June 2006 - Haskell - Haskell.org

Re: [Haskell] Edison 1.2 and profiling
by Mario Blazevic 04 Jun '06

04 Jun '06

> > Enabling profiling can significantly change the space behavior of > programs. There may be some important optimizations that don't occur > in the presence of profiling, and the profiling data itself takes > some memory. I have to admit, I don't know a lot about GHC profiling > so I'm afraid I can't be of much help. Me neither. I tried several things: Just turning off the optimizations doesn't cause the problem. Optimization + profiling (-O -prof) has the same problem as -prof alone. It doesn't matter if -auto-all option is there or not. And the profiled program gobbles up all memory with or without the +RTS -p option. > If you can scale down the > problem size to do your profiling, that may keep it from being OOM'd > and let you get some useful data. I'll try, but I doubt it. The program seems to misbehave at the very start of the execution. > > The > > non-profiled version of the program runs perfectly well. In fact, the > > PatriciaLoMap from Edison seems to be about two to three times faster > > than Data.Map library for my purposes. > > That's good to hear. What's the usage pattern? As soon as I get the execution profile, I'll send it to you. But it's basically a lot of small maps, with few insertions, moderate lookups and folds, and lots of unions. The one function I'm missing from Data.Map library is mapAssoc, but I can manage without it. One thing you might find interesting about the usage pattern is that my keys are not actually of type Int, but they contain an Int field. To get around PatriciaLoMap's restriction that keys must be Ints, I'm abstracting it as follows: import qualified Data.Edison.Assoc.PatriciaLoMap as PLM newtype Map term = PLM (PLM.FM (Label, term)) map f (PLM map) = PLM (PLM.map (\ (k, v)-> (k, f k v)) map) foldWithKey f x (PLM map) = PLM.fold (uncurry f) x map filter f (PLM map) = PLM (PLM.filter (f . snd) map) adjust f k (PLM map) = PLM (PLM.adjust (right f) (labelId k) map) where right f (x, y) = (x, f y) lookupM k (PLM map) = fmap snd (PLM.lookupM (labelId l) map) ... I wonder if PatriciaLoMap would be faster if it applied this pattern internally? Basically, could a structure like the following one make sense in Edison itself? class Countable x where toInt :: x -> Int data Countable k => GenericPLM k v = PLM.FM (k, v)

1 0

My summer of code project: HsJudy
by Caio Marcelo 04 Jun '06

04 Jun '06

Hey Haskell folks! I'm Caio Marcelo and my project for this Summer of Code is "Fast Mutable Collection Types for Haskell", I'll be implementing a lot of APIs for data collections (like Map and Arrays) using Judy library as backend. If you want to know more about it, take a look at my SoC blog at http://cmarcelo.blogspot.com. The repo is http://svn.openfoundry.org/pugs/third-party/HsJudy/. Comments, suggestions or ideas are all welcome. Thank you Haskell.org people for accepting my application! -- Caio Marcelo

10 11

Edison 1.2 and profiling
by Mario Blazevic 04 Jun '06

04 Jun '06

The profiled version of the latest Edison library doesn't work for me. I got Edison 1.2 via darcs. I used the regular 'make system' command to build and install the regular (non-profiled) version of Edison. Then I went to edison-api and edison-core subdirectories, in turn, and executed the Cabal commands: runhaskell Setup.hs clean runhaskell Setup.hs configure --enable-library-profiling runhaskell Setup.hs build runhaskell Setup.hs install The library seems to be properly installed: $ ls -l /usr/local/lib/EdisonCore-1.2/ghc-6.4.2/ total 9985 drwxr-xr-x 3 root root 72 Jun 3 20:28 Data -rw-r--r-- 1 root root 1939264 Jun 4 12:08 HSEdisonCore-1.2.o -rw-r--r-- 1 root root 2932114 Jun 4 12:08 libHSEdisonCore-1.2.a -rw-r--r-- 1 root root 5339788 Jun 4 12:08 libHSEdisonCore-1.2_p.a $ ls -l /usr/local/lib/EdisonCore-1.2/ghc-6.4.2/Data/Edison/Assoc/ total 720 -rw-r--r-- 1 root root 86508 Jun 4 12:08 AssocList.hi -rw-r--r-- 1 root root 86513 Jun 4 12:08 AssocList.p_hi -rw-r--r-- 1 root root 61267 Jun 4 12:08 Defaults.hi -rw-r--r-- 1 root root 61272 Jun 4 12:08 Defaults.p_hi -rw-r--r-- 1 root root 67740 Jun 4 12:08 PatriciaLoMap.hi -rw-r--r-- 1 root root 67745 Jun 4 12:08 PatriciaLoMap.p_hi -rw-r--r-- 1 root root 56208 Jun 4 12:08 StandardMap.hi -rw-r--r-- 1 root root 56213 Jun 4 12:08 StandardMap.p_hi -rw-r--r-- 1 root root 86239 Jun 4 12:08 TernaryTrie.hi -rw-r--r-- 1 root root 86244 Jun 4 12:08 TernaryTrie.p_hi I compile the profile version of my program using the following line: ghc -package text -package EdisonAPI -fglasgow-exts --make Main -o gens-prof -odir profiler -hidir profiler -prof -auto-all The ghc version is 6.4.2, by the way. When I run the compiled program, it keeps grabbing memory until it's killed by the system. The non-profiled version of the program runs perfectly well. In fact, the PatriciaLoMap from Edison seems to be about two to three times faster than Data.Map library for my purposes. Lots of unions. So you might say that I don't need profiling any more, but still... Any ideas on what might have gone wrong?

