- Haskell-Cafe - Haskell.org

What is Haskell used for?
by Jason Dagit 14 Nov '08

14 Nov '08

I was looking over the Hackage categories today to figure out what Haskell is *really* used for? I just really quickly sorted the categories by number of packages. It's really interesting to see System, Graphics and Network so high on the list. Clearly, next to data types and text processing, people are doing systems, graphics and network programming. C had better watch out! Jason Categories: Data (89) Text (84) System (83) Graphics (50) Network (48) Development (45) Web (41) Control (37) Language (36) Math (36) Database (33) Sound (31) Codec (26) Game (26) User Interfaces (23) Unclassified (22) Parsing (19) Compilers/Interpreters (16) Data Structures (16) Distribution (15) Testing (15) Algorithms (12) XML (12) Natural Language Processing (10) Monads (9) Pugs (9) Bioinformatics (8) Generics (8) GUI (8) Hardware (3) Foreign (7) FRP (6) Concurrency (6) Reactivity (5) Distributed Computing (5) Source-tools (5) Music (4) Cryptography (4) Editor (4) Interfaces (4) Physics (4) AI (3) Code Generation (3) Codecs (3) Other (3) Search (3) List (2) Data Mining (2) Combinators (2) Console (2) Debug (2) Numerical (2) Theorem Provers (2) Utils (2) Comonads (1) Composition (1) Desktop (1) FFI Tools (1) Finance (1) Help (1) Middleware (1) Monadic Regions (1) Numeric (1) Screensaver (1) Scripting (1) System.Console (1) Test (1) Trace (1) User-interface (1)

1 0

The Gtk Clutter API
by Darrin Thompson 13 Nov '08

13 Nov '08

http://www.clutter-project.org/ What is the degree-of-difficulty for creating Haskell bindings for Clutter? Is it simple to extend gtk2hs to cover it? Or has it been done and I just don't see it? Thanks. -- Darrin

1 0

Getting rid of functional dependencies with type proxies
by Tobias Bexelius 13 Nov '08

13 Nov '08

Hi, some time ago I asked about a problem I had with functional dependencies conflicting when using overlapping instances in a code like this: {-# OPTIONS_GHC -fglasgow-exts -fallow-undecidable-instances -fallow-incoherent-instances #-} data V2 a = V2 a a class Vec v a where dot :: v a -> v a -> a instance Num a => Vec V2 a where V2 a1 a2 `dot` V2 b1 b2 = a1*b1+a2*b2 class Mult a b c | a b -> c where (*.) :: a -> b -> c instance (Num x) => Mult x x x where (*.) = (*) instance (Vec a x) => Mult (a x) (a x) x where (*.) = dot {- Functional dependencies conflict between instance declarations: instance [overlap ok] (Num x) => Mult x x x instance [overlap ok] (Vec a x) => Mult (a x) (a x) x -} The problem here is that the dependency check is occuring too soon. Removing the dependency would solve the problem, but then you would need to put type annotations everywhere you use the multiplication operator, and that's not acceptable. However, I did found a solution I would like to share with you: Creating a "type proxy"! I find it strange that such a simple solution to such a (for me at least) common problem hasn't been documented more. This is the working code: {-# OPTIONS_GHC -fglasgow-exts -fallow-undecidable-instances -fallow-incoherent-instances #-} data V2 a = V2 a a class Vec v a where dot :: v a -> v a -> a instance Num a => Vec V2 a where V2 a1 a2 `dot` V2 b1 b2 = a1*b1+a2*b2 class Mult a b c | a b -> c where (*.) :: a -> b -> c class MultProxy a b c where (*..) :: a -> b -> c instance MultProxy a b c => Mult a b c where (*.) = (*..) instance (Num x) => MultProxy x x x where (*..) = (*) instance (Vec a x) => MultProxy (a x) (a x) x where (*..) = dot Thanks to Emil Axelsson that pointed me to http://haskell.org/haskellwiki/GHC/AdvancedOverlap, where similar problems where discussed. Again, I haven't found links to this page in any of the Haskell tutorials I've read so far. /Tobias

