January 2015 - Haskell-Cafe

Undefined symbols when trying to use SDL-mixer
by Peter Reiner Fels 15 Jan '15

15 Jan '15

I'm trying to use SDL-mixer v0.6.1. When I add a dependency to my cabal file and run cabal repl I get the following error: Loading package SDL-mixer-0.6.1 ... <command line>: can't load .so/.DLL for: /home/prf/haskell/link-problem/.cabal-sandbox/lib/x86_64-linux-ghc-7.8.3/SDL-mi xer-0.6.1/libHSSDL-mixer-0.6.1-ghc7.8.3.so (/home/prf/haskell/link-problem/.cab al-sandbox/lib/x86_64-linux-ghc-7.8.3/SDL-mixer-0.6.1/libHSSDL-mixer-0.6.1-ghc7 .8.3.so: undefined symbol: Mix_FreeChunk) I can see these undefined symbols when I run: nm -D SDL-mixer-0.6.1/libHSSDL-mixer-0.6.1-ghc7.8.3.so | grep Free U Mix_FreeChunk U Mix_FreeMusic Running ldd on this file and grepping for SDL gives libHSSDL-0.6.5-ghc7.8.3.so => /home/prf/haskell/link-problem/.cabal-sandbox/lib /x86_64-linux-ghc-7.8.3/SDL-0.6.5/libHSSDL-0.6.5-ghc7.8.3.so (0x00007fb20b30600 0) libSDL-1.2.so.0 => /usr/lib64/libSDL-1.2.so.0 (0x00007fb20a232000) It seems like a dependency to /usr/lib64/libSDL_mixer.so is missing. Line 1815 of SDL-mixer-0.6.1/configure contains the following definition of SDL_LIBS: SDL_LIBS="-L`$SDL_CONFIG --prefix`/lib -lSDL_mixer" This refers to the 32 bit lib directory, but I'm on a 64 bit system. I unpacked SDL-mixer and changed this line to refer to the 64 bit lib directory. Next I recreated my cabal sandbox from scratch and added the patched version of SDL-mixer as a dependency. When I run cabal repl again I get the same error as mentioned above. What am I doing wrong? How do I get SDL-mixer to build against /usr/lib64/libSDL_mixer.so? Thanks, Peter

1 0

uu-parsinglib - Greedy Parser
by Marco Vassena 15 Jan '15

15 Jan '15

In the uu-parsinglib the choice combinator (<|>) is symmetric and does not commit to any alternative. If the grammar being parsed is ambiguous, the corresponding parser will fail at run time on an ambiguous input. Consider for instance this html parser. pTag = pOpenTag <|> pCloseTag <|> pCommentTag <|> pContent pContent = Content <$> some (satisfy (/= '<')) pHtml = some pTag This parser will fail on "<a>123</a>", because this may be interpreted as: [ [Open a , Content "123", Close a], [Open a, Content "12", Content "3", Close a], [Open a, Content "1", Content "2", Content "3", Close a] ... ] In other parsing libraries such as parsec the choice operator <|> is greedy and commits to the first alternative that makes any progress, so that some is greedy and Content "123" would be parsed. The operator <<|> in uu-parsinglib has the same greedy behaviour. I would like to disambiguate the grammar so that the first result ([Open a, Content "123", Close a]) is selected (longest matching rule). I suppose that this may be done using <<|> to define a greedySome to be used in in pContent, however I am wondering whether there is another way to do this, without using such operator <<|>. Any help is appreciated. All the best, Marco

6 7

1/15/2015 3:46:00 PM
by Diony Rosa 14 Jan '15

14 Jan '15

http://www.1367.co.nz/basjaupc/pebrbhpjthzcrumhnityftkmrmxbjau.myfduxbsyxon… Diony Rosa 1/15/2015 3:46:00 PM

