- Haskell-Cafe - Haskell.org

Haskell meeting in Leipzig, Germany, June 4
by Johannes Waldmann 11 May '10

11 May '10

The Fifth Haskell in Leipzig meeting will be held on June 4, see http://www.iba-cg.de/hal5.html . You can still submit proposals (and wishes) for tutorials and talks (deadline: May 14). For instance, we'd love to hear about recent advances in Haskell IDEs (eclipsefp? leksah?), and parallel/multicore programming is always a hot topic as well. The primary language is German but we can switch to English if requested/agreed upon by audience and speaker. Best regards, Johannes Waldmann, HTWK Leipzig.

1 0

Why cannot ghc find a existng module ?
by zaxis 11 May '10

11 May '10

%pacman -Q|grep xmonad xmonad 0.9.1-4 xmonad-contrib 0.9.1-2 %%xmonad --recompile Error detected while loading xmonad configuration file: /home/sw2wolf/.xmonad/xmonad.hs xmonad.hs:20:7: Could not find module `XMonad.Layout.LayoutHints': Use -v to see a list of the files searched for. %ghc-pkg list|grep xmonad xmonad-0.9.1 xmonad-contrib-0.9.1 %pacman -Ql xmonad-contrib|grep LayoutHints xmonad-contrib /usr/lib/ghc-6.12.1/site-local/xmonad-contrib-0.9.1/XMonad/Layout/LayoutHints.hi Sincerely! ----- e^(π⋅i) + 1 = 0 -- View this message in context: http://old.nabble.com/Why-cannot-ghc-find-a-existng-module---tp28507704p285… Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

4 11

ANNOUNCE: hamlet 0.2.0
by Michael Snoyman 10 May '10

10 May '10

Hi all, I'm happy to announce the second major release of Hamlet[1]. Hamlet is a HTML templating library which works via quasi-quoting, giving you compile-time assurances, type safety and very efficient HTML generation. I wrote a blog post[2] describing the changes in this version of Hamlet. The two most major are reversing of identifier order in references and dropping the template argument. Combined, these changes make a Hamlet template fit in better with Haskell code. All the other changes are noted in that same blog post. The Hamlet documentation[3] has been updated to reflect all changes. Please let me know if the documentation needs clarification on any points. Cheers, Michael [1] http://hackage.haskell.org/package/hamlet-0.2.0 [2] http://www.snoyman.com/blog/entry/hamlet-version-0-2/ [3] http://docs.yesodweb.com/hamlet/

2 1

Well-Typed are hiring
by Duncan Coutts 10 May '10

10 May '10

We are looking to hire a Haskell expert to work with us at Well-Typed as a Haskell consultant. We are seeing an increasing demand for our services, and are thus seeking to expand our capacity. This is an exciting opportunity for someone who is passionate about Haskell and who is keen to improve and promote Haskell in a professional context. The role is quite general and could cover any of the projects and activities that we are involved in as a company. The tasks may involve: * working on the Haskell compilers, libraries and tools; * Haskell application development; * working directly with clients to solve their problems. Well-Typed has a variety of clients. For some we do proprietary Haskell development and consulting. For others, much of the work involves open-source development and cooperating with the rest of the Haskell community: the commercial, open-source and academic users. The position is initially as a contractor for one year with a salary comprised of a base level of 150 GBP per day, plus a profit-dependent component. We offer flexible hours and work from home. Living in England is not required. In the longer term there is the opportunity to become a member of the partnership with a full stake in the business: being involved in business decisions, and fully sharing the risks and rewards. If you are interested, please apply via info(a)well-typed.com. Send us your CV, and tell us why you are interested and why you would be a good fit for the job. We are more than happy to answer informal enquiries. Contact Duncan Coutts or Ian Lynagh for further information, either by email or IRC. The deadline for applications is Friday 11 June 2010. -- Duncan Coutts, Haskell Consultant Well-Typed LLP, http://www.well-typed.com/

1 0

polyvariadic function for Applicative instances
by Xiao-Yong Jin 10 May '10

10 May '10

Hi, Is it possible to have a function accept variable number of arguments, such that 'f' can be instantiated to different concrete types as f :: Applicative a => (e1 -> f) -> a e1 -> A f f g a = pure g <*> a f :: Applicative a => (e1 -> e2 -> f) -> a e1 -> a e2 -> A f f g a b = pure g <*> a <*> b Thanks, Xiao-Yong -- J c/* __o/* X <\ * (__ Y */\ <

3 3

Re: [Haskell-cafe] accents
by Dupont Corentin 10 May '10

10 May '10

Thank you Daniel and Ivan, with firefox i finded out that my text file was encoded in WINDOWS-1252. So a commande line such as: iconv -f WINDOWS-1252 -t ISO-8859-1 liste.txt > liste2.txt did the trick. Alternatively, i modified my code with: myReadFile a = do h <- openFile a ReadMode te <- mkTextEncoding "CP1252" hSetEncoding h te hGetContents h Corentin On 5/7/10, Daniel Fischer <daniel.is.fischer(a)web.de> wrote: > On Friday 07 May 2010 17:05:08 you wrote: >> I don't know the encoding of my file, how to deduce it? > > What happens if you open it in a text editor? > When it looks right, can you find out from the editor what encoding it > used? > > You could try opening the file in Firefox (or some other browser), go to > View -> Character Encoding > and see what it thinks it is (click Auto-Detect if you have selected some > default encoding) if it looks right. If it looks garbled, try selecting > different encodings to see if one looks right. > > You could send me [a part of] the file and I could try and find something > out (play a little with iconv). >

1 0

darcs to mercurial migration
by Günther Schmidt 10 May '10

10 May '10

Hello, I'm switching from darcs to mercurial with some of my projects. I'd like to retain as much of the history as possible, what tools are there available for this? Günther

6 6

Type of (>>= f) where f :: a -> m b
by Milind Patil 10 May '10

10 May '10

For a function f :: a -> m b f = undefined I am having trouble understanding how the type of (>>= f) is (>>= f) :: m a -> m b where, by definition, type of (>>=) is (>>=) :: (Monad m) => m a -> (a -> m b) -> m b I do not see how (>>= f) even unifies. I mean if I code a function with the same type as (>>=) ie. tt :: (Monad m) => m a -> (a -> m b) -> m b tt = undefined type of (tt f) does not infer to the same type as (>>= f), from ghc ... (tt f) :: (Monad ((->) b)) => (m a -> b -> b1) -> b -> b1 There seems to something special about (>>=) apart from its type. And whats (Monad ((->) b))? I am new to Haskell and I may have gaps in my understanding of type inference in Haskell. regards, Milind Patil

4 4

FRP demos and tutorials page
by Tom Poliquin 10 May '10

10 May '10

Haskell Cafe'rs, My student and I (sometimes it's difficult to determine who is who) have struggled to understand Reactive for several months and thanks to some initial help from the Reactive list (thanks Paul C) have been able to write some demos and tutorials. Apologies for posting this to the Haskell Cafe, instead of the FRP list but we wanted to make our efforts available to a wider audience, and the experts there are the choir. We have created are a set of demos and tutorials for FRP ("Conal reactive"). It consists of four demos: the simple furnace, the human controlled furnace, the hybrid robot sim, and the FRP robot sim. The furnace demos both consist of a thermostat like application, in which a furnace heats up and cools down in attempt maintain a constant temperature. In the simple furnace, the goal temperature is a fixed constant, while in the human controlled furnace, the user can control the goal temperature. The robot sim demos both consist of a 3D simulation in which a robot follows the walls of an arena using a simple control algorithm. The hybrid robot sim consists of two programs, one with FRP that simulates the robot, and one GLUT program that reads piped date from the first program to display the 3D view. The FRP robot sim is a single program that simulates the same robot scenario as the first, but uses reactive GLUT and FRP to do the whole demo in reactive style. We attempted to but as much documentation in the programs as possible, as well as give an overview on the webpage about the functions we used. Please note that both of us are rather new to Haskell, and so, we may have used poor programming practices in these tutorials and demos. If you have any suggestions about them, please feel free to write to us. Website: http://www.formicite.com/dopage.php?frp/frp.html Thanks, Tom and Tony

