November 2012 - Haskell-Cafe

Re: [Haskell-cafe] List all multiply/add combinations
by Rune Harder Bak 18 Nov '12

18 Nov '12

On Sun, Nov 18, 2012 at 2:04 PM, Stefan Klinger <all-lista(a)stefan-klinger.de> wrote: > Sounds like you would want to enumerate all possible *abstract* syntax > trees, these implicitly have exactly the necessary parentheses. I'd do > this recursively, splitting the sequence of numbers in two at all > possible places, and then combine the corresponding results with all > possible operators. That was my second idea, but just doing it naively resulted in many equivalent calculations, so I thought there might be a better way to view the problem. But as Artyom showed On Sat, Nov 17, 2012 at 11:37 PM, Artyom Kazak <artyom.kazak(a)gmail.com> wrote: > Indentation messed up… I have pasted the code here: http://hpaste.org/77864 enumerating the abstract syntax tree is actually (or at least could be) the way to go! Thanks a lot! What I need is a little bit different, but now I feel I'm on the right track! This is my first question to the café and it makes me a lot more certain using Haskell for production, when you can get this kind of quick and thorough help when in doubt. Great community! -Rune

4 3

Re: [Haskell-cafe] Automation of external library installation in haskell package.
by Maksymilian Owsianny 18 Nov '12

18 Nov '12

Hi Johan, Originally, that was my intention, to have the user install the library himself, and I have specified the extra-libraries field in the .cabal file. However, the problem with hackage not being able to build the package, and therefore generate documentation, remains. So conversely my question would be if there is any way to get around this? Best regards, Max. On Sun, Nov 18, 2012 at 8:21 PM, Johan Tibell <johan.tibell(a)gmail.com>wrote: > Hi Maksymilian, > > Generally I don't recommend that you bundle C libraries with your > Cabal package. On Linux users should get C libraries through their > distribution's package manager. The people who package Cabal packages > for Debian (as Debian packages) can then have the Debian package they > create depend on the C library's package, making sure it gets > installed automatically. > > You can list the library name in an extra-libraries field in your > .cabal file to communicate to users that they need this library > installed. There's a little bit of documentation on the > extra-libraries field in the Cabal User Guide > (http://www.haskell.org/cabal/users-guide/developing-packages.html#library > ). > > Cheers, > Johan >

3 2

Automation of external library installation in haskell package.
by Maksymilian Owsianny 18 Nov '12

18 Nov '12

Hello Haskell-Cafe, I have created a package that requires an external library to be present on the machine. Because of that, obviously, hackage fails to build it, and subsequently fails to generate documentation. So my question is, how should I go about configuring Cabal/Setup.hs to automatize the process to allow hackage to build it properly and also simplify the package installation for the user. The packages that I'm talking about are bindings to Awesomium<http://awesomium.com/> : awesomium-raw <http://hackage.haskell.org/package/awesomium-raw> and subsequently, the higher level ones that depend on it awesomium <http://hackage.haskell.org/package/awesomium> awesomium-glut <http://hackage.haskell.org/package/awesomium-glut> Specifically, the installation process for linux is described as follows: To install Awesomium, you'll need to add the shared library to your > system's library search path. On Ubuntu, you can use the following > commands: > > cd awesomium_v1.6.5_sdk_linux32 > sudo mkdir /usr/lib/awesomium-1.6.5 > sudo cp -r build/bin/release/* /usr/lib/awesomium-1.6.5 > sudo ldconfig /usr/lib/awesomium-1.6.5 > > So if anyone could give me some hints, or point me to package that have a similar, external-library-copying-and-registering installation process i would be grateful.

1 0

Parsing different types, same typeclass
by José Lopes 18 Nov '12

18 Nov '12

Hello everyone, I was wondering if you could help me! So, I have a typeclass "Action" which defines method "run": class Action a where run :: a -> Int and two data types that are instances of this typeclass: data A = A Int deriving (Read, Show) instance Action A where run (A n) = n data B = B Int deriving (Read, Show) instance Action B where run (B n) = n Now, I want to parse either "A" or "B" from a String. I was thinking about something like this... parseAction :: (Action a, Read a) => String -> a parseAction str | "(A " `isPrefixOf` str = (read :: String -> A) str | "(B " `isPrefixOf` str = (read :: String -> B) str The problem is that when calling "parseAction" I get "ambiguous type constraints". How to implement a parse function for two distinct types that share the same typeclass "Action". Because after calling "parseAction" I don't whether "A" or "B" was returned: I only care that they are "Action" instances so I can call "run". Best regards, José

