- Haskell-Cafe - Haskell.org

Hackage and popular packages: 2010 Q1 report
by Don Stewart 05 Apr '10

05 Apr '10

Hackage 2010 Q1 report http://donsbot.wordpress.com/2010/04/03/the-haskell-platform-q1-2010-report/ After the big move of Hackage from monk to the new abbot server, here's the first report on which packages are popular, and how Hackage is doing in general. -- Don

7 7

ANN: spec2code
by Greg Fitzgerald 05 Apr '10

05 Apr '10

After 5 years of R&D, I’m proud to announce the spec2code compiler. With spec2code, developers no longer need to acknowledge the mundane details of programming, such as memory allocation, bounds-checking, byte ordering, inheritance models or performance tuning. spec2code uses the latest techniques in compiler optimization to derive a deterministic implementation from only a high-level specification. For example, an optimal quicksort is derived from the following specification. sort :: Ord a => [a] -> (b | ordered b) sort = undefined ordered :: Ord a => [a] -> Bool ordered [] = True ordered [_] = True ordered (x1:x2:xs) = x2 > x1 && ordered (x2:xs) Read as, “the function 'sort' takes a list of orderable list items, and returns a list for which each list item is greater to the one preceding it.” How does it work? By telling the compiler about the logical properties of how its output relates to its input, the compiler is sufficiently constrained to apply its genetic algorithm. It iterates over the specification with all known optimization techniques, algebraically reducing the program to an underlying machine model. The compiler uses static analysis and complexity theory in conjunction with the machine model to order implementations by their performance characteristics and choose the optimal one. Conceptually, the first phase orders algorithms by complexity, using "average case", "worst case" and "memory consumption": [MergeSort (n*log n) (n*log n) n, QuickSort (n*log n) (n*n) (log n), InsertionSort (n*n) (n*n) 1, BubbleSort (n*n) (n*n) 1 ] After choosing a high-level algorithm, spec2code does the usual low-level optimizations via LLVM to map the algorithm to optimal native code. Does it scale? Absolutely, spec2code is not confined to specifications for which optimized algorithms are already known. spec2code can be used to implement operation systems, device drivers, build systems, package management tools, and even do your shirt laundry. spec2code will change your job from programmer to specification author, giving you more time for meetings, managing email, and browsing the web. Say goodbye to those dirty Perl scripts, broken C code, and ultimately, your job. spec2code is the future of programming and the beginning of the end of mankind. Looking forward to obsoleting you soon, Greg

2 1

Finish Gtk2hs APIs update!
by Andy Stewart 05 Apr '10

05 Apr '10

Hi all, After two weeks work, i have finish update Gtk2hs APIs to Gtk+2.18.3! Gdk and Pango APIs have update to newest version. Please report any problem to gtk2hs mail-list, we can fix it as soon as we can. Below is libraries that gtk2hs support: -----------------> libraries list start <--------------- Glib Gdk Gtk+ Pango Cairo GConf GIO Glade GnomeVFS Gstreamer OpenGL SourceView SourceView2 Mozembed SOE SvgCairo VTE Webkit ----------------> libraries list end <---------------- Cheers, -- Andy

1 0

Google Summer of Code - Distributed Haskell components with Apache Tuscany
by Jean-Sebastien Delfino 05 Apr '10

05 Apr '10

Hi all, The Apache Tuscany project [1] (an implementation of OASIS SCA [2], a programming model for distributed composite applications) is participating in the Google Summer of Code program [3] and one of the GSoC student project ideas is to develop support for Haskell SCA components. The project will help assemble Haskell components with SCA components written in other programming languages, and is also a great opportunity for students to research programming models and experiment with mixing component assembly/composition and Haskell function composition. More details can be found there [4]. If you're interested in the project, please post any questions to the Tuscany dev list dev(a)tuscany.apache.org or just send me an email. [1] http://tuscany.apache.org/ [2] http://www.oasis-open.org/committees/tc_home.php?wg_abbrev=sca-assembly [3] http://socghop.appspot.com/ [4] http://issues.apache.org/jira/browse/TUSCANY-3523 -- Jean-Sebastien

