- Haskell-Cafe - Haskell.org

matrix vector product
by kolli kolli 12 Sep '11

12 Sep '11

Hi, can anyone help me with the matrix vector product ....

2 1

Parsec: non-greedy 'between'
by Scott Lawrence 11 Sep '11

11 Sep '11

Hey all, Trying to match C-style comments, I have: between (string "/*") (string "*/") $ many anyChar Which doesn't work, because it is equivalent (ignoring returned values) to do {string "/*"; many anyChar; string "*/"} If the termination criterion was a single character, then I could use noneOf or (satisfy . not), but that doesn't help here. So... what am I missing? Thanks in advance. -- Scott Lawrence

3 3

Functional Programming eXchange 2012: Call for abstracts
by Robert Pickering 11 Sep '11

11 Sep '11

Hello all, I’m sending this email to several mailing lists so my apologies if you see this twice. Functional Programming eXchange is a developer conference that focuses on functional programming. The 2012 edition will take place on Friday March 16th March 2012, at the Skills Matter eXchange, in London. Typically talks are explore concepts that often crop up in functional programming such as parallelism or domain specific languages, look at frameworks built in functional languages or are cases studies of using functional programming in industry. Previous speakers have included: Simon Peyton Jones, Miles Sabin, David Pollak, Adam Granicz, Antonio Cisternino, Tomas Petricek, Jonas Boner, and Viktor Klang. Videos of the talks at the previous events can be found here: http://skillsmatter.com/event/scala/functionalpx-2011 http://skillsmatter.com/event/cloud-grid/functional-programming-exchange-20… If you are interested at speaking at this conference please send a talk abstract to talks_functionalpx(a)skillsmatter.com. I’m aiming to make the decision about which talks have been chosen around 14th October so if you submit an abstract expect to hear from me around this time. Kind regards, Robert

1 0

ANN: Netwire 1.2.4
by Ertugrul Soeylemez 11 Sep '11

11 Sep '11

Hello there, version 1.2.4 of netwire is out. Major changes: * Changed the semantics of time. Previously if a wire was not run (because of an earlier inhibiting wire), then its local time was suspended. Example: proc _ -> do t1 <- time -< () fps1 <- avgFps 1000 -< () require_ <<< wackelkontakt -< () t2 <- time -< () fps2 <- avgFps 1000 -< () identity -< printf "Time: %8.2f %8.2f\nFPS : %8.2f %8.2f" t1 t2 fps1 fps2 In the previous version t2 would be about half of t1, because the {require_ <<< wackelkontakt} wire inhibits for about half of all instances. Further fps1 and fps2 would be about equal. In other words, time was local, while FPS were global. This is not very useful in practical applications. With the new semantics t1 and t2 will be equal. Instead the framerate will drop after the inhibiting wire, such that fps2 will be about half of fps1. So now time is global and FPS are local. This is useful e.g. in a stateless network application (like a web app) to ban users, who are trying logins too fast. The following wire allows one login trial per second in average over the last 10 trials (the faster you try the longer you have to wait): login :: Wire MyApp IpAddress User login = context $ proc ipAddr -> do fps <- avgFps 10 -< () require_ -< fps < 1 {- login procedure -} Used with ArrowPlus this easily lets you fetch the currently logged in user from the session, and if there is none, request a login: getUser :: Wire MyApp IpAddress User getUser = getCurrentUser <+> login * Context-sensitive wires. There is now a new wire transformer called 'context', which is essentially a context switcher establishing multiple wire threads of control: context :: (Ord a, Monad m) => Wire m a b -> Wire m a b It takes a base wire and evolves it individually for each possible input value. This is a very convenient alternative to parallel routing switches. The 'context' wire transformer never forgets a context. For many applications (like sessions in a web application) this is not desirable. For them there is an alternative transformer 'contextLimited', which allows you to garbage-collect. Minor changes: * More responsive calculus wires. Reflect changes immediately without the one-instant delay. If you need the delay for feedback, use 'delay' explicitly. * New analysis wires 'collect' and 'lastSeen'. * New inhibition wires 'forbid_', 'inhibit_' and 'require_'. * Separate session starting/cleaning functions in FRP.NetWire.Session. However, the exception-safe withWire function is the preferred way. * Optimized internal representation. There are now only two constructors WGen and WArr for the Wire type instead of four. I dropped WConst and WId, because they are representable as special cases of WArr without noticable performance loss. This improves overall performance, because now there are only two constructors to pattern-match against. Right now I'm working intensively on netwire, so if you have any feature requests, just mail me. There is a good chance to find your desired feature implemented quickly. Note: I have been too quick with version numbers. Major version 2 will be the one with the stabilized API, so for now with increasing minor version numbers (but not patch levels!) you have to be prepared for API changes, although I don't expect big changes anymore. Enjoy and keep the feedback coming! Greets, Ertugrul -- nightmare = unsafePerformIO (getWrongWife >>= sex) http://ertes.de/

