- Haskell-Cafe - Haskell.org

On the Meaning of Haskell 8
by John D. Earle 03 Dec '09

03 Dec '09

This is a continuation of what I wrote on the Haskell Prime mailing list which accounts for why this is the eighth in the series. These are lecture notes and if anyone has not already noticed from the Haskell Prime mailing list. I am giving a lecture. Type aliases allow you to extend the type system beyond the capabilities of the language. For example, natural numbers are not part of the type system, but you can say that a number has the natural number type by using a type alias. Ensuring that a number is indeed a natural number is your problem. The compiler will only ensure that it is at least an integer. There are a variety of instances where type aliases are useful. A prominent application in computer programming languages is to establish a form verses function relationship. Types in functional languages are formal types and not functional types ironically. A common example of this is to define an address type alias where you declare it as a string. Yes, it is a string formally, but it does not function as such. Its semantics are more specific and so it is not merely a string. How it behaves in context will differ and be more specific than a string. To quote Dune, "There is a place terrifying to us, women, that we cannot go." The compiler cannot go there, but it can verify that it is at least a string. The form verses function distinction is especially useful when unifying against a tuple. Functional types can in large measure supplant labels. When you know its functional type you will usually know enough to know what it is referring to. When all you know is its formal type it can be touch and go. It is possible as the programmer to develop a set of rules beyond those that ensure that the address type is a string that for the benefit of the programmer describe where it is grammatically correct for variables of the address type may appear to which the compiler is wholly ignorant. When defining tuple types, type aliases were intended to replace labels. The result is more compact. The down side is that to access members unification must be employed. In the C language type aliases can be created using either the C preprocessor or the typedef construct. Consequently, this notion of type alias does not apply exclusively to functional languages.

1 0

Haddock Secrets?
by Gregory Crosswhite 03 Dec '09

03 Dec '09

Hey everyone, Is there some secret to getting Haddock to work with literate Haskell sources that I am missing? For example, when I download Takusen and type cabal configure cabal haddock It produces HTML files complete with a table of contents, but with all of the documentation stripped out. Oddly, I know that it is *possible* to process the literate sources into documentation because it appears on Hackage! I am doing this on OSX Snow Leopard with GHC 6.10.4 and Haddock 2.5. Thanks, Greg

2 1

Re: [Haskell] Interesting experiences of test automation in Haskell? Automation of Software Test 2010
by Tom Hawkins 03 Dec '09