3 3

Set Operations In Haskell's Type System
by John Creighton 09 May '10

09 May '10

This is partly a continuation from: http://groups.google.ca/group/haskell-cafe/browse_thread/thread/4ee2ca1f5eb… and http://hpaste.org/fastcgi/hpaste.fcgi/view?id=25265 Also of relevance: http://groups.google.ca/group/haskell-cafe/browse_thread/thread/9cc8858a2e5… http://www.haskell.org/haskellwiki/GHC/AdvancedOverlap http://homepages.cwi.nl/~ralf/HList/paper.pdf http://okmij.org/ftp/Haskell/typecast.html http://www.haskell.org/haskellwiki/User:ConradParker/InstantInsanity http://okmij.org/ftp/Haskell/types.html (haven't looked at this link yet) I will continue to try to solve the problem on my own but at the moment I'm able to get IsSuperSet to work but not the classes Isa, Child and IsSubSet to work. Unlike set theory IsSubSet is not the same as switching the order arguments in IsSuperSet because the searching is done in the opposite direction. In one case we are searching the parents and each child only has one parent. In the other Case we are searching the children and each parent could have multiple children). Bellow is my current code: {-# LANGUAGE EmptyDataDecls, MultiParamTypeClasses, ScopedTypeVariables, FunctionalDependencies, OverlappingInstances, FlexibleInstances, UndecidableInstances, TypeFamilies #-} {-# LANGUAGE TypeOperators #-} --10 {-# LANGUAGE FlexibleContexts #-} --{-# LANGUAGE IncoherentInstances #-} ----------------------IsSubSet ----------------------------------- data ItsNotAParrent class IsSubSet a b c | a b -> c where -- General Definition isSubSet :: a->b->c class Child a b|b->a where {} instance ( Parrent b a2, TypeEq a2 a itsAParrent, -- Child' itsAParrent a b ) => Child a b class Child' itsAParrent a b where {} instance (TypeCast b ItsNotAParrent)=>Child' F a b --No Childern instance (TypeCast b c, Parrent c b)=>Child' T a b instance (TypeCast b M)=>Child' itsAParrent a b --- Fail Case instance ( TypeEq ItsNotAParrent a itsNotAParrent, TypeEq a b iseq, IsSubSet' itsNotAParrent iseq a b c3 -- ) => IsSubSet a b c3 where -- isSubSet a b = undefined::c3 class IsSubSet' itsNotAParrent iseq a b c| itsNotAParrent iseq a b -> c where {} instance (TypeCast c T)=>IsSubSet' F T a b c where {} instance (TypeCast c F)=>IsSubSet' T iseq a b c where {} --Not sure which logic value is best for this case. instance (TypeCast c M)=>IsSubSet' itsNotAParrent iseq a b c where {} --Fail Case instance ( Child a d, IsSubSet d b c )=> IsSubSet' F F a b c where {} --bla11=isSubSet Cat Animal -----------------------Isa --------------------------------------------- class Isa' a b c|a b->c where {} --Direct Relationship class Isa a b c|a b->c where isa::a->b->c instance ( Isa' a1 b1 c1, --Direct Relationship IsSuperSet a1 a c2, --Check --20 IsSuperSet b b1 c3, -- Isa'' c1 c2 c3 a1 b1 c4 -- Decesion function -- )=>Isa a b c4 where isa a b = undefined::c4 class Isa'' c1 c2 c3 a b c4|c1 c2 c3 a b->c4 where {} -- isa :: c1->c2->c3->a->b->c4 instance Isa'' T T T a1 b1 T where {} -- isa'' c1 c2 c3 a b = T --30 instance Isa'' F c2 c3 a1 b1 F where {} -- -- isa'' c1 c2 c3 a b = F instance Isa'' c1 F c3 a1 b1 F where {} -- isa'' c1 c2 c3 a b = F instance Isa'' c1 c2 F a1 b1 F where {} -- isa'' c1 c2 c3 a b = F ---------------- Instance Isa Relations ---------------------------------- instance Isa' Animal Noun T instance (TypeCast F result) => Isa' a b result -----------------Test Relationships ----------------------------------40 --bla6 = isa Cat Noun -- --bla4 = isa Cat Verb -----------------------Basic Type Declarations --------------------------- data Noun = Noun deriving (Show) --15 data Verb = Verb deriving (Show) -- data Adjactive = Adjactive deriving (Show) data Animal=Animal deriving (Show) data Feline=Feline deriving (Show) --50 data Cat = Cat deriving (Show) data Taby_Cat=Taby_Cat deriving (Show) -----------------------Instance SubType Relations -------------------------------- data ItsAnOrphan = ItsAnOrphan instance Show ItsAnOrphan where show _ = "ItsAnOrphan" --60 class Parrent a b| a->b where -- Specific Cases parrent :: a->b -- instance Parrent Cat Feline where -- parrent a = Feline --40 instance Parrent Feline Animal where -- parrent a = Animal -- instance (TypeCast result ItsAnOrphan) => Parrent anyChild result where parrent a = undefined::result ----------------------- Generic subType Relations ------------------------------ class IsSuperSet a b c | a b -> c where -- General Definition isSuperSet :: a->b->c --instance (TypeEq b Animal T,TypeEq c F T)=>IsSuperSet a b c where --85 -- isSuperSet a b = F -- u=undefined instance ( TypeEq ItsAnOrphan b isOrphan, TypeEq a b iseq, IsSuperSet' isOrphan iseq a b c3 -- ) => IsSuperSet a b c3 where --S isSuperSet a b = undefined::c3 -- isSuperSet a b=(isSuperSet' (u::isaninmal) (u::iseq) (a::a) (b::b))::c3 class IsSuperSet' isOrphan iseq a b c| isOrphan iseq a b -> c where {} -- isSuperSet' :: isOrphan->iseq->a->b->c instance (TypeCast c T)=>IsSuperSet' F T a b c where {} -- isSuperSet' isOrphan iseq a b = T instance (TypeCast c F)=>IsSuperSet' T T a b c where {} --Not sure which logic value is best for this case. instance ( Parrent b d, IsSuperSet a d c )=> IsSuperSet' F F a b c where {} -- isSuperSet' isOrphan iseq a b = (isSuperSet a ((parrent (b::b))::d))::c instance (TypeCast c F)=>IsSuperSet' T F a b c where {} -- isSuperSet' isOrphan iseq a b = F ----------- Logical Types ------------------ data T=T -- deriving (Show) data F=F -- deriving (Show) --25 data M=M -- Fail case instance Show T where show _ = "T" instance Show F where show _ = "F" instance Show M where show _ = "M" class ToBool a where toBool :: a->Bool instance ToBool T where toBool a = True instance ToBool F where toBool a = False -----------------------Loical Type Relations ----------------------------- class TypeOr a b c|a b->c where typeOr :: a->b->c instance TypeOr T T T where typeOr a b = T --50 instance TypeOr T F T where typeOr a b = T instance TypeOr F T T where typeOr a b = T instance TypeOr F F T where typeOr a b = T ----------------------- OKMIJ ------------------------------------------ class TypeEq' () x y b => TypeEq x y b | x y -> b instance TypeEq' () x y b => TypeEq x y b class TypeEq' q x y b | q x y -> b --60 class TypeEq'' q x y b | q x y -> b instance TypeCast b T => TypeEq' () x x b instance TypeEq'' q x y b => TypeEq' q x y b instance TypeEq'' () x y F -- see http://okmij.org/ftp/Haskell/typecast.html class TypeCast a b | a -> b, b->a where typeCast :: a -> b class TypeCast' t a b | t a -> b, t b -> a where typeCast' :: t->a- >b class TypeCast'' t a b | t a -> b, t b -> a where typeCast'' :: t->a- >b --70 instance TypeCast' () a b => TypeCast a b where typeCast x = typeCast' () x instance TypeCast'' t a b => TypeCast' t a b where typeCast' = typeCast'' instance TypeCast'' () a a where typeCast'' _ x = x ----------------------- Testing Code -------------------------------- myCat=Cat bla=isSuperSet Animal Cat bla2=isSuperSet Cat Animal bla3=isSuperSet Animal Noun bla4=isSuperSet Noun Animal bla5=isSuperSet Noun Noun

3 6

Re: [Haskell-cafe] Haskell and the Software design process
by John Lato 09 May '10

09 May '10

> From: Rafael Cunha de Almeida <assinante(a)kontesti.me> > > Ivan Miljenovic <ivan.miljenovic(a)gmail.com> disse: >> On 3 May 2010 14:17, aditya siram <aditya.siram(a)gmail.com> wrote: >>> I'm a little confused about this too. I've seen many functions defined like: >>> f x = (\s -> ...) >>> which is a partial function because it returns a function and is the same as: >>> f x s = ... >> >> No, that's a partially applied function. >> >> A partial function is one such as: >> >> secondElement (_:x:_) = x >> >> Note that it is only defined for lists with at least two elements, for >> any other list (i.e. singleton or empty) it will throw an error; >> -fwarn-incomplete-patterns (which is included in -Wall) tells you >> about these. >> >> You can also argue that functions such as head are partial, as they >> explicitly throw an error if the input data isn't correct. >> >> Partial functions are bad because if you accidentally use one the >> wrong way, your entire program crashes in a flaming wreck. It's much >> better to do something like this: >> >> safeSecondElement (_:x:_) = Just x >> safeSecondElement _ = Nothing >> >> This will work with all possible input types. > > I don't think that safeSecondElement is worse than secondElement. I think it's > better for the program to crash right away when you try to do something that > doesn't make sense. > > Getting the secondElement of a list with one or less elements doesn't make > sense, so you are definetely doing something wrong if you expected the list to > have two elements, but it doesn't. It's best that the program crashes there, > than propagate the error and crash somewhere else or, worse, not crash at all > and give a wrong answer. Stepping in on the tail end of this discussion, however... The reason people argue for safeSecondElement over secondElement is exactly the reason you argue against it. Calling safeSecondElement on a list with < 2 elements forces the programmer to handle the result immediately by returning a Maybe, which requires the programmer to handle the Nothing case, corresponding to invalid input for this function. This is better than the program crashing because the programmer has to fix it before it's even released. This is exactly how to use the type system to your advantage; the error condition is indicated by the return type of the function. "Crashing at the point of the error" isn't necessarily useful in Haskell due to lazy evaluation. The code will crash when the result of the partial function is evaluated, which may be quite far away (in terms of function calls) from where the programmer would expect. John

10 14

Chicago Haskell Users Group Meeting Tuesday, May 11, 7PM
by Jeremy Shaw 09 May '10