2 1

ANNOUNCE: Edison 1.2
by Robert Dockins 03 Jun '06

03 Jun '06

Fellow Haskellers, I am pleased to announce the final, stable release of Edison 1.2. Edison is a library of efficient, purely-functional data structures for Haskell. Many thanks to all of you who helped with your comments and suggestions during the release candidate phase! Edison 1.2 represents the first major release of Edison in several years and my first release as maintainer. The following list highlights major changes since the 1.1 release: * Typeclasses updated to use functional dependencies * Several new implementations * Modules renamed to use the hierarchical module extension * Documentation haddockized * Source moved to a darcs repository * Build system cabalized * Unit tests integrated into a single driver program which exercises all the concrete implementations shipped with Edison The project homepage is located at: http://www.eecs.tufts.edu/~rdocki01/edison.html API documentation is avaliable at: http://www.eecs.tufts.edu/~rdocki01/docs/edison/index.html And the main source distribution may be downloaded from: http://www.eecs.tufts.edu/~rdocki01/projects/edison-1.2-source.tar.gz Distribution package maintainers may be interested in the following split-source distributions: http://www.eecs.tufts.edu/~rdocki01/projects/edison-core-1.2-source.tar.gz http://www.eecs.tufts.edu/~rdocki01/projects/edison-api-1.2-source.tar.gz I would also like to take this opportunity to invite you to participate in making Edison better. The following list represents the things that Edison currently most needs. If you are interested in helping, please let me know! * Additional implementations. Edison currently has a fairly large collection of sequence implementations, but few collection or associated collection implementations. * Benchmarking suite. Edison needs a comprehensive benchmarking suite to allow objective comparison of different data structure implementations on various workloads. (perhaps the Auburn project could be revived?) * Documentation. The API is extensively documented, but the individual implementations are rather sparsly documented. Also, only sequences currently have time complexity data. * Users! Your experiences help me know how to improve Edison. If you decide to use or experiment with Edison, consider droping me an email to let me know what you like and don't like about Edison. After all, the squeaky wheel gets the grease! Enjoy! Rob Dockins Talk softly and drive a Sherman tank. Laugh hard, it's a long way to the bank. -- TMBG

1 0

Last CFP: FM-Ed'06 - Formal Methods in the Teaching Lab
by J.N. Oliveira 03 Jun '06