2 1

Proof that Haskell is RT
by Andrew Birkett 13 Nov '08

13 Nov '08

Hi, Is a formal proof that the Haskell language is referentially transparent? Many people state "haskell is RT" without backing up that claim. I know that, in practice, I can't write any counter-examples but that's a bit handy-wavy. Is there a formal proof that, for all possible haskell programs, we can replace coreferent expressions without changing the meaning of a program? (I was writing a blog post about the origins of the phrase 'referentially transparent' and it made me think about this) Andrew -- - http://www.nobugs.org -

17 28

HOpenGL shading problem
by Trin 13 Nov '08

13 Nov '08

Hi, last few days I spent making remake of glxgears. Nothing fancy, I just wanted easily configurable gears. The current performance is only about 70% of glxgears. But I have a problem with shading. Whatever I have been trying, I still don't understand how the shading can be activated. I have light, I have color, but not even Flat shading. Please can you give me a hand? Thanks, Martin PS: I didn't find any attachment size policy for this list, so I hope that the attachment is small enough.

2 1

re: [Haskell-cafe] ANNOUNCE: rewriting-0.1
by Greg Meredith 13 Nov '08

13 Nov '08

Thomas, Did you explore nominal rewrite at all? Do you know if it might be possible to use the scrap-your-nameplate package to implement some useful subset of the nominal rewrite machinery? Best wishes, --greg -- L.G. Meredith Managing Partner Biosimilarity LLC 806 55th St NE Seattle, WA 98105 +1 206.650.3740 http://biosimilarity.blogspot.com

2 1

Turning all the Nothings into Just defaultValue using Data.Generics
by David Fox 13 Nov '08

13 Nov '08

I want to use Data.Generics to write a function to turn all the Nothings in a data structure into Just defaultValue, as shown below. I get the following error because the compiler doesn't know enough about Maybe a for mkT to create the generic function that everywhere requires, I guess. Test.hs:26:16: Ambiguous type variable `a' in the constraints: `Typeable a' arising from a use of `mkT' at Senior/Test2.hs:26:16-30 `Default a' arising from a use of `justDefault' at Senior/Test2.hs:26:20-30 Probable fix: add a type signature that fixes these type variable(s) Here is the example. It all works except for "test". Any suggestions how to do this? {-# LANGUAGE DeriveDataTypeable, FlexibleContexts, FlexibleInstances, MultiParamTypeClasses, RankNTypes, TemplateHaskell, TypeSynonymInstances #-} {-# OPTIONS_GHC -fallow-undecidable-instances #-} module Test where import Data.Generics class Default a where defaultValue :: a instance Default Int where defaultValue = 0 instance Default String where defaultValue = "" instance Default (Maybe a) where defaultValue = Nothing data A = A {b :: Int, c :: Maybe String} deriving (Show, Data, Typeable) instance Default A where defaultValue = A {b = defaultValue, c = defaultValue} test = everywhere (mkT justDefault) (defaultValue :: A) where justDefault Nothing = Just defaultValue justDefault (Just x) = Just x