09 May '10

Hello Chicago People! The Chicago Haskell Users Group will be meeting this Tuesday, May 11, at 7PM at The Grafton. Say goodbye to Tom Tobin (our founder), and hello to Haskell world domination! The Grafton is in Lincoln Square next to the Old Town School of Folk Music, close to the Western stop on the Brown line. We will be wishing Tom Tobin goodbye as he prepares to head back to New York. Leon Smith is possibly doing a presentation on concurrency and/or circular programming. And we will have open discussion on Haskell and non-Haskell related topics. Ideally we will be sitting in the back on the couches near the fireplace, subject to availability (alas, we can't reserve them). The Grafton has free wi-fi, and great Irish food, beer, and drinks. Stay until 9PM and hear a variety of musicians playing rare and original folk music songs. http://www.thegrafton.com/ Also, be sure to join our brand new facebook group, http://www.facebook.com/group.php?gid=119966508024820 And google discussion group: http://groups.google.com/group/haskell-chicago Thanks! - jeremy

1 0

IO (Either a Error) question
by Eugeny N Dzhurinsky 09 May '10

09 May '10

Hello! I have some sort of strange question: assume that there are 2 functions func1 :: Int -> IO (Either Error String) func2 :: String -> IO (Either Error [String]) in case if there will be no IO involved, I could use Control.Monad.Either and write something like runCalc :: Int -> IO (Either Error [String]) runCalc param = func1 param >>= func2 but with that IO stuff I can't simply do in this way. Can somebody please suggest, how to combine IO and Either monads, if that's even possible? Thank you in advance! -- Eugene Dzhurinsky

17 35

mixing map and mapM ?
by briand＠aracnet.com 09 May '10

09 May '10

I was doing the following: do status <- mapM PF.getFileStatus filenames let times = map PF.modificationTime status let sorted = sortBy (\(_, t1) (_,t2) -> compare t1 t2) (zip filenames times) and I thought, surely I can combine the status and times definitions into one line, only I can't. Hint ? Thanks, Brian

10 14

nun.haskell.org http services down?
by Jens Petersen 09 May '10

09 May '10

http://{code,community,projects}.haskell..org/ seem to be inaccessible. Could someone please look into it? Thanks, Jens

11 11

Re: [Haskell-cafe] Would it be evil to add "deriving Typeable" to newtype Q?
by Ivan Miljenovic 09 May '10

09 May '10

Re-CC'ing -cafe: On 6 May 2010 12:54, Leonel Fonseca <leonelfl(a)gmail.com> wrote: > I wasn't aware of GeneralizedNewtypeDeriving. > I just edited the source file Language.Haskell.TH.Syntax > and left: > > newtype Q a = Q { unQ :: forall m. Quasi m => m a } > deriving Typeable Hang on, is Q something actually in the template-haskell library? In that case, you can't just do "deriving (Typeable)" . However, you might be able to generate the Typeable instance using DrIFT (see http://hackage.haskell.org/package/DrIFT-cabalized for a cabal-install-able package). -- Ivan Lazar Miljenovic Ivan.Miljenovic(a)gmail.com IvanMiljenovic.wordpress.com

7 6

Haskell and the Software design process
by Jaco van Iterson 09 May '10

09 May '10

Hi I was just wondering what methods are best to design/model the software in bigger projects when you are planning to use Haskell. Is there no difference compared to other languages? Are there any Haskell tools? Cheers, Jaco PS. "Software design in Haskell" didn't give relevant hits. :(

17 22

installing haskell platform on 64-bit ubuntu 10.04 (Lucid)
by Andy Lee 09 May '10

09 May '10

Has anyone had problems getting Haskell Platform installed on Ubuntu 10.04? I've got GHC 6.12.1 installed via apt-get, and I've downloaded/untarred haskell-platform-2010.1.0.0.tar.gz. But when I run ./configure, it fails fairly quickly, with an error 'cause it can't build the "hello world" test program, complaining: <command line>: unknown package: haskell98 Anyone else seen this? Am I missing something obvious? --Andy

2 3

GHC 6.12 on OS X 10.5
by Conor McBride 08 May '10

08 May '10

Hi I thought I'd record my upgrade exerience (so far) in case anyone else finds it useful, and (more selfishly) in case anyone has some helpful advice. Summary of situation * I got 6.12 working. * It took a lot of fumbling around. * The eventual "fix" (renaming /opt/local/lib/libiconv.dylib) was a bit dodgy, but is holding for now. Long rambling narrative follows. Various features of GHC 6.12 made it very tempting to go for an earlier upgrade than I normally would. (SHE really needs GADT-style data families; everything I do needs deriving Traversable (deriving HalfZip would be nice too!).) Already there was a handy installer for Intel+Leopard macs (is anybody on the PPC+Leopard case? if not, I will be soon...). I thought "go for it". Installation, no problem. Compiling something: whoops, no mtl. OK, cabal install mtl: whoops, no acceptable cabal. Fair dos, if I'd read the warnings I'd have known that my cabal-install would be useless and tried to build the new one. QUESTION: Is the sensible upgrade path to build cabal-install 0.8 with your old GHC, before installing 6.12? Does this even work? At one point, I tried this and had some peculiar issues involving zlib... Anyhow, I've got 6.12 and I need cabal-install 0.8. Can I find it? Tricky, and http://haskell.org/cabal/ didn't help much. I'm very tolerant of busy people not quite getting round to it, and I can use google, so I find the darcs version and get cracking. This happens: ----------------------------------------------------------------------------- bash-3.2$ ./bootstrap.sh Checking installed packages for ghc-6.12.1... parsec-2.1.0.1 will be downloaded and installed. network-2.2.1.5 will be downloaded and installed. Cabal is already installed and the version is ok. mtl-1.1.0.2 will be downloaded and installed. HTTP-4000.0.8 will be downloaded and installed. zlib-0.5.2.0 will be downloaded and installed. Downloading parsec-2.1.0.1... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 15430 100 15430 0 0 12743 0 0:00:01 0:00:01 --:--:-- 19312 [1 of 1] Compiling Main ( Setup.hs, Setup.o ) Linking Setup ... Undefined symbols: "_iconv_close", referenced from: _hs_iconv_close in libHSbase-4.2.0.0.a(iconv.o) "_iconv", referenced from: _hs_iconv in libHSbase-4.2.0.0.a(iconv.o) "_iconv_open", referenced from: _hs_iconv_open in libHSbase-4.2.0.0.a(iconv.o) ld: symbol(s) not found collect2: ld returned 1 exit status Error during cabal-install bootstrap: Compiling the Setup script failed ----------------------------------------------------------------------------- At lthis point, I suspected a linker/dylib problem, but foolishly, I wanted the problem to be pretty much anything else, so I spent far too long looking the wrong way. I thought that if I could just get cabal-install working somehow, I'd be ok. I tried installing parsec with runhaskell Setup.hs {configure,build,install} but each of these commands appeared content to do precisely nothing: I found this perplexing. I tried reverting to 6.10.4 and compiling cabal-install. This made more progress, but died with some sort of zlib version snafu. (I'm sorry, I should have recorded the details but didn't.) The zlib-0.5.2.0 package did install successfully, but somehow the build for cabal-install itself went awry with an "even though this is the right version I can't find X" sort of a problem. I uninstalled 6.10.4, deleted my .cabal directory, nuked a few other relics showing up from what's probably a dumb choice of PATH setting. I had another go at 6.12, and this time I tried compiling a wee program of my own with no exciting package dependencies. Should have done that straight away, of course. Same iconv problem. This left no alternative. I found I had a /usr/lib/libiconv.dylib etc and an /opt/local/lib/libiconv.dylib etc, so I renamed the latter to opt/local/lib/moolibiconv.dylib, and suddenly, GHC became capable of linking stuff. The darcs version of cabal-install then built just fine; SHE rebuilt ok; Epigram needed a few extra LANGUAGE pragmas, but no trouble. I'm almost back where I was. I worry about this. (1) What have I broken by shifting the opt/local/lib copy of libiconv? (2) Why did that break things anyway? (3) How come I ended up with a broken library setup? (4) What else is broken? (What's worse than finding a maggot in your apple? Finding half a maggot in your apple.) I'm not at all a mac expert (I use a mac because I'm too stupid to maintain a linux installation -- the Nottingham grad students (all grown up now) told me they were fed up fixing my computer back in 04.) This dylib nonsense seems quite common. Is it MacPorts messing things up? Is there a more principled fix than the brutal thing I did? I wonder if it might be possible to build some sort of diagnostic tool to check paths and libraries for conflicts and omissions. Anyhow, sorry for the dreariness of this story, but it ends in an acceptable configuration, for the moment. I hope it can be of some small help. Happy hols Conor

10 19

Ideas for better network IO in Combinatorrent?
by Jesper Louis Andersen 08 May '10