1 0

Re: Haskellers hate GUIs!!
by Andrzej Jaworski 05 Apr '10

05 Apr '10

Hello Wisecrackers, Cannot resist shering with you my perspective. Computer is a logic device and logic builds complexity bottom up while graphics builds it top down. They are therefore antagonistic by nature and in the single case of their glorious interplay - The Euclidean Geometry - it blocked abstract thinking for two millennia (no zero, no negative numbers, no continuity). The implications of that facts are 3-fold: (1) Compositional GUI needs some non-geometric and perhaps non-commutative concepts (2) GUI will always set limit on the scope of your thinking (3) conceptually rich DSL is your best friend as human-computer interface. The role of GUI has gone way beyond its usefulness and now serves the industry to sell computers as candies. Apps like Acrobat Reader became little more than GUI when judged by size and only spymasters' conflicting interests strip some GUI excess to see you swim in sewers of the Internet-turned-video. This visual mania is dangerous as it discourages investigative thinking and reliance on concepts. It is shocking and in my judgement not a coincidence that major theories (like Chaos T.) were discovered using punched cards or plain pen and paper, while Artificial Intelligence research has been very acutely wounded by mouse. Down to Earth TTS reading should have human quality by now, had the researchers used conceptual and compositional approach. If linguists knew Haskell they could write DSLs to intelligently query large data sets and step by step discover human algorithms. Instead they use mostly statistics for pattern recognition. If experts could build DSLs for amplifying their own human concepts they would multiply computer power by human intelligence. Instead they are taught how to press GUI buttons. And what is behind this buttons? Statistics and designer crap! Down to the dust of your heels: make Acrobat Reader TTS read some PDF and you will hear main text intermingled with unrelated footnotes and paragraph titles and copyrights of every picture or graph repeated forever. You can correct this of course with several lines of Haskell code but then Wall Street should sell Adobe and buy You;-) I have been using Haskell for bizarre mathematical stuff but rarely feel the need for anything more than Unicode. Still I would appreciate greatly a simple compositional GUI like Clean has. Perhaps Haskell friendly drivers for selected cards is another path to simplicity? The ultimate GUI waits for operator algebra action where Haskell may show its Category Theoretical teeth. I doubt real world will ever learn Haskell so why Haskellers should pull every piece of real world crap? My advice is: stay clean, the world is wrong. Alleluia! Cheers, - Andrzej Jaworski

2 4

Happy Easter / ZJD!
by Gregory Crosswhite 04 Apr '10

04 Apr '10

Happy Easter to those of you who are Christian, and Happy Zombie Jesus Day to those of you who aren't! - Greg

2 1

HXT Namespaces and XPath
by Mads Lindstrøm 04 Apr '10

04 Apr '10

Hi all, In HXT, namespace prefixes bound by an XML document are valid in the context of an XPath. How do avoid that? An example program will clarify: simpleXml :: String simpleXml = "<soap:Body xmlns:soap=\"http://www.w3.org/2003/05/soap-envelope\"/>" nsEnv :: [(String, String)] nsEnv = [ ("s" , "http://www.w3.org/2003/05/soap-envelope") ] evalXPath :: String -> String -> [XmlTree] evalXPath xpath xml = runLA ( xread >>> propagateNamespaces >>> getXPathTreesWithNsEnv nsEnv xpath ) xml Here: evalXPath "//s:Body" simpleXml == evalXPath "//soap:Body" simpleXml Even though I only mentions the prefix "s" (and not "soap") in the function nsEnv. I do not want the XPath to see prefixes declared in the xml-document, as it means that two semantically similar XML documents can get different results when applied to the same XPath. Also, is there any way to get 'getXPathTreesWithNsEnv' to recognize XPath namespace declarations? I am thinking about declarations like: declare namespace foobar='http://foobar.org/foobar' Greetings, Mads Lindstrøm

1 0

Correct way to record state in OpenGL?
by Mark Spezzano 04 Apr '10

04 Apr '10

Hi What is the correct way to record "custom" state when using OpenGL? By this, I refer to, say, properties of a square--say it's x,y coordinates as it moves across the screen. This requires that the program keep track of the object's state as it moves. These coordinates are _not_ part of the OpenGL state machine, so is the correct way to use the State monad? If so--or if not so--how would I proceed to keep track of the state of my objects as they move about the screen? Maybe the HOpenGL implementation comes with such state-tracking devices? Please post code snippets if necessary. Thanks in advance, Mark Spezzano

3 2

A function that makes list of Bool lists from list of Int lists
by Maur Toter 04 Apr '10

04 Apr '10

Hey, I am new with Haskell so I think this will be a pretty dumb question. I would like to make a function that makes this: listbool :: [[Int]] -> [[Bool]] in this way: listbool [[1,2],[3,4]] == [[True, True],[False, False]] listbool [[1],[5,5],[5,4],[2]] == [[True],[False, False],[True, True], [False]] So always True from the elements of the first list, False from the elements of the second one, etc... Can you help me to do that? Thanks for your help! Maur Toter

7 11

replicateM over vectors
by Scherrer, Chad 04 Apr '10

04 Apr '10

Hi, I'd like to be able to do replicateM, but over a vector instead of a list. Right now I'm doing this: import qualified Data.Vector.Generic as G import qualified Data.Vector.Unboxed.Mutable as M replicateM n action = do mu <- M.unsafeNew n let go !i | i < n = action >>= M.unsafeWrite mu i >> go (i+1) | otherwise = G.unsafeFreeze mu go 0 I thought it might be useful to express this in terms of the available primitives, since this might fuse more easily, but I don't yet see a way to do it. Is there a better (more elegant and/or faster) way to write this? Thanks, Chad

4 6