3 2

Linker errors to OpenGL with GHC 6.10.1
by Greg Fitzgerald 13 Nov '08

13 Nov '08

Do you know how I can fix these linker errors? C:\projects\fun>cat HelloWorld.hs import Graphics.Rendering.OpenGL import Graphics.UI.GLUT main = do (progname, _) <- getArgsAndInitialize createWindow "Hello World" displayCallback $= clear [ColorBuffer] mainLoop C:\projects\fun>ls lib GlU32.Lib glut32.lib OpenGL32.Lib glut.def glut32.dll C:\projects\fun>ghc -Llib -lglut32 -lglu32 -lopengl32 HelloWorld.hs --make Linking HelloWorld.exe ... C:\Program Files\Haskell\GLUT-2.1.1.2\ghc-6.10.1/libHSGLUT-2.1.1.2.a(Begin.o):fake:(.text+0x1cb): undefined reference to `glutGet' C:\Program Files\Haskell\GLUT-2.1.1.2\ghc-6.10.1/libHSGLUT-2.1.1.2.a(Begin.o):fake:(.text+0x8af): undefined reference to `glutMainLoop' C:\Program Files\Haskell\GLUT-2.1.1.2\ghc-6.10.1/libHSGLUT-2.1.1.2.a(Window.o):fake:(.text+0x3cd): undefined reference to `glutEntryFunc' C:\Program Files\Haskell\GLUT-2.1.1.2\ghc-6.10.1/libHSGLUT-2.1.1.2.a(Window.o):fake:(.text+0x40d): undefined reference to `glutVisibilityFunc' C:\Program Files\Haskell\GLUT-2.1.1.2\ghc-6.10.1/libHSGLUT-2.1.1.2.a(Window.o):fake:(.text+0x3595): undefined reference to `glutPassiveMotionFunc' ...hundreds more... Thanks, Greg

2 1

Re: [Haskell-cafe] Proof that Haskell is RT
by Andrew Birkett 13 Nov '08

13 Nov '08

Edsko de Vries wrote: > See "What is a purely functional language" by Sabry. Not quite a formal > proof about *Haskell*, but then we would first need a formal semantics > of Haskell to be able to do that proof ;-) Thanks for the reference, and also to everyone who replied - all very useful and interesting. For what it's worth, the blog posts I was writing are here: http://www.nobugs.org/blog/archives/2008/11/12/why-do-they-call-it-referent… http://www.nobugs.org/blog/archives/2008/11/12/why-do-they-call-it-referent… Andrew

2 1

ANNOUNCE: hpapi 0.1 release
by Michael D. Adams 12 Nov '08

12 Nov '08

I am pleased to announce the first release of hpapi, Performance API (PAPI) bindings for Haskell. The release is on Hackage: http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hpapi The source repository is at: http://code.haskell.org/hpapi/ Please send bug reports to <adamsmd [AT] cs.indiana.edu> What is it ========== This is a simple set of bindings for the Performance API (PAPI). PAPI provides access to various CPU counters such as cache-miss, instruction and pipeline stall counts. The easiest way to use this library is the withCounts function: withCounts :: [EventCode] -> IO a -> IO (a, [Integer]) Given a list of events to count and an IO operation it will run the operation and return the resulting counts. So you could do something like: main = do [l1,tlb,fp] <- withCounters [papi_l1_dcm, papi_tlb_dm, papi_fp_ins] ioOperationToMeasure printLn (l1, tlb, fp) To use it one will of course need the PAPI C library and any necessary kernel modifications installed. Status ====== The High Level interface (which will be sufficient for most people's needs) should be usable, but the low level interface is still alpha quality. That is to say, it is ready for people to start banging on it, but it is largely untested and the modules/interfaces may be reorganized based on user feedback. All of PAPI is supported with the exception of the following "advanced" features: - Multiplexing - Threads - Locks - Overflow - Statistical Profiling Future Directions ================= Future directions will be determined largely by user feedback. Principal areas that I am looking for feedback on are: - documentation - module organization - the handling of PAPI errors (currently it throws an IO error) - the design of the Haskell interface build on top of the C level PAPI_get_opt and PAPI_set_opt - what advanced features of PAPI to build bindings for next Comparison with GHC's built-in PAPI =================================== GHC supports a built-in version of PAPI (see http://hackage.haskell.org/trac/ghc/wiki/PAPI) This system is different in a few ways. First, hpapi allows you to measure specific parts of a program while GHCI's PAPI only does whole program measurement Second, since hpapi is a library you can have fine grained and sophisticated control over how things are measured. On the other hand GHC's PAPI as part of the RTS gives you less control, but on the plus side doesn't require any modification to the program. Finally, GHC's PAPI splits apart the counts that come from the garbage collector from those that come from the mutator, while hpapi combines them. This is a limitation of hpapi we hope to correct some time in the future. It remains to be seen whether it would be worth while to unify these two systems. ================ Bug reports as well as suggestions, feedback or contributions to the API or the documentation most welcome. Michael D. Adams

1 0