08 May '10

Hi, Cafe! (This post got rather long, and in writing it, I probably answered some of my own questions. I'll still post it because it might be interesting to read for somebody). This post concerns the currently central problem in Combinatorrent. But the problem is interesting from another angle. If you want fast servers in Haskell, I am pretty sure you will need to crack this nut as well. Status is that my Erlang-client, which handles all IO by hand in Erlang uses much less CPU resources than the Combinatorrent client in Haskell. Especially under load. Productivity is usually around 70-75% so we spend some considerable time in the GC still. It is up from some 30%, so we are faring much better than earlier. Still, we use about 4 times as many CPU seconds per downloaded megabyte compared to Erlang. And it is not like the Erlang code is optimized: The parser loop is Erlang and not using the C drivers packeting facility. (Aside: For non-Erlangers -- In erlang you can set options on a TCP socket about how you want data to be delivered. There are several (C-hand-optimized protocol parsers for ASN.1, FastCGI, or even HTTP), but one of them designates that the TCP stream will receive data in the format of a 4 byte big endian integer header, the length, followed by the payload of that length. By setting the option {packet, 4} on the TCP link, you can "outsource" this packeting to the Erlang VM kernel in C and receive a message (in a process channel) whenever such a message arrives.) In Erlang I don't use it because it gives too erratic speed measurements. I need to know how fast I am communicating because it is used in the BitTorrent protocol to gauge which peers it is interesting to talk to. Rather, I run a handrolled parser loop, doing exactly the thing from above. Essentially the loop collects data from the socket in what is equivalent to a Lazy ByteString. When the LBS is full, it is converted to a Strict ByteString and handed on to the parser. While doing this measurements are done on the size of incoming data from the operating system. The parser is fairly naive, but it uses the Erlang-bit-syntax pattern matches to essentially decode the binary chunk with a pointer walking over it. Note that The largest packet type is 16 kilobytes + a_small_constant. Rather than reallocate the binary, the Erlang system just stores a Slice-triple (u, o, l), of the underlying byte array, u, the offset into that array, o and the length from the offset, l. When the slice disappears, the garbage collector can evict the binary. Binaries are ref-counted since they can not contain pointers and they are stored outside the process heap. This has to do with the Erlang VMs garbage collection strategy: Each Erlang process has its own GC and all message passing happens by copying. Since copying around large chunks of binary data is expensive, they are stored immutably and shared by all processes. It is fairly efficient. There is only one major copy going on when assembling the LBS into a Strict BS and that copy knows the exact size of the target allocation area, so it can just allocate the right size and then copy data into that area. Enough Erlang, what does Combinatorrent do? Combinatorrents receiver processes (There are one for each Peer we are communicating with) currently takes 46% of all allocation on their own. Why? Here is why. Our main loop in the receiver process looks like this (explanation below the source code, read it first): bs <- liftIO $ recv s 2048 when (B.length bs == 0) stopP loop s c (A.parse getMsg bs) where loop s c (A.Done r msg) = do liftIO . atomically $ writeTChan c (FromPeer (msg, fromIntegral $ msgSize msg)) loop s c (A.parse getMsg r) loop s c (prt(a)(A.Partial _)) = do bs <- liftIO $ recv s 4096 when (B.length bs == 0) stopP loop s c (A.feed prt bs) loop _ _ (A.Fail _ ctx err) = do warningP $ "Incorrect parse in receiver, context: " ++ show ctx ++ ", " ++ show err stopP [full code at: http://github.com/jlouis/combinatorrent/blob/master/src/Process/Peer/Receiv… ] The 'A' module you see is Bryan O'Sullivan's attoparsec package. This parser is incremental. Either it returns A.Done r msg, where msg is a decoded message and r is the rest of the input not parsed yet. Or, it returns a Partial parse where it demands more input to successfully parse. If you feed that continuation with more data, it happily continues on. The 'recv' you see is from Johan Tibell's network-bytestring package. It will block on the Socket s until data arrives. Then it will fill up to 4096 bytes into the strict bytestring bs. We use this to feed more data to attoparsec. One problem is that usually, we are so fast that the packet is only some 1448 bytes which make sense: the MTU of the ethernet wire is 1500 bytes and sizeable chunks are taken by TCP/IP and also the internet routers PPPoE/A. This means the bytestring is trimmed down most of the time, incurring a copy. BitTorrent connections are either dead-slow, only transferring a few bytes each second on average, to dead-fast, easily filling up the 4k window. I've toyed with the idea of an adapting receiver window size, but it is peanuts compared to the next problem it seems. The next problem is what attoparsec ends up doing. The relevant parser code is: getAPMsg :: Int -> Parser Message getAPMsg l = do c <- A.anyWord8 case c of 0 -> return Choke 1 -> return Unchoke ... 6 -> (Request <$> apW32be <*> (Block <$> apW32be <*> apW32be)) 7 -> (Piece <$> apW32be <*> apW32be <*> A.take (l - 9)) ... k -> fail $ "Illegal parse, code: " ++ show k [full code at: http://github.com/jlouis/combinatorrent/blob/master/src/Protocol/Wire.hs#L1… ] Where the attoparsec parser first reads of a character, c, designating the message type of the parse. Then it handles the message type by returning an ADT of the parse. The killer is A.take (l -9). If we look into the attoparsec source code, we see a call to its 'ensure' function. This function ensures we have enough data available for parsing. If not, it demands more input through A.Partial results. When it gets input it concatenates with '+++' which is really the 'append' function of Strict ByteStrings which are O(n) and destroys the allocation rate of the application :) My current plan for a solution is this: Read the socket and gather up chunks for a lazy bytestring. We know, due to the 32bit BE header how many bytes more to expect. When we have all of them, do the Strict BS conversion and hand that to attoparsec. It means no +++ call inside attoparsec. Furthermore I get a SBS.splitAt, yielding the slice-construction I am longing for. But are there any ideas out there for doing it even better than this? Conjure by David Himmelstrup removes the Lazy -> Strict conversion as well by preallocating the Strict ByteString and filling it directly, but I think I can live with the simpler solution for now. The only important thing currently is to get attoparsec happy, without feeding it sangria or beer :) In any event, there is food for thought on some possible extensions of network-bytestring I think. Another possibility is going back to Handle's as they have a buffer construction, but Handle's had even worse CPU resource usage if I remember correctly. I have not analyzed that problem to this depth however. I have also considered John Lato's iteratee's, but we are sorely lacking an iteratee 101 introduction. Can iteratees parse infinite streams for instance? -- J.

1 0

Shorthand method of enumerating a list a gotcha ... or is it just me?
by Gene A 08 May '10

08 May '10

The problem I see is that in both: Version: September 2006 of hugs, which is the one that is current for Ubuntu 9.10 release, and ghci 6.10.4, they both exhibit a {I think} strange behaviour, in regards to the shorthand way of calling out a list of enumerable values. I will explain the problem that I have run into with examples: Hugs> [3,7..22] [3,7,11,15,19] <- OK Hugs> map (* 1.0) [3,7,11,15,19] <- manual spec OK [3.0,7.0,11.0,15.0,19.0] Hugs> map (* 1.0) [3,7..22] <- same spec as first but !!! when mapped to with a (*1.0) to coerce them to reals: [3.0,7.0,11.0,15.0,19.0,23.0] <- went one outside of range spec. ======================================== Exactly the same behaviour from ghci 6.10.4 : Prelude> [3,7..22] [3,7,11,15,19] Prelude> map (* 1.0) [3,7..22] <- using a range [3.0,7.0,11.0,15.0,19.0,23.0] <- it screws up {at least it is not a "feature' to me} Prelude> map (* 1.0) [3,7,11,15,19] <- spelled out it acts "right". [3.0,7.0,11.0,15.0,19.0] This seems like a possible bug? or at least a sure fire trap waiting to be sprung ... one of those nasties that could really create havoc if someone is not aware of this behaviour and buries a function that include something that unwittingly coerces from an Integral to a Realfrac or Fractional. Is this a well known thing to watch out for.. or is it something that can be worked around, other then having to enumerate every value in a list rather then use the handiness of the range notation as shorthand? cheers, gene

5 4

ANN: test-framework 0.3.0
by Max Bolingbroke 08 May '10

08 May '10

I'm pleased to announce the release of version 0.3 of test-framework (http://hackage.haskell.org/package/test-framework-0.3.0) This package provides a nice test runner for HUnit, QuickCheck 1 and QuickCheck 2 tests. To see a detailed example of the runner in action, please check out our website at http://batterseapower.github.com/test-framework The highlights of this release are: * Thanks to the contributions of Rogan Creswick at Galois, you can now get JUnit XML output from the test runner using the --jxml=FILENAME option * There is a new command line option (--plain), which tells the test runner to avoid using any ANSI features - this can be handy if you are (for example) viewing test output in Emacs Cheers, Max

3 3

Re: [Haskell-cafe] Haskell and the Software design process
by Henning Thielemann 08 May '10

08 May '10

On Mon, 3 May 2010, Gregory Collins wrote: > Don Stewart <dons(a)galois.com> writes: > >> Some key points: >> >> * Avoid partial functions > > As an important corollary to this one I would add: "never throw > exceptions from pure code". They often leak out from "catch" blocks and > ruin your day. It's not possible to throw exceptions from pure code. You can only call 'error' and that's another name for 'undefined', i.e. you have a partial (non-total ?) function.