5 7

A small step towards solving cabal hell.
by timothyhobbs＠seznam.cz 18 Nov '12

18 Nov '12

I've read a lot of negative regarding the problems with cabal and hackage. I've written quite a few of them myself. I want to propose a simple change in philosophy of packages. Haskell has inherited a philosophy from the imperative world, that there are two types of packages: Libraries and applications. Libraries are big collections of modules. Applications are big collections of modules. There is a difference from the perspective of the build system. Applications are big collections of modules that belong together and mutually rely upon each other for the application to work. Such that if one module is missing, the application cannot build and thus cannot do useful tasks. If an application doesn't build, then the maintainer has to go and fix that problem. This isn't perfect, I do not solve this problem in this email. Libraries don't always have this property. For example, XMonad.Layout.Column (http://hackage.haskell.org/packages/archive/xmonad-contrib/0.10/doc/html/XM…) has no mutual dependency on XMonad.Layout.Circle (http://hackage.haskell.org/packages/archive/xmonad-contrib/0.10/doc/html/XM…) . This is a feature of libraries that can and should be taken advantage of wherever possible :) We can and should, break up libraries into "books". I hold this opinion, because libraries are hard to maintain. If xmonad-contrib refuses to build, I open up it's source code, and see some hundred modules. I cannot, as a non-xmonad developer, imagine myself fixing such a large library. But if it was just one small module from that library I wanted, and it refused to build, and I opened it up, and there were just 3 files there. I wouldn't feel so overwhelmed. I would fix the problem myself. What I'm trying to say, is that "books" (small packages containing 3 modules or less) are so easy to maintain, that they really need no maintainer at all. Any idiot can fix one. But libraries, with their hundreds of modules, and seemingly endless dependencies, Can be much harder to maintain, and require a knowledgeable maintainer. So lets stop publishing libraries, and maintaining libraries, and listening to people crying when libraries don't build, and lets start writing books, and publishing really small packages, that are so simple anyone can fix them when they break even when they have no in depth knowledge of the package. Lets make a new guideline for "libraries" that if it is possible to split a library into two or more, non-mutually dependent parts, than we *should* split them such. Timothy PS, I noticed that yesod is already a collection of "books", but it appears to me, that these packages are STILL mutually dependent :(

3 6

I killed performance of my code with Eval and Strategies
by Janek S. 17 Nov '12

17 Nov '12

Dear Haskellers, I am reading Simon Marlow's tutorial on parallelism and I have problems with correctly using Eval monad and Strategies. I *thought* I understand them but after writing some code it turns out that obviously I don't because parallelized code is about 20 times slower. Here's a short example (code + criterion benchmarks): {-# LANGUAGE BangPatterns #-} module Main where import Control.Parallel.Strategies import Criterion.Main main :: IO () main = defaultMain [ bench "Seq" $ nf calculateSeq xs , bench "Par" $ nf calculatePar xs ] where xs = [1..16384] calculateSeq :: [Double] -> [Double] calculateSeq [] = [] calculateSeq (x:xs) = (sin . sqrt $ x) : xs calculatePar :: [Double] -> [Double] calculatePar xss = runEval $ go xss where go :: Strategy [Double] go [] = return [] go xs = do lsh <- (rpar `dot` rdeepseq) $ calculateSeq as lst <- go bs return (lsh ++ lst) where !(as, bs) = splitAt 8192 xs Compiling and running with: ghc -O2 -Wall -threaded -rtsopts -fforce-recomp -eventlog evalleak.hs ./evalleak -oreport.html -g +RTS -N2 -ls -s I get: benchmarking Seq mean: 100.5990 us, lb 100.1937 us, ub 101.1521 us, ci 0.950 std dev: 2.395003 us, lb 1.860923 us, ub 3.169562 us, ci 0.950 benchmarking Par mean: 2.233127 ms, lb 2.169669 ms, ub 2.296155 ms, ci 0.950 std dev: 323.5201 us, lb 310.2844 us, ub 344.8252 us, ci 0.950 That's a hopeless result. Looking at the spark allocation everything looks fine: SPARKS: 202 (202 converted, 0 overflowed, 0 dud, 0 GC'd, 0 fizzled) But analyzing eventlog with ThreadScope I see that parallel function spends most of the time doing garbage collection, which suggests that I have a memory leak somewhere. I suspected that problem might be caused by appending two lists together in the parallel implementation, but replacing this with difference lists doesn't help. Changing granularity (e.g. splitAt 512) also brings no improvement. Can anyone point me to what am I doing wrong? Janek PS. This is of course not a real world code - I know that I'd be better of using unboxed data structures for doing computations on Doubles.