1 0

ParallelListComp strange behavior
by Michael Jones 14 Jan '15

14 Jan '15

I am trying to use ParallelListComp and have code that compiles but only produces one value, but I am not clear why. The code is something like: gen :: (a -> b) -> [MyType] gen f = [MyType (Just []) (Just c) (Just d) Nothing | c <- f | d <- [1,2,3,4] When the gen function is used with take or any other function, only one value is returned. Is there something here that can’t desugar properly to mzip that will still compile? Note I did not test this specific code, I am just writing it here to represent the different pieces of a larger piece of real code which is below. Mike ———— Real Code ———— data Config = Config { onOffControl::Maybe [DEVICE_ON_OFF_CONTROL_BITS] , seqUpPos::[Maybe ([DEVICE_SEQ_UP_POSITION_BITS], Word16)] , seqDownPos::[Maybe ([DEVICE_SEQ_DOWN_POSITION_BITS], Word16)] , tonTimers::[Maybe ([DEVICE_TON_TIMERS_BITS], Word16)] , toffTimers::[Maybe ([DEVICE_TOFF_TIMERS_BITS], Word16)] , vRange::Maybe [DEVICE_V_RANGE_BITS] , vThresh::[Maybe (Word16, Word16)] , rstbConfig::Maybe [DEVICE_RSTB_CONFIG_BITS] , faultResponse::Maybe [DEVICE_FAULT_RESPONSE_BITS] , breakpoint::Maybe ([DEVICE_BREAK_POINT_BITS], Word16) , writeProtect::Maybe ([DEVICE_WRITE_PROTECTION_BITS], Word16) , specialLot::Maybe Word16 , deviceId::Maybe Word16 , statusInfo::Maybe ([DEVICE_STATUS_INFORMATION_BITS]) , monitorStatus::Maybe ([DEVICE_MONITOR_STATUS_BITS]) , monitorStatusHistory::Maybe ([DEVICE_MONITOR_STATUS_HISTORY_BITS]) , monitorBackup::Maybe ([DEVICE_MONITOR_BACKUP_BITS]) , seqPosCount::Maybe ([DEVICE_SEQ_POSITION_COUNT_BITS], Word16) generateConfigs seqValues delayValues threshValues = return [Config (Just []) -- On off ([Just (suPol, suPos),Just (suPol, suPos),Just (suPol, suPos),Just (suPol, suPos),Just (suPol, suPos),Just (suPol, suPos)]) ([Just (sdPol, sdPos),Just (sdPol, sdPos),Just (sdPol, sdPos),Just (sdPol, sdPos),Just (sdPol, sdPos),Just (sdPol, sdPos)]) ([Just ([tonMax], tonDelay),Just ([tonMax], tonDelay),Just ([tonMax], tonDelay),Just ([tonMax], tonDelay),Just ([tonMax], tonDelay),Just ([tonMax], tonDelay)]) ([Just ([toffMax], toffDelay),Just ([toffMax], toffDelay),Just ([toffMax], toffDelay),Just ([toffMax], toffDelay),Just ([toffMax], toffDelay),Just ([toffMax], toffDelay)]) (Just [rng1, rng2, rng3, rng4, rng5, rng6]) ([Just (threshOv, threshUv),Just (threshOv, threshUv),Just (threshOv, threshUv),Just (threshOv, threshUv),Just (threshOv, threshUv),Just (threshOv, threshUv)]) (Just ([rst_en] ++ [rst_time])) (Just ([fault_number] ++ [fault_action] ++ [fault_delay] ++ fault_other ++ [fault_count])) (Just ([], 0)) (Just ([], 0)) (Just 0) Nothing Nothing Nothing Nothing Nothing Nothing | suPol <- [DEVICE_seq_up_position_bits] | suPos <- seqValues | sdPol <- [DEVICE_seq_down_position_bits] | sdPos <- seqValues | tonMax <- DEVICE_ton_timers_bits | tonDelay <- delayValues | toffMax <- DEVICE_toff_timers_bits | toffDelay <- delayValues | rng1 <- DEVICE_v_range_bits1 | rng2 <- DEVICE_v_range_bits2 | rng3 <- DEVICE_v_range_bits3 | rng4 <- DEVICE_v_range_bits4 | rng5 <- DEVICE_v_range_bits5 | rng6 <- DEVICE_v_range_bits6 | threshOv <- threshValues | threshUv <- threshValues | rst_time <- DEVICE_rstb_config_bits_time | rst_en <- DEVICE_rstb_config_bits_en | fault_number <- DEVICE_fault_response_bits_retry_number | fault_other <- [DEVICE_fault_response_bits_other] | fault_count <- DEVICE_fault_response_bits_count | fault_delay <- DEVICE_fault_response_bits_delay | fault_action <- DEVICE_fault_response_bits_action ] generateSeqValues :: IO [Word16] generateSeqValues = do vs <- replicateM 10 (randomRIO (0x0000, 0x03FF)) return vs generateDelayValues :: IO [Word16] generateDelayValues = do vs <- replicateM 10 (randomRIO (0x0000, 0x1FFF)) return vs generateThreshValues :: IO [Word16] generateThreshValues = do vs <- replicateM 10 (randomRIO (0x0000, 0x00FF)) return vs