5 7

Haskell interface to Frama-C
by Tom Hawkins 08 May '10

08 May '10

I just started using Frama-C [1] for analyzing some of our embedded C programs. Pretty awesome suite of tools. Especially its ability to describe and verify function contracts with ACSL [2]. The tool suite is primarily build with OCaml. Has anyone considered building a Haskell interface to Frama-C? I lack the expertise and time for such a project. But it sure would be nice to be able to write verification plug-ins with Haskell. -Tom [1] http://frama-c.com/ [2] http://en.wikipedia.org/wiki/ANSI/ISO_C_Specification_Language

1 0

ANNOUNCE: network-protocol-xmpp 0.3
by John Millikin 07 May '10

07 May '10

XMPP[1] is a protocol for asynchronous communication via long-lived XML documents. It is most famous for underlying the Jabber and Wave protocols, but is also used for a variety of cases where security and extensibility are useful. My library, network-protocol-xmpp[2], is an implementation of most of RFC 3920 and a bit of RFC 3921. It supports opening client-to-server and component-to-server sessions, which is useful for implementing XMPP-based clients. This library's interface is very simple: clients may start a session with 'runClient' or 'runComponent', send and receive stanzas, and resume stored sessions. Later, I intend to add additional modules to support features such as MUC or file transmission, but for now I'd like to make sure the core library works. There's an example XMPP client, echo.hs[3], in the Darcs repository. Server-to-server sessions and accepting connection requests are not yet supported, so it's not usable as the foundation of an XMPP server yet. However, given that XMPP prefers to place features on the servers rather than clients, it's likely that writing an XMPP server in Haskell would take a *lot* of effort anyway. You're probably better off just using ejabberd or OpenFire. My thanks go to Stephan Maka, who let me know that this library actually has users, and who contributed most of the Component connection implementation. ---------- Additionally, I'm trying an experiment intended to break the gridlock of Haskell XML libraries. Currently, there are multiple options for XML handling (HXT, HaXml, libxml, tagsoup, xml, xml-basic) but each defines their own data types. This makes it very difficult to mix-and-match -- I can't easily use one library's parser with another's tree manipulation combinators. Simple use cases, such as using HXT for serializing, drag in dozens of dependencies on libraries like HTTP and QuickCheck. A library's poor data type definitions (such as HaXml's lack of support for namespaces) means that an entire set of potentially useful operations is unusable. And a library's choice propagates! In n-p-xmpp 0.2, which used HXT, clients were forced to figure out arrows before they could perform even rudimentary inspection of received stanzas. To fix this, I've created the xml-types[4] package. Its only dependency is the "text" library, and it defines full types for most XML structures[5]. My hope is that other libraries will replace their own types with these, so that (for example) I can parse events with libxml, manipulate them with HXT's arrow API, and write them back out using a custom serialiser without having to jump through type conversion hoops. It might take some time before any major XML libraries are transitioned to these types, but maybe in a year or two there'll be enough to be useful. [1] http://en.wikipedia.org/wiki/Extensible_Messaging_and_Presence_Protocol [2] http://hackage.haskell.org/package/network-protocol-xmpp [3] http://patch-tag.com/r/jmillikin/xmpp/snapshot/current/content/pretty/examp… [4] http://hackage.haskell.org/package/xml-types [5] I haven't included types for inner entity declarations in DTDs, out of laziness, because does anybody actually *use* those?

2 2

{-# LANGUAGE DeriveApplicative #-} ?
by Pavel Perikov 07 May '10

07 May '10

Hi, list!. Now in 6.12.1 we have DeriveFunctor, DeriveFoldable and DeriveTraversable. This greatly simplifies the "reuse structure" style of programming. Some structure (not just _data_ structure) got captured in ADT and can be reused for various purposes. Wouldn't it be nice to have the ability to derive Applicative as well? It shouldn't be more difficult than deriving Functor but will provide exciting possibilities. Just think about liftA2. Please note that deriving Monad is quite different story and seem generally unfeasible. Suggestions welcome. P.

3 3

accents
by Dupont Corentin 07 May '10

07 May '10

Hello, i have a list of french words with accents. How could i handle them? If i load them with ghci i get: > a <- readFile "list.txt" > head $ lines a "abn\233gation" putStrLn displays a strange character for the "é". Cheers, Corentin

3 4

Type system game (was: Nomic game in Haskell)
by Dupont Corentin 07 May '10

07 May '10

Hello, I think your type should be: type Board a b c d e f g h i = Either (Three a b c) (Either (Three d e f) (Either (Three g h i) (Either (Three a d g) (Either (Three b e h) (Either (Three c f i) (Either (Three a e i) (Three c e g))))))) as far as i can understand, it seems thats it's mandatory for Player 1 to play in a,b,c to win? So victory should go to player 2? Cheers, Corentin On 4/16/10, Dan Piponi <dpiponi(a)gmail.com> wrote: > On Thu, Apr 15, 2010 at 4:58 PM, Ashley Yakeley <ashley(a)semantic.org> wrote: > >> type Board a b c d e f g h i = >> Either (Three a b c) >> (Either (Three d e f) >> (Either (Three g h i) >> (Either (Three a d g) >> (Either (Three b e h) >> (Either (Three c f i) >> (Either (Three a e i) >> (Either (Three c e g) >> ))))))) > > In the service of readability we could also define: > > data X = X > data O > > Though the victory conditions aren't precisely the usual ones. > -- > Dan > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe >

1 0

[Long, probably not-beginners anymore] Parallel folds and folds as arrows (was: Re: [Haskell-beginners] Re: When, if ever, does Haskell "calculate once"?)
by Maciej Piechotka 07 May '10