1 0

ANN: cabal-ghci 0.1
by Etienne Laurin 11 Sep '11

11 Sep '11

Hello fellow hackers. Here is a helpful package I wrote to ease the development of projects using cabal. It includes a :cabalset ghci command to set ghci options for your project, and a cabal-ghci executable to launch ghci with those options. Do you get errors when loading your project into ghci? Prelude> a ← return 1 <interactive>:1:3: Not in scope: `←' Prelude> :load Main Could not find module `Main': it is not a module in the current program, or in any known package. cabalset fixes these problems for you. Prelude> :cabalset :set "-i/home/atnnn/up1/src/dist/build/autogen" :set "-i/home/atnnn/up1/src/web" :set "-XUnicodeSyntax" :set "-Wall" Prelude> a ← return 1 <interactive>:1:12: Warning: Defaulting the following constraint(s) to type `Integer' (Num t0) arising from the literal `1' Prelude> :load Main Ok, modules loaded: Main, Application, Handlers. This works from any sub-directory of your project and accepts the following arguments: -fflag enable flag -f-flag disable flag exec load options for the exec executable To install: $ cabal update $ cabal install cabal-ghci $ cat >> ~/.ghci :m + Distribution.Dev.Interactive :def cabalset cabalSet :cabalset :m - Distribution.Dev.Interactive ^D Etienne Laurin

5 9

hledger subcommand detection, RFC
by Simon Michael 11 Sep '11

11 Sep '11

When I split up the hledger package, I always intended to make the hledger program act as a single front end for hledger-* executables. I finally got around to trying that, just pushed to darcs [1]. So hledger now searches your PATH at startup and offer any hledger-* executables as subcommands. To make this fully modular, so that options can be reused and third-party add-ons don't need to be baked in to hledger in any way, there has also been a rather extensive options overhaul, using cmdargs. Below is an example [2] of the new help output with all hledger-* packages plus a local hledger-report.hs script installed. I'm not sure if this is fully safe, cross-platform, robust, quick enough to be unnoticeable, etc. I have about 20 dirs and 4k files in my PATH. I haven't noticed a slowdown on macbook or vps, but have not measured. More testing of this on diverse platforms and machines, or code review, would be very welcome. -Simon [1] http://joyful.com/darcsden/simon/hledger/browse/hledger/Hledger/Cli/Options… [2] $ hledger hledger [COMMAND] ... [OPTIONS] run the specified hledger command. hledger COMMAND --help for more detail. In general, COMMAND should precede OPTIONS. Misc commands: add prompt for new transactions and append them to the journal convert show the specified CSV file as hledger journal entries test run self-tests, or just the ones matching REGEXPS Report commands: accounts (or balance) show matched accounts and their balances entries (or print) show matched journal entries postings (or register) show matched postings and running total activity show a barchart of transactions per interval stats show quick statistics for a journal (or part of it) Add-on commands found: chart [-- OPTIONS] run the hledger-chart program interest [-- OPTIONS] run the hledger-interest program report [-- OPTIONS] run the hledger-report program vty [-- OPTIONS] run the hledger-vty program web [-- OPTIONS] run the hledger-web program -? --help Display help message --debug Show extra debug output -V --version Print version information

3 4

Improvements to Vim Haskell Syntax file - Is anyone the maintainer?
by David Terei 10 Sep '11

10 Sep '11

Hi All, I have made some improvements to the syntax highlighting file for Haskell included with Vim. I'd like to push them upstream. The haskell.vim syntax file included with Vim lists this mailing list as the maintainer. I take that to mean no one. Is this right or does someone actually have responsibility for the Vim Haskell syntax file? here are some screenshots of the changes I've made: http://www.scs.stanford.edu/~davidt/vim-syntax.html Cheers, David

3 2

compiler "pragma"'s
by Vasili I. Galchin 10 Sep '11

10 Sep '11

Hello, Is there a list of compiler pragmas? Specifically I am looking at how to specify more than one type variable in a class definition. Also I have forgotten the "meta syntax" for specifying a pragma ... some kind of Haskell comment? Thanks, Vasili

3 2

matrix manipulations and calculations
by Thomas Friedrich 09 Sep '11

09 Sep '11

Hi everyone, I need to do a lot of matrix manipulation and it needs to be fast. There seems to be a whole bunch of packages dealing with matrix calculations on HackageDB. I am wondering, what is the current state-of-the-art? I started of using hmatrix [1] now. However, I eventually would like to use some functions from the statisics-package [2]. They both use different types for their vector representations. In hmatrix a vector is of type Data.Packed.Vector. In the statistics package, vectors are based on [3]. I found this: http://www.haskell.org/haskellwiki/Linear_algebra but the last post is from 2005. That is 6 years ago. What is the state now? What would you recommend me to use? I don't want to code and then realise in a few month, I made the wrong decision. So, help is greatly appreciated. Thanks, Thomas -- [1] http://hackage.haskell.org/package/hmatrix-0.12.0.0 [2] http://hackage.haskell.org/package/statistics-0.9.0.0 [3] http://hackage.haskell.org/package/vector-0.9