2 4

Re: [Haskell-cafe] working my way through Data.Type.Equality...my head hurts...
by oleg＠okmij.org 14 Jan '15

14 Jan '15

Richard Eisenberg wrote: > `~` is simply equality on types. The constraint `a ~ b` means that `a` > is the same type as `b`. We could have spelled it `=` if that weren't > used elsewhere. Before `~` was added to GHC, it might have been > implemented like this: > class Equals a b | a -> b, b -> a > instance Equals x x > -- no other instances! Actually, `~' has been implemented, back in 2004, and the implementation quite a bit more involved than the above (although still taking a couple of lines). The `~' constrain was called TypeCast back then, and there is a page about it explaining the benefits and peculiarities of the implementation. http://okmij.org/ftp/Haskell/typecast.html Perhaps the many examples of TypeCast make it less mysterious.

1 0

Deploying Scotty App to AWS
by Stephen Portanova 13 Jan '15

13 Jan '15

I'm having trouble running my Scotty binary on amazon web services... I'm using Ubuntu 15.04, this gist (https://gist.github.com/yantonov/10083524) to install the haskell platform dependencies, and then I do a git pull to my repo and build my binary. Everything compiles, but when I do ./cabal-sandbox/bin/app or ./dist/build/app/app, nothing happens. No error messages or anything - the terminal just goes to the next line and hangs (on my Mac development machine the server starts up fine and prints a startup message). Any ideas or experience with deploying Scotty / Haskell apps would be really appreciated, thanks! - Stephen

5 7

Future of the boxes package--call for ideas
by David Feuer 13 Jan '15

13 Jan '15

I've just taken over maintainership of the boxes package, and will be making a maintenance release shortly (as soon as I figure out how and get added to the maintainers group). The package, however, currently suffers from a paucity of bug reports (no problem) and feature requests (not so great). To keep things lively, I need a bit of help from two groups of people. If you use the package but wish it could do something more for you, I want to know about it. If you considered using the package but rejected it because it couldn't quite handle your job, I want to know about that too. I'd prefer if people would open issues and pull requests at https://github.com/treeowl/boxes/issues but I will also accept requests by email. Thanks, David

1 0

working my way through Data.Type.Equality...my head hurts...
by Nicholls, Mark 13 Jan '15

13 Jan '15

Full code here… https://github.com/goldfirere/nyc-hug-oct2014/blob/master/Equality.hs Lets start at the beginning… Ø data a :~: b where Ø Refl :: a :~: a Looks reasonable….” a :~: b” is inhabited by 1 value…of type “a :~: a” Ø sym :: (a :~: b) -> (b :~: a) Ø sym Refl = Refl hmmm… (a :~: b) implies that a is b so (b :~: a) brilliant Ø -- | Transitivity of equality Ø trans :: (a :~: b) -> (b :~: c) -> (a :~: c) Ø trans Refl Refl = Refl Ø -- | Type-safe cast, using propositional equality Ø castWith :: (a :~: b) -> a -> b Ø castWith Refl x = x fine…. But…. Ø -- | Generalized form of type-safe cast using propositional equality Ø gcastWith :: (a :~: b) -> ((a ~ b) => r) -> r Ø gcastWith Refl x = x I don’t even understand the signature What does “~” mean…(it’s something that comes out in error messages when my types are all messed up)….and then there’s a “=>” going on…in the middle of a signature….I know “=>” in the context of “Num a => a -> a -> a” What would a value of type “((a ~ b) => r)” look like? CONFIDENTIALITY NOTICE This e-mail (and any attached files) is confidential and protected by copyright (and other intellectual property rights). If you are not the intended recipient please e-mail the sender and then delete the email and any attached files immediately. Any further use or dissemination is prohibited. While MTV Networks Europe has taken steps to ensure that this email and any attachments are virus free, it is your responsibility to ensure that this message and any attachments are virus free and do not affect your systems / data. Communicating by email is not 100% secure and carries risks such as delay, data corruption, non-delivery, wrongful interception and unauthorised amendment. If you communicate with us by e-mail, you acknowledge and assume these risks, and you agree to take appropriate measures to minimise these risks when e-mailing us. MTV Networks International, MTV Networks UK & Ireland, Greenhouse, Nickelodeon Viacom Consumer Products, VBSi, Viacom Brand Solutions International, Be Viacom, Viacom International Media Networks and VIMN and Comedy Central are all trading names of MTV Networks Europe. MTV Networks Europe is a partnership between MTV Networks Europe Inc. and Viacom Networks Europe Inc. Address for service in Great Britain is 17-29 Hawley Crescent, London, NW1 8TT.