07 May '10

On Thu, 2010-05-06 at 23:46 +0200, Daniel Fischer wrote: > Share.share :: GHC.Types.Int > GblId > [Str: DmdType] > Share.share = > case GHC.List.$wlen @ GHC.Integer.Type.Integer Share.share_a 0 > of ww_amc { __DEFAULT -> > GHC.Types.I# (GHC.Prim.+# ww_amc ww_amc) > } > Hmm. What's the name of this form and how to get it? > > No. > > cFoldl' f g (b0,c0) xs0 = lgo b0 c0 xs0 > where > lgo b c [] = (b,c) > lgo !b !c (x:xs) = lgo (f b x) (g c x) xs > Ok. Fixed (I tried fast rewrite from foldr') > > 64-bit system? I get > 64 bit, GHC 6.12.2. % ghc -V The Glorious Glasgow Haskell Compilation System, version 6.12.2 % file a.out a.out: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.9, not stripped > > And that is strange, because I get the same figures for that one as for the > first (times differ by a few hundredths of a second). Fixed or non-fixed version? > Is that a difference between 32-bit code generator and 64-bit or between > GHC versions (6.12.2 here, but 6.10.3 gives roughly the same results)? > Hmm. Compiler and platform matches. Unless you use some other 64-bit platform - not x86-64 ;) > > > > main = > > > print $! uncurry (+) (cFoldl' lengthFold lengthFold > > > (0, 0) [1..size]) > > And that gives the same figures as the other two (plus/minus 0.05s). > > > > > All are compiled with optimizations. > > All compiled with -O2. > Hmm. Difference between -O1 and -O2 Fixed versions and with sum (and -O2): > main = let a = [1..size] > l = length a + sum a > in print $! l Lot's of memory(Over 3 GiB). I voluntarily killed process > main = print $! length [1..size] + sum [1..size] Lot's of memory(Over 3 GiB). I voluntarily killed process. So far as being inplace. > main = print $! uncurry (+) (cFoldl' lengthFold (+) (0, 0) [1..size]) 5000000150000000 16,889,753,976 bytes allocated in the heap 3,356,480 bytes copied during GC 1,976 bytes maximum residency (1 sample(s)) 28,200 bytes maximum slop 1 MB total memory in use (0 MB lost due to fragmentation) Generation 0: 32216 collections, 0 parallel, 0.29s, 0.38s elapsed Generation 1: 1 collections, 0 parallel, 0.00s, 0.00s elapsed INIT time 0.00s ( 0.00s elapsed) MUT time 14.89s ( 15.12s elapsed) GC time 0.29s ( 0.38s elapsed) EXIT time 0.00s ( 0.00s elapsed) Total time 15.18s ( 15.50s elapsed) %GC time 1.9% (2.4% elapsed) Alloc rate 1,134,094,110 bytes per MUT second Productivity 98.1% of total user, 96.1% of total elapsed ./a.out +RTS -s 15.18s user 0.11s system 98% cpu 15.503 total Lowered to size = 100000 (bigger causes stack overflow in first main): > main = let a = [1..size] > l = length a + sum a > in print $! l 5000150000 22,045,352 bytes allocated in the heap 18,781,768 bytes copied during GC 6,316,904 bytes maximum residency (4 sample(s)) 3,141,912 bytes maximum slop 17 MB total memory in use (0 MB lost due to fragmentation) Generation 0: 23 collections, 0 parallel, 0.05s, 0.05s elapsed Generation 1: 4 collections, 0 parallel, 0.03s, 0.04s elapsed INIT time 0.00s ( 0.00s elapsed) MUT time 0.02s ( 0.03s elapsed) GC time 0.08s ( 0.09s elapsed) EXIT time 0.00s ( 0.00s elapsed) Total time 0.10s ( 0.12s elapsed) %GC time 78.4% (75.8% elapsed) Alloc rate 1,002,334,818 bytes per MUT second Productivity 20.6% of total user, 17.2% of total elapsed ./a.out +RTS -s 0.10s user 0.02s system 96% cpu 0.129 total > main = print $! length [1..size] + sum [1..size] 5000150000 30,077,024 bytes allocated in the heap 17,482,888 bytes copied during GC 5,602,600 bytes maximum residency (4 sample(s)) 3,144,312 bytes maximum slop 15 MB total memory in use (0 MB lost due to fragmentation) Generation 0: 38 collections, 0 parallel, 0.04s, 0.05s elapsed Generation 1: 4 collections, 0 parallel, 0.03s, 0.03s elapsed INIT time 0.00s ( 0.00s elapsed) MUT time 0.02s ( 0.03s elapsed) GC time 0.07s ( 0.08s elapsed) EXIT time 0.00s ( 0.00s elapsed) Total time 0.09s ( 0.11s elapsed) %GC time 72.0% (70.0% elapsed) Alloc rate 1,156,986,613 bytes per MUT second Productivity 26.9% of total user, 22.9% of total elapsed ./a.out +RTS -s 0.09s user 0.02s system 97% cpu 0.116 total > main = print $! uncurry (+) (cFoldl' lengthFold (+) (0, 0) [1..size]) 5000150000 17,128,128 bytes allocated in the heap 10,608 bytes copied during GC 2,072 bytes maximum residency (1 sample(s)) 28,024 bytes maximum slop 1 MB total memory in use (0 MB lost due to fragmentation) Generation 0: 32 collections, 0 parallel, 0.00s, 0.00s elapsed Generation 1: 1 collections, 0 parallel, 0.00s, 0.00s elapsed INIT time 0.00s ( 0.00s elapsed) MUT time 0.02s ( 0.02s elapsed) GC time 0.00s ( 0.00s elapsed) EXIT time 0.00s ( 0.00s elapsed) Total time 0.02s ( 0.02s elapsed) %GC time 5.3% (3.9% elapsed) Alloc rate 951,721,286 bytes per MUT second Productivity 94.7% of total user, 90.2% of total elapsed ./a.out +RTS -s 0.02s user 0.00s system 96% cpu 0.024 total It seems that the best, at least for large inputs. ----------------------------------------------------------------------- On the other hand it seems to form an arrow[1]. First the result of test: 5000000150000000 12,800,063,440 bytes allocated in the heap 2,545,048 bytes copied during GC 1,968 bytes maximum residency (1 sample(s)) 28,216 bytes maximum slop 1 MB total memory in use (0 MB lost due to fragmentation) Generation 0: 24414 collections, 0 parallel, 0.24s, 0.29s elapsed Generation 1: 1 collections, 0 parallel, 0.00s, 0.00s elapsed INIT time 0.00s ( 0.00s elapsed) MUT time 7.18s ( 7.35s elapsed) GC time 0.24s ( 0.29s elapsed) EXIT time 0.00s ( 0.00s elapsed) Total time 7.42s ( 7.64s elapsed) %GC time 3.3% (3.8% elapsed) Alloc rate 1,782,759,994 bytes per MUT second Productivity 96.7% of total user, 93.9% of total elapsed ./a.out +RTS -s 7.42s user 0.09s system 98% cpu 7.648 total (Yes - lower then the code above - result reproducable). Code: > {-# LANGUAGE BangPatterns #-} > import Control.Arrow > import Control.Category > import Data.List > import Prelude hiding (id, (.)) > > newtype FoldlArrow a b c = FoldlArrow (a -> b -> c) > > instance Category (FoldlArrow a) where > id = FoldlArrow $ \a !b -> b > (FoldlArrow f) . (FoldlArrow g) = > FoldlArrow $ \a !b -> let !c = g a b > in f a c > > instance Arrow (FoldlArrow a) where > arr f = FoldlArrow $ const f > first (FoldlArrow f) = FoldlArrow $ \a (!b, !c) -> (f a b, c) > second (FoldlArrow f) = FoldlArrow $ \a (!b, !c) -> (b, f a c) > (FoldlArrow f) *** (FoldlArrow g) = > FoldlArrow $ \a (!b, !c) -> (f a b, g a c) > (FoldlArrow f) &&& (FoldlArrow g) = > FoldlArrow $ \a !b -> (f a b, g a b) > > instance ArrowChoice (FoldlArrow a) where > left (FoldlArrow f) = FoldlArrow left' > where left' a (Left !l) = (Left $! f a l) > left' a (Right !r) = (Right r) > right (FoldlArrow f) = FoldlArrow right' > where right' a (Left !l) = (Left l) > right' a (Right !r) = > (Right $! f a r) > (FoldlArrow f) +++ (FoldlArrow g) = > FoldlArrow choice > where choice a (Left !l) = (Left $! f a l) > choice a (Right !r) = (Right $! g a r) > (FoldlArrow f) ||| (FoldlArrow g) = > FoldlArrow choice > where choice a (Left !l) = f a l > choice a (Right !r) = g a r > > instance ArrowApply (FoldlArrow a) where > app = FoldlArrow $ \a (FoldlArrow !f, !b) -> f a b > > doFoldl :: FoldlArrow a b b -> b -> [a] -> b > doFoldl (FoldlArrow f) = foldl' (flip f) > > element :: FoldlArrow a b a > element = FoldlArrow const > > lengthA :: FoldlArrow a Int Int > lengthA = arr (+1) > > sumA :: Num a => FoldlArrow a a a > sumA = arr (uncurry (+)) . (element &&& id) > > size = 100000000 > > main = > print $! uncurry (+) $ doFoldl (lengthA *** sumA) (0, 0) [1..size] On the other hand if I use arrow syntax: > myArr :: Num a => FoldlArrow a (a, a) (a, a) > myArr = proc (l, s) -> do e <- element -< () > returnA -< (l + 1, s + e) It starts consuming memory as well. Somewhere is lazy passing of value but I cannot find where Regards PS. As it is probably out of scope and topic of beginners mailing list I'm CC'ing cafe (possibly beginners should be dropped). [1] It can be extended to work on other arrows as well - not only (->).

2 3

ANNOUNCE: gt-tools-0.1.4
by Sergei Trofimovich 07 May '10

07 May '10

Hey, -cafe! I'm glad to announce release of Google Translate Service[1] Tools[2]. Currently it's a command line tool 'gtc' (queen of this post) and 'gtg' (more a hello-world type gtk2hs user). So let's play with that beast a little! PREREQUISITES * ghs-6.12+ - gt-tools relies on proper locale detection for passed command line arguments INSTALLATION $ cabal update Downloading the latest package list from hackage.haskell.org $ cabal install gt-tools -f-gui # avoid 'gtg' for now ... Done! It was simple. USE $ gtc # sniff the beast Usage: Standart mode: gtc 'fromLang' 'toLang' 'text' Interactive mode: gtc -i 'fromLang' 'toLang' Where 'Lang' is 2 sumbols code, i.e. be, en, ru. Commandline keys: --interactive or -i - interactive mode. --help or -h - to see this help. --list or -l - list of available languages. $ gtc -l # language list # a lot $ gtc en no "good evening" Source language: en Translation: god kveld Original text: good evening Translit: Whoa! $ gtc -i en fi # interactive mode! > # waits for prompt! > who are you? -------------------------------------------------------------------------------- Source language: en Translation: kuka sinä olet? Original text: who are you? Translit: # leading exclamation'!' mark allows you to invert source and destination language > !päivää -------------------------------------------------------------------------------- Source language: fi Translation: days Original text: päivää Translit: interjection: good afternoon Integration with other translating services is considered by author as well as user interface should be polished a little more too. Any feedback is appreciated! /me wonders if Miss lambdabot might like to have such functionality. What do you think? Have fun! [1] It's a simple translation service: http://translate.google.com/ [2] "home page" for gt-tools: http://github.com/styx/gtc -- Sergei

4 3

Database connection pool
by Michael Snoyman 07 May '10

07 May '10

Hi all, I would like to pool my database connections in an application I'm writing, and so far haven't found any prior art on the subject (besides this[1]). I was wondering if: * There's a package somewhere that does this * Others have implemented it and have suggestions * There's some big gotchas The stackoverflow that I linked to gives one approach to this that seems valid. My original thought on an implementation was to use a Chan. That way: * When a connection is released, is goes to the end of the pool, so connections get used evenly (not sure if this actually matters in practice). * If a process dies and for some reason doesn't return its connection, there shouldn't be a problem. Of course, I'd use bracket, so such a situation should never arise anyway. * Keep the coding against MVar to a minimum to avoid normal multi-threaded pitfalls. I'm using HDBC in this case, but I don't think I'll be using any HDBC-specific code. Thanks, Michael [1] http://stackoverflow.com/questions/1141677/concurrent-db-connection-pool-in…

3 7

Mirror repository for gtk2hs
by Andy Stewart 07 May '10

07 May '10

Hi all, Because code.haskell.org is unstable recently, and many gtk2hs users can't get code from http://code.haskell.org/gtk2hs So i build a mirror repository at http://patch-tag.com/r/AndyStewart/gtk2hs-sync-mirror/home You can access this mirror repository when code.haskell.org down. This mirror repository synchronous http://code.haskell.org/gtk2hs Hourly. Because mirror repository just read-only, please send any patches to gtk2hs main repository (http://code.haskell.org/gtk2hs) I need thanks "Thomas Hartman" (the administor of patch-tag.com) i can't build this mirror repository if haven't his help. Thanks Thomas! Enjoy, everybody! -- Andy

