -> (Port.T -> IO a) -> IO a *Main> :t Queue.with Queue.with :: SndSeq.T mode -> (Queue.T -> IO a) -> IO a There is example code, where a full setup is created by a number of nested "do" blocks. The repeating pattern there is: something1 $ \x -> do something2 $ \y -> do something3 $ \z -> do What is this all about? I particularly would like to understand, when this parttern is needed and what determines the the number of nested "do" blocks. -- Martin

3 4

simple parsec question
by Immanuel Normann 05 Mar '13

05 Mar '13

Hi, I am trying to parse a semi structured text with parsec that basically should identify sections. Each section starts with a headline and has an unstructured content - that's all. For instance, consider the following example text (inside the dashed lines): --------------------------- top 1: some text ... bla top 2: more text ... bla bla --------------------------- This should be parsed into a structure like this: [Section (Top 1) (Content "some text ... bla"), Section (Top 1) (Content "more text ... bla")] Say, I have a parser "headline", but the content after a headline could be anything that is different from what "headline" parses. How could the "section" parser making use of "headline" look like? My idea would be to use the "manyTill" combinator, but I don"t find an easy solution. Many thanks for any hint Immanuel

4 7

Haskell syntax/indentation for vim
by Tristan Ravitch 05 Mar '13

05 Mar '13

Cafe, I've recently been playing with vim and wasn't quite satisfied with the existing syntax highlighting and indentation, so I thought I'd try my hand at a new Haskell mode: https://github.com/travitch/hasksyn It is minimal in that it doesn't provide support for running external commands over code or anything fancy. It just does syntax highlighting and reasonably-smart indentation. There is no support for literate Haskell since supporting both with one mode is very tricky. It might be useful to some people. Comments, bug reports, and suggestions welcome.

5 7

LambdaJam 2013
by Tony Morris 04 Mar '13

04 Mar '13

Hey haskell guys, LambdaJam 2013 is a functional programming conference to be held in Brisbane in early May. Call for submissions ends this Friday 08 March. If you are planning to make a submission, please make sure you do soon! If you are stuck somehow, I'd love to be able to help you get to the goal. So feel free to send me an email or I am on IRC (dibblego), lurking around the #haskell channel or privmsg if you like. Personally, I would love see more haskell submissions :) LambdaJam2013 call for papers: http://www.yowconference.com.au/lambdajam/Call.html -- Tony Morris http://tmorris.net/

1 0

Re: [Haskell-cafe] Trouble installing and using Chart/cairo on windows 7
by Brent Yorgey 04 Mar '13

04 Mar '13

On Mon, Mar 04, 2013 at 02:30:55PM -0800, briand(a)aracnet.com wrote: > On Mon, 4 Mar 2013 17:27:29 -0500 > Brent Yorgey <byorgey(a)seas.upenn.edu> wrote: > > > On Sun, Mar 03, 2013 at 05:38:02PM -0800, briand(a)aracnet.com wrote: > > > On Sun, 3 Mar 2013 19:58:37 -0500 > > > Brent Yorgey <byorgey(a)seas.upenn.edu> wrote: > > > > > > > > > > > Good access to fonts and font metrics is the kicker. Otherwise I'd > > > > say to switch to using diagrams as a backend, hence getting a whole > > > > bunch of actual backends for free. I would love to see development of > > > > some good Haskell font packages -- maybe it would even make a good > > > > GSoC project? Unfortunately I don't know enough about it to even know > > > > what would be involved, or how much work it would be. > > > > > > > > > > I assume that to use diagram the font package would have to be a vector font system, or could bit-mapped fonts be used ? > > > > A vector font system would be ideal, as then you could go crazy > > turning glyphs into paths and doing whatever the heck you want with > > them using the diagrams framework. However, bit-mapped fonts could be > > used too (as long as there is a backend to support them), it would > > just be less useful. > > > > yes - seems to me that vector fonts are definitely the way to go. > > it seems like this problem has to already have a solution. maybe all that's needed is a font importer which understands some, already established, vector font representation. > Well, there is the SVGFonts package: http://hackage.haskell.org/package/SVGFonts I don't have a good sense of how widely supported this representation is, or how easy it is to convert between it and other font representations. -Brent

1 0