03 Jun '06

Call for Papers -- Submission deadline: June 9, 2006 FORMAL METHODS IN THE TEACHING LAB Examples, Cases, Assignments and Projects Enhancing Formal Methods Education http://www.di.uminho.pt/FME-SoE/FMEd06/ A Workshop at the Formal Methods 2006 Symposium Workshop: Saturday, August 26, 2006 Symposium: August 21 - 27, 2006 McMaster University, Hamilton, Ontario, Canada I. ORGANIZATION This workshop is organized by the Formal Methods Europe Subgroup on Education. Dines Bj?rner (JAIST, Japan) Eerke Boiten (University of Kent, UK) Raymond Boute (Universiteit Gent, Belgium) Andrew Butterfield (Trinity College, Dublin) John Fitzgerald (University of Newcastle upon Tyne, UK) Randolph Johnson Steve King (University of York, UK) Peter Lucas Michael Mac an Airchinnigh (Trinity College, Dublin) Dino Mandrioli (Politecnico di Milano, Italy) Andrew Martin (Oxford University, UK) Jos? Oliveira (Universidade do Minho, Portugal) -- Convenor Kees Pronk (Technische Universiteit Delft, NL) Sim?o Melo de Sousa (Universidade da Beira Interior, Portugal) Wolfgang Reisig (Humboldt-Universit?t zu Berlin, Germany) Workshop Co-Chairs Raymond Boute Formal Methods Group, Department of Information Technology (INTEC), Ghent University, Ghent (Belgium) E-mail: boute(a)intec.UGent.be Jos? Oliveira Departamento de Informatica Universidade do Minho, Braga, Portugal E-mail: jno(a)di.uminho.pt II. CONCEPT OF THE WORKSHOP Motivation Quoting Dines Dines Bj?rner: "Formal Methods Education is currently facing a `trichotomy': - On the one hand, industries dealing with the design of complex and critical systems have an increasing need for methods that provide a certain degree of confidence in the result, and are often looking for external assistance in the area of formal methods from consulting companies and academia. - On the other hand, a growing number of university staff enjoys the intellectual challenge of research in this area and teaching formal techniques to students. - On the "third hand", an increasing number of students de-select formal methods in the curriculum, due to various causes and trends." One cause of the problem is a general mathphobic trend in society and education. Another cause is that, intellectually, Information Technology is the victim of its own success. Indeed, the rapid growth creates so many design and implementation tasks that can be done and, more importantly, are being done with negligible educational or scientific background that it is difficult to argue convincingly in favor of formal methods on the basis of immediate everyday necessities. Critical systems, of course, are a notable exception. These trends are so pervasive that the small minority of FM educators has little hope to curb them in the near future. More effective in the long term is instilling a higher degree of professionalism in the next generation. This requires in particular a directed, positive action towards motivating students. Theme This workshop solicits short papers, presentations, demonstrations and evaluations describing sharp classroom or lab experiments which have proved particularly beneficial to the students' understanding and motivation for formal methods. The emphasis is not on (new) theories or methods but on specific illustrations and exercises that can be used by colleagues in their own courses, perhaps applying their own formalisms. The main goals are: - to share knowledge and experience on the practicalities of teaching and learning formal methods; - to build a collection of interesting cases, examples, assignments and projects that FM teachers can use in educational activities. Format The workshop will be a forum-like event, with short presentations, demos and informal discussion slots. After the workshop, if the evaluation committee decides that there is a sufficient number of high-quality submissions, an agreement will be sought with Springer Lecture Notes in Computer Science about publishing a special volume, and authors will be invited to submit their contribution for refereeing. III. SUBMISSIONS Call for Contributions This workshop solicits papers, presentations, demonstrations and evaluations describing such material in detail and how it has been beneficial to the students' understanding and motivation. The emphasis should not be primarily on new theories or methods but on specific illustrations and exercises that can be used by colleagues in their own courses, perhaps applying their own formalisms. The central problem(s) should be clearly stated and a typical solution outline provided (using the author's preferred method), accompanied by a discussion of what educational aspect is meant to be enhanced. Contributors should motivate their techniques with a discussion of the desired knowledge and skill outcomes of the examples/case studies or projects, and a frank appraisal of their effectiveness, insofar as such an appraisal is meaningful and instructive, which we expect to be the case for most topics. Papers should be kept short (maximum 6 pages). They should be prepared preferably in LaTeX, and a pdf-file should be sent to jno(a)di.uminho.pt. Submitted papers will be evaluated by the Subgroup on Education. Timings Submission deadline: Friday, June 9, 2006 Acceptance notification: Friday, July 1, 2006 IV. REPOSITORY AND FOLLOW-UP The organization will produce a web-based resource of output from the workshop housed under http://www.fmeurope.org. Contributors willing to allow their teaching materials to be made publicly available for the community are invited to send source files, links or tools and other information that would be suitable for such an on-line repository, which the organization will keep alive on a wiki-like basis. The collected material will form the start of a compendium of examples, cases, assignments and projects, according to the following (rough) categorization. Examples are shorter items, ranging in length from a single observation to over a full page. An example is aimed at clarifying a single aspect where the essence is captured in a somewhat condensed form, with minimal clutter from side-issues. Cases are taken from situations encountered in practice, where the problems may appear in various forms: from immediately appealing (and hence motivating) but not very challenging to subtly hidden and requiring major research. Side-issues and secondary problems may be included to clarify the setting or to illustrate the need for abstraction. Assignments and projects correspond to examples and cases respectively, but the difference is that they are elaborated by the students rather than the instructors. The repository is expected to evolve in at least 3 dimensions: new items are added in their original form as time proceeds; existing items are reworked in various formalisms; experience in teaching is reported. Every one or two years, people who submit the most suitable contributions will be invited to join forces for combining their work into a "laboratory notebook". Any further suggestions are welcome.