1 0

"Dynamic" CSV Parsing - Parsec
by Günther Schmidt 07 May '10

07 May '10

Hello, I'm trying to build a CSV parser that can dynamically assemble a parser from the values of the first line. As the most simple example the parse of the first line would return a parser with which subsequent lines would then be parsed. This parser would, for instance, only parse lines with the exact number of columns as found in the first line. Where I eventually want to go is a bit more complicated than this, but for now I'd be grateful for suggestions on how to go about the simple case. Best regards Günther

2 1

Fwd: Error instaling Happstack on Windows - cabal bug?
by Flavio Botelho 06 May '10

06 May '10

Trying to install on Windows, i am having a strange problem (may actually be some cabal bug?), More specifically trying to install happstack-util [13 of 19] Compiling Happstack.Crypto.MD5 ( src\Happstack\Crypto\MD5.hs, dist\bu ild\Happstack\Crypto\MD5.o ) cabal: Error: some packages failed to install: happstack-util-0.5.0 failed during the building phase. The exception was: ExitFailure (-1073741819) A Windows prompt shows problems (Application not properly initialized) with a "perl.exe" program. Does cabal use perl (that's completely unexpected for me)? I did install perl and put it on the path and that didn't solve it (maybe there is a configuration parameter to point to perl's executable?).... Thanks, Flavio

4 4

DTP10 Call for Participation
by Conor McBride 06 May '10

06 May '10

Remember, Haskell is the world's most popular dependently typed functional programming language... (s:S)*(p:P s)->(s:S)*(p:P s)->(s:S)*(p:P s)->(s:S)*(p:P s)->(s:S)*(p:P s)-> DTP 2010 --- Call for Participation EARLY REGISTRATION ENDS 17 MAY 2010 Workshop on DEPENDENTLY TYPED PROGRAMMING Edinburgh, Scotland, 9&10 July 2010 (a FLoC workshop, affiliated with LICS) http://sneezy.cs.nott.ac.uk/darcs/dtp10/ (s:S)*(p:P s)->(s:S)*(p:P s)->(s:S)*(p:P s)->(s:S)*(p:P s)->(s:S)*(p:P s)-> Roll up! Roll up! Register early, register often! http://www.floc-conference.org/registration.html Attendance at DTP10 can be yours at a BARGAIN price if you register BEFORE 17 MAY 2010. The preliminary programme for DTP10 is here: http://sneezy.cs.nott.ac.uk/darcs/dtp10/programme.html Invited Talks: Ana Bove, Chalmers University, "10 Years of Partiality and General Recursion" Matthieu Sozeau, Harvard University, "Elaborations in Type Theory" Contributed Talks: Edwin Brady, "Practical, efficient programming with dependent types" James Caldwell, "Extracting Monadic Programs form Proofs", (joint work with Josef Pohl) Adam Chlipala, "Generating Pieces of Web Applications with Type-Level Programming" Nils Anders Danielsson, TBA Larry Diehl, "Unit & integration test composition via lemmas" Makoto Hamana, "Another Initial Algebra Semantics of Inductive Families for Programming" Hugo Herbelin, "A sequent calculus presentation of the Calculus of Inductive Constructions" (joint work with Jeffrey Sarnat, Vincent Siles) Karim Kariso, "Integrating Agda and Automated Theorem Proving Techniques" (joint work with Anton Setzer) Dan Licata, "Security-Typed Programming within Dependently Typed Programming" (joint work with Jamie Morgenstern) Ulf Norell, TBA Carsten Schuermann, "The HOL-Nuprl connection in Delphin", (joint work with Adam Poswolsky) Anton Setzer, "Coalgebras in dependent type theory" Antonis Stampoulis, "VeriML: Type-safe computation of logical terms inside a language with effects" Tarmo Uustalu, TBA Sean Wilson, "Supporting Dependently Typed Functional Programming with Proof Automation and Testing" If, by some chance, you are interested in talking at DTP10, please do get in touch. Space in the programme is now very tight, but we remain open to proposals. See you in Edinburgh in July! Thorsten and Conor _______________________________________________ Agda mailing list Agda(a)lists.chalmers.se https://lists.chalmers.se/mailman/listinfo/agda

2 2

Interest in a Mathematics & AI strike force ?
by Alp Mestanogullari 06 May '10

06 May '10

Hello -cafe, When I started learning Haskell, I saw the AI page [1] which aimed at creating a sound, uniform and handy framework for AI programming in Haskell. I added my name on it and thought a bit about it. I even wrote a first version of HNN [2], a neural network library, quite early in my Haskell days. I found that idea to be great but did not see any actual effort around this. So, I'm now thinking again about that and even enlarging it to mathematics & AI. Thus, I would like to have an idea of the number of people interested in being involved in such an effort. There are several tools out there on hackage but they aren't that much uniform and neither play nicely together. I'm pretty convinced this could be improved and as a Mathematics student I'm highly interested in that. If enough people are interested, we could for example set up a mailing list and a trac to organize the effort and then people could just discuss and write Haskell modules when time permits. Any comment, idea, reaction, interest ? [1] http://www.haskell.org/haskellwiki/AI [2] http://www.haskell.org/haskellwiki/HNN -- Alp Mestanogullari http://alpmestan.wordpress.com/ http://alp.developpez.com/

6 16

Mystery Parse Error in LHS file
by Iæfai 06 May '10

06 May '10

I have an lhs file, 'starsystem.lhs' that is not compiling because of a parse error. Specifically: $ cabal build Preprocessing executables for starsystem-0.1... Building starsystem-0.1... [1 of 4] Compiling Render ( Render.lhs, dist/build/ starsystem/starsystem-tmp/Render.o ) [2 of 4] Compiling Passive ( Passive.lhs, dist/build/ starsystem/starsystem-tmp/Passive.o ) [3 of 4] Compiling Active ( Active.lhs, dist/build/ starsystem/starsystem-tmp/Active.o ) [4 of 4] Compiling Main ( starsystem.lhs, dist/build/ starsystem/starsystem-tmp/Main.o ) starsystem.lhs:29:0: parse error on input `main' The location is at the first instance of main: > data Action = Action (IO Action) The main program is mostly book-keeping such as initializing OpenGL and GLFW, creating window, setting up viewport, etc. \begin{code} main :: IO () -- ** line 29 main = do c_unbundled GLFW.initialize -- open window I cannot figure out what the problem here is. I have tried it in a regular hs file and it works. I have posted full source here: http://hpaste.org/fastcgi/hpaste.fcgi/view?id=25352#a25352 Any thoughts would be welcome.

3 5

ANNOUNCE: wai-handler-fastcgi 0.0.0
by Michael Snoyman 06 May '10

06 May '10

Hi all, I finally scratched an itch I've had for a while, and put together wai-handler-fastcgi. It is built on top of Dan Knapp's wonderful direct-fastcgi[1] package, so it is free from dependencies on C libraries. This package allows WAI[2] applications to be run on a FastCGI-supporting server. The WAI, for those not familiar, is a standard interface between web servers and web applications, allowing you to write your web applications once and deploy in multiple manners. Michael [1] http://hackage.haskell.org/package/direct-fastcgi [2] http://hackage.haskell.org/package/wai

1 0

forall (What does it do)
by John Creighton 06 May '10

06 May '10

I've seen forall used in a few places related to Haskell. I know their is a type extension call "explicit forall" but by the way it is documnted in some places, the documentation makes it sound like it does nothing usefull. However on Page 27 of Haskell's overlooked object system: >We define an existential envelope for shape data. >data OpaqueShape = forall x. Shape x => HideShape x >“Opaque shapes are (still) shapes.” Hence, a Shape instance: >nstance Shape OpaqueShape >where > readShape f (HideShape x) = readShape f x > writeShape f (HideShape x) = HideShape $ writeShape f x > draw (HideShape x) = draw x >When building the scribble list, we place shapes in the envelope. >let scribble = [ HideShape (rectangle 10 20 5 6) >, HideShape (circle 15 25 8) " It seems that forall can be used as a form of up casting. That is if we have a parametric type we can treat the parametric types as homogeneous: The paper did not seem to regard this feature too highly: >By contrast, the explicit constraint for the existential envelope cannot be eliminated. >Admittedly, the loss of type inference is a nuance in this specific example. In >general, however, this weakness of existentials is quite annoying. It is intellectually >dissatisfying since type inference is one of the added values of an (extended) Hindley/ >Milner type system, when compared to mainstream OO languages. Worse than >that, the kind of constraints in the example are not necessary in mainstream OO >languages (without type inference), because these constraints deal with subtyping, >which is normally implicit. >We do not use existentials in OOHaskell http://homepages.cwi.nl/~ralf/OOHaskell/paper.pdf but can't we just use the read shape function above if we want to be able to infer based on types. I understand that haskers tend to like things well typed but surely their are times when we would like to be able to flatten the type of our data somewhat. On another note, I also wonder how forall relates to type families. I know their are certain restrictions on where we can use type families in haskell. I wonder if we can get around these somewhat using forall.