3 2

How to determine correct dependency versions for a library?
by Janek S. 17 Nov '12

17 Nov '12

Recently I started developing a Haskell library and I have a question about package dependencies. Right now when I need my project to depend on some other package I only specify the package name in cabal file and don't bother with providing the package version. This works because I am the only user of my library but I am aware that if the library were to be released on Hackage I would have to supply version numbers in the dependencies. The question is how to determine proper version numbers? I can be conservative and assume that version of libraries in my system are the minimum required ones. This is of course not a good solution, because my library might work with earlier versions but I don't know a way to check that. What is the best way to determine a minimal version of a package required by my library? I also don't see any sane way of determining maximum allowed versions for the dependencies, but looking at other packages I see that this is mostly ignored and package maintainers only supply lower versions. Is this correct approach? Janek

16 36

List all multiply/add combinations
by Rune Harder Bak 17 Nov '12

17 Nov '12

Given a list of numbers of fixed length I need to list all possible values (and the associated computation) you get by inserting +,-,*,/ between the numbers, and also set parentheses where you please. It shouldn't list computations with unnecessary parentheses. Example list of length 3 [a,b,c] and only with + and *: a*b+c, a*(b+c),a*b*c,a+b*c,(a+b)*c,a+b+c What would be a good way to do this, and a good representation in Haskell? Best, Rune

3 6

Call for Participation: Programming Languages Mentoring Workshop - a POPL workshop.
by Gareth Smith 17 Nov '12

17 Nov '12

Apologies for any duplicates: CALL FOR PARTICIPATION SIGPLAN Programming Languages Mentoring Workshop, Rome Tuesday January 22, 2013 Co-located with POPL 2013 PLMW web page: http://www.doc.ic.ac.uk/~gds/PLMW/index.html After the resounding success of the first Programming Languages Mentoring Workshop at POPL 2012, we proudly announce the 2nd SIGPLAN Programming Languages Mentoring Workshop (PLMW), co-located with POPL 2013 and organised by Nate Foster, Philippa Gardner, Alan Schmitt, Gareth Smith, Peter Thieman and Tobias Wrigstad. The purpose of this mentoring workshop is to encourage graduate students and senior undergraduate students to pursue careers in programming language research. This workshop will provide technical sessions on cutting-edge research in programming languages, and mentoring sessions on how to prepare for a research career. We will bring together leaders in programming language research from academia and industry to give talks on their research areas. The workshop will engage students in a process of imagining how they might contribute to our research community. We especially encourage women and underrepresented minority students to attend PLMW. Since PLMW will be in Rome this year, we particularly look forward to seeing Eastern European students at the workshop. This workshop is part of the activities surrounding POPL, the Symposium on Principles of Programming Languages, and takes place the day before the main conference. One goal of the workshop is to make the POPL conference more accessible to newcomers. We hope that participants will stay through the entire conference, and will also attend the POPL tutorials on Monday 21st January which are free to PLMW registered attendees. Through the generous donation of our sponsors, we are able to provide scholarships to fund student participation. These scholarships will cover reasonable expenses (airfare, hotel and registration fees) for attendance at both the workshop and the POPL conference. Students attending this year will get one year free student membership of SIGPLAN The workshop registration is open to all. Students with alternative sources of funding are welcome. APPLICATION for PLMW scholarship: The scholarship application can be accessed from the workshop web site (http://www.doc.ic.ac.uk/~gds/PLMW/index.html) The deadline for full consideration of funding is 9th December, 2012. Selected participants will be notified from Friday 14th December, and will need to register for the workshop by December 24th. SPONSORS: Imperial College London Jane Street Monoidics NSF Resource Reasoning SIGPLAN vmware

1 0

Custom Enum instance with non-consecutive indices
by Nicolas Trangez 17 Nov '12

17 Nov '12

All, I've been working on a server implementation of an existing networking protocol. The protocol uses "magic constants" in some places (e.g. to tag message types), as well as bitfields, or a combination of both packed in a single value. I created data types for both the identifiers as well as the bitfield masks, e.g. > import Data.Bits > data MessageType = Message1 > | Message2 > | Message5 > data MessageFlag = Flag1 > | Flag2 Since I need to be able to get a numeric representation of them, I thought making a custom Enum instance would make sense: > instance Enum MessageType where > fromEnum a = case a of > Message1 -> 1 > Message2 -> 2 > Message5 -> 5 > toEnum n > | n == 1 = Message1 > | n == 2 = Message2 > | n == 5 = Message5 > instance Enum MessageFlag where > fromEnum a = case a of > Flag1 -> 1 `shiftL` 0 > Flag2 -> 1 `shiftL` 1 > toEnum n > | n == 1 `shiftL` 0 = Flag1 > | n == 1 `shiftL` 1 = Flag2 This is not a complete definition (not to mention non-exhaustive pattern matches), so I was wondering what the best practices are to extend this so these instances are 'correct'? I've been trying several approaches, but all of them failed on code like > getFlags :: Int -> [MessageFlag] > getFlags i = filter (\v -> (i .&. fromEnum v) /= 0) [Flag1 ..] Unless I hard-code all options in the last list, of course, but this is obviously not the intention. Any help or pointers (including "This is not how Enum is intended to be used") would be appreciated! Thanks, Nicolas

3 5