1 0

CFP: 2nd Int'l Workshop on Automated Specification and Verification of Web Systems (WWV'06)
by Santiago Escobar 02 Jun '06

02 Jun '06

[ We apologize for multiple copies ] ******************************************************************* 2nd International Workshop on Automated Specification and Verification of Web Systems (WWV'06) Cyprus, November 15-19, 2006 (Track of ISoLA 2006) http://www.dsic.upv.es/workshops/wwv06 ******************************************************************* SCOPE The increased complexity of Web sites and the explosive growth of Web-based applications has turned their design and construction into a challenging problem. Nowadays, many companies have diverted their Web sites into interactive, completely-automated, Web-based applications (such as Amazon, on-line banking, or travel agencies) with a high complexity that requires appropriate specification and verification techniques and tools. Systematic, formal approaches to the analysis and verification can address the problems of this particular domain with automated and reliable tools that also incorporate semantic aspects. We solicit paper on formal methods and techniques applied to Web sites, Web services or Web-based applications, such as: * rule-based approaches to Web site analysis, certification, specification, verification, and optimization * formal models for describing and reasoning about Web sites * model-checking, synthesis and debugging of Web sites * abstract interpretation and program transformation applied to the semantic Web * intelligent tutoring and advisory systems for Web specifications authoring The WWV series provides a forum for researchers from the communities of Rule-based programming, Automated Software Engineering, and Web-oriented research to facilitate the cross-fertilization and the advancement of hybrid methods that combine the three areas. LOCATION WWV'06 will be held in November in Cyprus as a Special Track of the 2006 International Symposium on Leveraging Applications of Formal Methods, Verification, and Validation (ISoLA 2006). HISTORY The first WWV'05 edition was held in Valencia, Spain, during March 14-15, 2005. The workshop was attended by 42 participants from universities, research institutes, and companies from 11 countries: Austria, Canada, France, Germany, India, Italy, Japan, Mexico, Spain, UK, and USA. A selection of the papers appeared as a volume of the Elsevier series Electronic Notes in Theoretical Computer Science (ENTCS). SUBMISSION PROCEDURE Submissions must be received by July 16, 2006. In addition, an ASCII version of the title and abstract must have been submitted by July 3, 2006. Submitted papers should be at most 15 pages in the Lecture Notes in Computer Science (LNCS) style. Submitted papers should include an abstract and the author's information. See the author's instructions of LNCS style at http://www.springer.de/comp/lncs/authors.html. PUBLICATION Accepted papers will be published in a preliminary proceedings volume, which will be available during the workshop. Publication of the workshop post-proceedings in the Springer series Lecture Notes in Computer Science (LNCS) is envisaged. PROGRAM CO-CHAIRS Maria Alpuente (Technical University of Valencia, Spain) Moreno Falaschi (University of Siena, Italy) WORKSHOP CHAIR Santiago Escobar (Technical University of Valencia, Spain) PROGRAM COMMITTEE Jose Julio Alferes (Universidade Nova de Lisboa, Portugal) Maria Alpuente (Technical University of Valencia, Spain) Demis Ballis (University of Udine, Italy) Francois Bry (University of Munich, Germany) Santiago Escobar (Technical University of Valencia, Spain) Francois Fages (INRIA Rocquencourt, France) Moreno Falaschi (University of Siena, Italy) Gopal Gupta (University of Texas at Dallas, USA) Shriram Krishnamurthi (Brown University, USA) Tiziana Margaria (University of Potsdam, Germany) I.V. Ramakrishnan (State University of New York at Stony Brook, USA) Leon van der Torre (University of Luxembourg, Luxembourg) IMPORTANT DATES Abstract Submission July 3, 2006 Full Paper Submission July 16, 2006 Acceptance Notification September 17, 2006 Camera Ready October 15, 2006 Workshop November 15-19, 2006 (one day)

1 0

Announce of Arrays & References library version 0.1
by Bulat Ziganshin 01 Jun '06