2 2

lhs2TeX - lhs2TeX.fmt missing
by Ozgur Akgun 06 May '10

06 May '10

Hi all, I am trying to get lhs2TeX to work. I installed the package using cabal, and now I try to run it on a very simple *.lhs file. But it blames me and says user error, cannot find lhs2TeX.fmt: lhs2TeX: user error (File `lhs2TeX.fmt' not found in search path: And, when I check the search path, there really is no lhs2TeX.fmt file. Should I download it separately or something like that? PS: Thanks for the great package to the authors! Best, Ozgur Akgun

5 6

A query to users of graphviz
by Ivan Lazar Miljenovic 06 May '10

06 May '10

For anyone who uses my graphviz library: who uses the graphNodes and functions? Note that this also applies to anyone who uses augmentation functions as well (more specifically, when they define global attributes for use with an augmentation function). If so, would it be a problem if I switched it from returning a list to returning a Set? The rationale being: Dot allows nodes to be "defined" twice (to specify different attributes) and also that some nodes might only be defined because they appear in an edge. Here is an example containing both problems: ,---- | digraph { | a [label="hello"] | subgraph { rank=same a b } | b -> c | } `---- Currently, graphNodes ignores the first problem and doesn't even consider the second (though I thought I had implemented support for this); however, I myself am needing this functionality soon, hence why I'm considering this (note that this isn't a problem for graphEdges, as multiple edges are OK). Currently, graphNodes doesn't take this into account (for edges, multiple definitions means the edge is drawn multiple times, so this isn't a problem for graphEdges); as such, graphviz should really (but doesn't) use nub on this. A better solution would be to use Sets and merge the results by merging together the applied attributes (and adding the global attributes for nodes at the same time). However, the problem with using Sets is that it means node type has to be an instance of Ord. I can't think of many types which you might use as a node type that _wouldn't_ have an applicable Ord instance, but I suppose it's possible. As such, there are three possible options: 1) Leave as-is (just get it working properly with hack-y nubs, etc.): this will be extremely inefficient. 2) Use Sets (or a Map n Attributes, which will make merging Attributes easier); this will require an Ord constraint on the label type for the DotRepr class (i.e. it will be needed _everywhere_). It may also not be as lazy as the current implementation, and the node order will be changed. 3) As with 2), but convert the result back to a list so that the current interface remains unchanged (with the added benefit of not requiring a major version bump). With whichever option is chosen, then there is also the consideration of whether any global attributes should be considered and added to the list of attributes (this goes for graphEdges as well). I personally am thinking of going with 3) and adding in global attributes as well. What do other people think? -- Ivan Lazar Miljenovic Ivan.Miljenovic(a)gmail.com IvanMiljenovic.wordpress.com

1 0

Mixing mtl- and transformers-using libraries
by Joachim Breitner 06 May '10

06 May '10

Hi, I have a rather large project here that uses standard mtl stuff. This program now wants to use salvia, which in turn uses the transformers library. The result is this: X.hs:163:29: Overlapping instances for Monad (Either String) arising from a use of `neuePersonCommit' at X.hs:163:29-48 Matching instances: instance (transformers-0.1.4.0:Control.Monad.Trans.Error.Error e) => Monad (Either e) -- Defined in transformers-0.1.4.0:Control.Monad.Trans.Error instance (Control.Monad.Error.Class.Error e) => Monad (Either e) -- Defined in Control.Monad.Error Is there any way of using salvia without having ghc consider instances from transformers? Thanks, Joachim -- Joachim "nomeata" Breitner mail: mail(a)joachim-breitner.de | ICQ# 74513189 | GPG-Key: 4743206C JID: nomeata(a)joachim-breitner.de | http://www.joachim-breitner.de/ Debian Developer: nomeata(a)debian.org

3 2

Is anyone using Haddock's support for frames?
by David Waern 06 May '10

06 May '10

Hi Since version 2.4.0 Haddock has generated HTML output that uses frames (index-frames.html) in addition to the normal output. We'd like to deprecate this feature unless there is a significant amount of users. The reason is two-fold: * We probably want to replace the frames with something more modern (like a sidebar on the same page) in the future * We are rewriting the HTML backend and it would be nice to avoid unnecessary work So if you're using this feature and want to keep it, please speak up! cc:ing cvs-ghc@ in case they have any users of the frames due to the size of the GHC code base. (This might have been the the original motivation for the feature). Thanks, David

5 7

Why cabal forces reinstallation of happstack-utils?
by Maciej Piechotka 06 May '10

06 May '10

I try to configure happstack with parsec 3.1. It seems to fail due to cabal: I installed happstack-util editing happstack-util.cabal by hand: % grep parsec ~/.ghc/x86_64-linux-6.12.2/package.conf.d/happstack-util-0.5.0-6e27d5d3ba1c07f259d463ee3036c92b.conf parsec-3.1.0-5842597f447f82b210228391f371cff1 % cabal install --constraint 'parsec > 3' --dry-run happstack Resolving dependencies... cabal: cannot configure happstack-util-0.5.0. It requires parsec <3 For the dependency on parsec <3 there are these packages: parsec-2.0, parsec-2.1.0.0 and parsec-2.1.0.1. However none of them are available. parsec-2.0 was excluded because of the top level dependency parsec >3 parsec-2.1.0.0 was excluded because of the top level dependency parsec >3 parsec-2.1.0.1 was excluded because of the top level dependency parsec >3 % ghc-pkg list | grep happstack-util happstack-util-0.5.0 Regards

6 14

class Arbitrary in quickcheck
by Tim Docker 06 May '10

06 May '10

I've notice a behaviour of quickcheck that is unexpected to me. With this code: import Test.QuickCheck main = check myconfig ((\v -> v == v) :: (Maybe Double,Maybe Double) -> Bool) myconfig = defaultConfig{configMaxTest=100000, configEvery = \n args -> show n ++ ":\n" ++ unlines args} I am relying on the default Arbitrary instances to generate pairs of maybe doubles. But to my surprise, all of the pairs have these patterns: (Just _ Just _) (Nothing, Nothing) I never see patterns: (Just _, Nothing) (Nothing, Just _) Why is this the case? Thanks, Tim

3 6

Haskell XML Parsers
by R Senington 06 May '10

06 May '10

Dear all, I have been looking at using XML for a little program I have been writing. The file I am currently trying to load is about 9MB, and I have now tried to use HaXml and HST. Without any of my own code, just a simple call to the basic parsers, they both use huge amount of memory. HST is the worst and about 7GB and climbing. HaXml uses 1.3Gb. The code I am using is HST xml <- readFile file_name_here;k<-runX (parseXmlDocument True) xml;print k and for HaXml x<-readFile file_name_here let (Document _ _ e _) = xmlParse "t" x let t = myFilter $ CElem e print $ length t I have seen on previous posts to the cafe that other people have run into this problem with HST. Is this a general problem with XML in Haskell (I know that XML parsing is a slow and bulky process but this seems excessive)? Is there a known solution? Does anyone have any advice? Cheers RS

4 4

posting UTF8 data with Curl library
by Eugeny N Dzhurinsky 06 May '10

06 May '10

Hello! I need to submit data to HTTP server using UTF8 encoding. I found out that libcurl for haskell can work with Data.ByteString - but it seems not able to work with Data.ByteString.UTF8. Can you please advice, how do I convert Data.ByteString.UTF8 into Data.ByteString and visa versa? Thank you in advance! -- Eugene Dzhurinsky

3 2

macosx-app for wxHaskell
by Bill Atkins 06 May '10

06 May '10

In order to run apps built with wxHaskell on OS X, you're supposed to wrap the binary using a script called macosx-app ( http://www.haskell.org/haskellwiki/WxHaskell/MacOS_X) Unfortunately, after running "cabal install wx" this file is not in "~/.cabal" - it's also not in the wxHaskell source. Does anyone know where I can find it? Thanks, Bill

2 1

Would it be evil to add "deriving Typeable" to newtype Q?
by Leonel Fonseca 06 May '10

06 May '10

Hi everybody, Is it reasonable to add "deriving Typeable" to newtype Q? In case you wonder why I want to do that, it is because I've constructed a Q [Dec] inside a monad, I want to extract it from the monad (via runIO) and the monad has constraint "Typeable" over this parameter. I've also tried to write the Typeable instance in my own module (not Language.Haskell.TH.Syntax). But I've got no luck since the Q type constructor is exported but the data constructor is not. Thanks. -- Leonel Fonseca.

2 1

ANN: Holumbus-Distribution 0.1.0 - distributed data structures
by Stefan Schmidt 05 May '10

05 May '10

Hello, after a long year, I've finally managed to upload the new version of Holumbus-Distribution to Hackage. Holumbus-Distribution offers distributed data structures like Chan, MVar or functions. These data types can be used for inter-process communication. With the help of this library it is possible to build Erlang-Style mailboxes for an easy implementation of distributed systems in Haskell. You can find the library on Hackage at: http://hackage.haskell.org/package/Holumbus-Distribution And the project homepage at: http://www.holumbus.org (for the current documentation please refer to the example programs in the source bundle, the Wiki will be updated soon) The following data structures are implemented so far: * distributed Chan (DChan) like the Chan datatype, but it allows the writing and reading from other programs * distributed MVar (DMVar) like the MVar data type, but the content of the MVar is shared among multiple programs. * distributed Functions (DFunction) an easy way to do remote procedure calls, just like haxr * distributed Values (DValue) a variable which could only be written once and which could easily read by other programs * distributed Streams and Ports (DStream, DPort) just like the DChan, but this time, you are only allowed to read from the channel from one program This library was completely rewritten between version 0.0.1 and 0.1.0. The old packages can be found under Holumbus.Network.* while the new ones are located at Holumbus.Distribution.*. The old packages will be removed in future versions (0.2.0). This library itself is independent from other Holumbus libraries although it contains some modules which are only to be used by the other Holumbus libraries but this may change in the near future. Regards, Stefan

1 0