4 3

First CfP: International Conference on Test and Proofs (TAP 2012), Next Deadline: 2011-12-14
by Achim D. Brucker 09 Sep '11

09 Sep '11

(Apologies for duplicates) CALL FOR PAPERS Full and short Research Paper, Industrial Expierence Papers, Tool Papers Abstract submission: Dec 14, 2011, Paper submission Dec 21, 2012 6th INTERNATIONAL CONFERENCE ON TEST AND PROOFS (TAP 2012) http://lifc.univ-fcomte.fr/tap2012/ May 31 - June 1, 2012, Prague, Czech Republic Part of the TOOLS Federated Conferences 2012 http://tools2012.fit.cvut.cz/ The TAP conference is devoted to the convergence of proofs and tests, to the application of techniques from both sides and their combination for the advancement of software quality. Test and Proof seem to be contradictory techniques: if you have proved your program to be correct, it is fruitless to comb it for bugs; and if you are testing it, that is surely a sign that you have given up on any hope to prove its correctness. Accordingly, proofs and tests have, since the onset of software engineering research, been pursued by distinct communities. However, the development of both approaches lead to the discovery of common issues and to the realization that each may need the other. The emergence of model checking has been one of the first signs that contradiction may yield to complementarity. Further evidence give test data generation techniques from models or programs which boil down to constraint resolution techniques for relatively large formula; the advent of powerful SMT solvers have therefore powered new testing techniques. Finally, since formal, proof-based verification is costly, testing invariants and background theories can be helpful to detect errors early and to improve cost effectivity. Summing up, in the past few years an increasing number of research efforts have encountered the need for combining proofs and tests, dropping earlier dogmatic views of incompatibility and taking instead the best of what each of these software engineering domains has to offer. The TAP conference aims to bring together researchers and practitioners working in the converging fields of testing and proving, and will offer a generous allocation of papers, panels and informal discussions. Topics of interest cover theory definitions, tool constructions and experimentations and include (other topics related to TAP are welcome): - Transfer of concepts from testing to proving (e.g., coverage criteria) and from proving to testing - Program proving with the aid of testing techniques - New problematics in automated reasoning emerging from specificities of test generation - Verification and testing techniques combining proofs and tests - Generation of test data, oracles, or preambles by deductive techniques such as: theorem proving, model checking, symbolic execution, constraint logic programming - Model-based testing and verification - Generation of specifications by deduction - Automatic bug finding - Debugging of programs combining static and dynamic analysis - Formal frameworks - Tool descriptions and experience reports - Case studies combining tests and proofs - Applying combination of test and proof techniques to new application domains such as validating security procotols or vulnerability detection of programs Important Dates: ================ Abstract submission: December 14, 2011 Paper submission: December 21, 2011 Notification: March 2, 2012 Camera ready version: March 19, 2012 TAP conference: May 31 - June 1, 2012 (to be confirmed) TOOLS conferences (TOOLS, ICMT, SC, TAP): May 29 - June 01, 2012 Conference Chairs: Bertrand Meyer Program Chairs: Achim D. Brucker and Jacques Julliand Program Committee (to be extended): =================================== Nazareno Aguirre, Bernhard K. Aichernig, Paul Ammann, Dirk Beyer, Nikolaj Bjorner, Robert Clarisó, Marco Comini, Catherine Dubois, Gordon Fraser, Angelo Gargantini, Patrice Godefroid, Martin Gogolla, Arnaud Gotlieb, Reiner Hähnle, Bart Jacobs, Thierry Jéron, Gregory Kapfhammer, Nikolai Kosmatov, Victor Kuliamin, Karl Meinke, Holger Schlingloff, T.H. Tse, Margus Veanes, Luca Viganò, Burkhart Wolff, Fatiha Zaidi Submission: =========== Please submit your papers via http://lifc.univ-fcomte.fr/tap2012/. TAP 2012 will accept two types of papers: - Research papers: full papers with at most 16 pages in LNCS format (pdf), which have to be original, unpublished and not submitted elsewhere. - Short contributions: work in progress, (industrial) experience reports or tool demonstrations, position statements; an extended abstract with at most 6 pages in LNCS format (pdf) is expected. Accepted papers will be published in the Springer LNCS series and will be available at the conference. The contents of previous TAP proceedings is available at: http://www.informatik.uni-trier.de/~ley/db/conf/tap/ -- Dr. Achim D. Brucker, SAP Research, Vincenz-Priessnitz-Str. 1, D-76131 Karlsruhe Phone: +49 6227 7-52595, Fax: +49-6227 78-50502, http://www.brucker.ch

1 0