01 Jun '06

Good day, Haskellers! i glad to present my new library that is compatible with Hugs 03, Hugs 05 and GHC. the library features: - Unboxed references in IO and ST monads, namely IOURef and STURef. It is a direct replacement for widely used "fast unboxed variables" modules, with interface modelled after the IORef/STRef interface and support for all integral, floating point and pointer types plus Char and Bool (see "Examples/URef.hs" for usage details). - Monad-independent interfaces to boxed and unboxed references. This allows to write algorithms that works with mutable state but still can be executed both in IO and ST monads (look at "Examples/Universal.hs" for a cool demonstration). - Syntax sugar to make using of mutable objects easier (=:, +=, -=, val, ref, uref). Mutable objects that can be used with this sugar includes all sorts of references, mutable arrays and hash tables The library also includes refactored implementation of Data.Array.* modules. The main user-visible change comparing to original is the support for dynamic arrays, i.e. arrays whose bounds may be changed at runtime ("Examples/Array/Dynamic.hs" shows various ways of their usage). I also implemented long-awaited polymorphism for unboxed arrays. The library can be downloaded as http://freearc.narod.ru/ArrayRef.tar.gz Full documentation is at http://haskell.org/haskellwiki/Library/ArrayRef I thanks Oleg Kiselyov who is pointed how to implement unboxed arrays better, Simon Marlow who gives an idea of unboxed references, and John Meacham and Jared Updike who helped me to implement dynamic arrays. I hope that this library will become a part of the next GHC and Hugs releases and may be even Haskell' std library, so i accept suggestions on its further improvements. On the other side, i will be glad to pass maintenance of the library, including it's darcs'izing, hackaging, testing (especially performance testing) and documenting to someone's other hands - volunteers welcome -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

1 0

ANNOUNCE: Kamiariduki - the system to judge your derivative work's purpose and license is valid with Ceative Commons License Works..
by shelarcy 01 Jun '06

01 Jun '06

Daar all, I am pleased to announce the preview release of kamiariduki, the system to judge your derivative work's purpose and license is valid with other works what is mainly under Creative Commons License. I don't know good phrase that explain this software directory and simply. I used DRM (Degital Rights Management) for people (or Creators) a few place. People missunderstood this system. Because DRM is usually used for closed system, not open system. sourcefourge site : https://sourceforge.net/projects/kamiariduki/ files: https://sourceforge.net/project/showfiles.php?group_id=168626 Creative Common web site: http://creativecommons.org/ This Application depend on many libraries. So if you want to know about that, you must read README file. library dependencied includes * patched version of wxHaskell (* if build with Mac OS X platform, you need to put those libraries for dependencies directory * SYB3 * Subversion trunk of AutoFroms: http://autoforms.sourceforge.net/ ) The current status of kamiariduki is very experimental. If you want to management derivative works lisence, this worls. But it lacks many fundamental functionalities and will have many bugs. Any comments are welcome. And I also release my patched version of wxHaskell above site. That includes: * http://sourceforge.net/mailarchive/forum.php?thread_id=6395572&forum_id=341… * http://sourceforge.net/mailarchive/forum.php?thread_id=8202721&forum_id=341… * http://sourceforge.net/mailarchive/forum.php?thread_id=6542156&forum_id=341… * I think calendarEvent calls not defined function onList, so I replaced by onCal * and a few fix for wxWidgets 2.6.x * I'm not use this version usually, so I don't know this works fine as wxWindows 2.4.x. That doesn''t include * unicode support http://sourceforge.net/tracker/index.php?func=detail&aid=1369929&group_id=7… * because this dropped a few functionalty now. * Haste's wxScintilla and wxSTC support http://haste.dyndns.org:8080/build_wxhaskell.php * because wxSTC is separeted from wxWidgets' build process, and doesn't have any USE_WX* macro. So change configure and make file are difficult work. * vertical toolbar http://sourceforge.net/mailarchive/forum.php?thread_id=7823794&forum_id=341… * because this fixes half of problem. If just set wxTB_VERTICAL, toolbar menu added horizonal. * resource file link patch for Windows http://sourceforge.net/mailarchive/forum.php?thread_id=7953050&forum_id=341… * because this problem isn't occure on wxWidgets 2.6.x. * Spin Cotrol Documentation patch http://sourceforge.net/mailarchive/forum.php?thread_id=8440712&forum_id=341… * I don(t see that yet. I released this file on my place, because it seems that development of wxHaskell is stopped now. And I don't have wxHaskell project's account. Patch is included kamiariduki source's patch directory, If anyone interested in to contribute above problem or others and extend for my one, I appreciate that. Thanks, -- shelarcy <shelarcy capella.freemail.ne.jp> http://page.freett.com/shelarcy/

1 0