5 13

--disable-{executable-}profiling cabal-install-1.22.0.0 with ghc-7.6.3
by Ozgur Akgun 13 Jan '15

13 Jan '15

A recent change in cabal-install's command line interface seems not to be compatible with ghc-7.6.3. cabal-install-1.22.0.0 doesn't like --disable-executable-profiling as a command line option any more. (I could only find this related to this change: https://github.com/haskell/cabal/pull/2286) $ cabal install --disable-executable-profiling cabal: unrecognized 'install' option `--disable-executable-profiling' I try --disable-profiling instead. $ cabal install transformers --disable-profiling -j3 Resolving dependencies... Notice: installing into a sandbox located at [snip] Configuring transformers-0.4.2.0... Failed to install transformers-0.4.2.0 Build log [snip]: unrecognized 'configure' option `--disable-profiling' cabal: Error: some packages failed to install: transformers-0.4.2.0 failed during the configure step. The exception was: ExitFailure 1 Using the following versions of cabal and ghc. $ cabal -V cabal-install version 1.22.0.0 using version 1.22.0.0 of the Cabal library $ ghc -V The Glorious Glasgow Haskell Compilation System, version 7.6.3 --disable-profiling works fine with ghc-7.8.3, this is only a problem with 7.6.3. PS: transformers is only a dependency of an executable I am trying to build. But this happens with all the other dependencies as well. -- Ozgur Akgun

3 14

FP Complete is hiring: Software engineer
by Mathieu Boespflug 13 Jan '15

13 Jan '15

Hi list, FP Complete is expanding yet again! We are looking to hire for several new engineers to join our Haskell development team, both to build great new core products and in partnership with our clients to apply Haskell at a large scale. Some of our recently announced core products include the Integrated Analysis Platform, Stackage and LTS Haskell, with much more to come. If you’d like to be part of our team and shape the future of Haskell, please send a resume or CV to jobs+dev(a)fpcomplete.com. Any existing work - either a running website or an open source codebase - which you can include as links be greatly appreciated as well. We will want you to start right away. Depending on your current jurisdiction, this will either be a full-time contractor position, or an employee position. This is a telecommute position: you can work from home or wherever you choose, with little or no travel. Location in North America preferred; but you will work with colleagues who are both on North American and European hours. Activities: * distribute existing and new code over large clusters, * code parallelization and performance tuning, * interface with foreign math and scientific libraries, * relentlessly refactor for composability, testability and clarity, * identify performance bottlenecks and debug known issues, * implementing unit tests, integration tests, acceptance tests, * write clear and concise documentation. Skills: * strong experience writing process driven application code in Haskell. * experience building reusable components/packages/libraries and demonstrated ability to write well structured and well documented code, * ability to evolve and refactor large code bases, * experience working under test driven methodologies, * Ability to interact with a distributed development team, and to communicate clearly on issues, in bug reports and emails. These further skills are a plus: * Bachelor’s or Master’s degree in computer science or mathematics, * experience developing products in a regulated environment (avionics/medical/finance). * experience building scalable server/Web applications, * (web) UI design and implementation experience, * an understanding of the implementation of GHC’s multithreaded runtime, * experience as an architect and/or a creator of technical specs.

1 0