03 Dec '09

On Fri, Nov 27, 2009 at 9:30 PM, John Hughes <rjmh(a)chalmers.se> wrote: > This is a "heads up" about a workshop on test automation that I just joined > the programme committee of. Automation of Software Test will be co-located > with ICSE in Cape Town in May--the workshop home page is here: > > http://www.cs.allegheny.edu/ast2010/ > > Tools like QuickCheck, SmallCheck and HUnit fit the call for papers > perfectly. So if you're doing some interesting automated testing in Haskell, > why not submit a paper about it, and show a new audience what the Haskell > community is up to? Both research papers and case studies are welcome, and > the latter can even be in the form of a presentation of up to 15 slides--so > there's no excuse for not putting something together! Recently we added a unit testing framework to Atom: a haskell DSL for hard realtime embedded applications. The framework hooks into Atom's assertion and functional coverage statements and provides some basic randomization utilities for automatic test generation. We've used this framework to test the embedded software of Eaton's parallel hydraulic hybrid application (HLA). One nice feature we've incorporated for our application is in the event of an assertion violation, simulation data is formated such that we can use the same analysis and debug tools that we use for real vehicle telemetry. I don't think we can make it to Cape Town, but I'll see if one of us can write up a case study. (Shruti, Tejas, Jake, any interest?) Here's an example test log from Language.Atom.Unit: pass: processTest1 cycles = 1000 pass: dualProcessTest cycles = 1000 pass: processNoDeactivationGuard cycles = 1000 pass: processNoActivationGuard cycles = 1000 pass: processDuelingProcesses cycles = 1000 pass: arbiterTest1 cycles = 1000 pass: arbiterTest2 cycles = 1000 pass: arbiterTest3 cycles = 1000 pass: arbiterTest8 cycles = 1000 pass: arbiterTest10 cycles = 1000 pass: absDisengagementTest cycles = 35000 pass: cruiseControlDisengagementTest cycles = 35000 pass: engineSpeedDisengagementTest cycles = 35000 pass: pumpSpeedDisengagementTest cycles = 35000 pass: ignitionDisengagementTest cycles = 35000 pass: selectedGearDisengagementTest cycles = 35000 pass: currentGearDisengagementTest cycles = 35000 pass: modeDisengagementTest cycles = 35000 pass: commissioningProcessDisengagementTest cycles = 35000 pass: canEEC1DisengagementTest cycles = 35000 pass: canEEC2DisengagementTest cycles = 35000 pass: canETC1DisengagementTest cycles = 35000 pass: canETC2DisengagementTest cycles = 35000 pass: canEBCDisengagementTest cycles = 35000 pass: canECCDisengagementTest cycles = 35000 pass: driverisolationFaultDisengagementTest cycles = 35000 pass: drivermodeDrainFaultDisengagementTest cycles = 35000 pass: drivermodeFillFaultDisengagementTest cycles = 35000 pass: drivermainBypassFaultDisengagementTest cycles = 35000 pass: driverchargeBypassFaultDisengagementTest cycles = 35000 pass: driverclutch1FaultDisengagementTest cycles = 35000 pass: driverclutch2FaultDisengagementTest cycles = 35000 pass: driverclutch3FaultDisengagementTest cycles = 35000 pass: driverpumpFaultDisengagementTest cycles = 35000 pass: drivermotorFaultDisengagementTest cycles = 35000 pass: coolerFanOnOffTest cycles = 40000 pass: stateTransitionTest cycles = 75000 pass: swashCmdInBootstrapTest cycles = 40000 pass: brakePressureAboveRangeAndTxSpeedDisengagementTest cycles = 35000 pass: brakePressureAboveRangeDisengagementTest cycles = 35000 pass: accumulatorPressureBelowRangeDisengagementTest cycles = 35000 pass: accumulatorPressureAboveRangeDisengagementTest cycles = 35000 pass: sawsh1BelowRangeDisengagementTest cycles = 35000 pass: swash1AboveRangeDisengagementTest cycles = 35000 pass: swash2BelowRangeDisengagementTest cycles = 35000 pass: swash2AboveRangeDisengagementTest cycles = 35000 pass: swashDiffOnBoundaryDisengagementTest cycles = 35000 pass: swashDiffAboveRangeDisengagementTest cycles = 35000 pass: pumpCaseOilTempCAboveRangeDisengagementTest cycles = 35000 pass: pumpCaseOilTempCBelowRangeDisengagementTest cycles = 35000 pass: pumpCaseOilTempCBoundaryDisengagementTest cycles = 35000 pass: reservoirOilTempCAboveRangeDisengagementTest cycles = 35000 pass: reservoirOilTempCBelowRangeDisengagementTest cycles = 35000 pass: reservoirOilTempCBoundaryDisengagementTest cycles = 35000 pass: tempSensorMiscompareAboveRangeDisengagementTest cycles = 35000 pass: tempSensorMiscompareBoundaryDisengagementTest cycles = 35000 pass: reservoirOilTempCTooExtremeDisengagementTest cycles = 35000 pass: reservoirOilTempCTooExtremeBoundaryDisengagementTest cycles = 35000 pass: reservoirOilLevelGalLDisengagementTest cycles = 35000 pass: reservoirOilLevelGalHDisengagementTest cycles = 35000 pass: pumpSpeedAboveRangeLDisengagementTest cycles = 35000 pass: pumpSpeedAboveRangeHDisengagementTest cycles = 35000 pass: accumOverPressureBoundaryDisengagementTest cycles = 35000 pass: accumOverPressureAboveRangeDisengagementTest cycles = 35000 pass: filterPressureBelowRangeDisengagementTest cycles = 35000 pass: filterPressureAboveRangeDisengagementTest cycles = 35000 pass: 5VSupplyBelowRangeDisengagementTest cycles = 35000 pass: 5VSupplyAboveRangeDisengagementTest cycles = 35000 pass: 12VSupplyBelowRangeDisengagementTest cycles = 35000 pass: 12VSupplyAboveRangeDisengagementTest cycles = 35000 pass: stuckEngagedDisengagementTest cycles = 35000 pass: stuckOilLevelDisengagementTest cycles = 35000 pass: clutchdisengagementFaultTest cycles = 40000 pass: clutchSensorMiscompareFault cycles = 40000 pass: clutchSensorMiscompareFaultAbovePumpSpeedandabsRangeTest cycles = 40000 pass: clutchSensorMiscompareFaultAbovePumpSpeedandBelowabsRangeTest cycles = 40000 pass: clutchSensorMiscompareFaultBoundaryTest cycles = 40000 pass: clutchSensorMiscompareFaultAboveRangeTest cycles = 40000 pass: unexpectedDisengagementTest cycles = 40000 pass: failedEngagementAttemptsTest cycles = 40000 pass: fanOffTest cycles = 30000 pass: evuTest cycles = 30000 pass: swashCalTest cycles = 40000 pass: seedKey32Test1 cycles = 100 pass: seedKey32Test2 cycles = 100 pass: seedKey32Test3 cycles = 100 pass: seedKey32Test4 cycles = 100 pass: seedKey32Test5 cycles = 100 pass: seedKey32Test6 cycles = 100 pass: seedKey32Test7 cycles = 100 Total Passing Tests : 90 / 90 Total Simulation Cycles : 2650700 Total Function Coverage : 102 / 102 GREEN LIGHT > > So how about it? It would be great to see some Haskell papers at the > workshop! Deadline 20 January. > > John Hughes > > PS Check out the ICSE web site for information on the location: > http://www.sbs.co.za/ICSE2010/ > > _______________________________________________ > Haskell mailing list > Haskell(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell > >

1 0

Existencial Types
by rodrigo.bonifacio 02 Dec '09

02 Dec '09

4 6

O(n) algorithm for determining subset
by Michael Mossey 02 Dec '09

02 Dec '09

Can someone tell me if this is correct. I'm guessing that if I represent two sets of integers by Word32 (where ith bit set means i is in the set), then an algorithm to determine if x is a subset of y would go something like (y - x) == (y `exclusiveOr` x)

7 11

Type synonym family inside type class
by Martijn van Steenbergen 02 Dec '09

02 Dec '09

Hello, I have a type family and a type class: > type family ErrorAlg (f :: (* -> *) -> * -> *) e ix :: * > class MkErrorAlg f where > mkErrorAlg :: ErrorAlg f e a -> f (K0 a) ix -> Either e a Instances for these two really go hand in hand, so I thought I would move the type family into the type class. However, this causes GHC to complain about the type variables that are bound on the LHS of the type synonym: > Not in scope: type variable `e' > Not in scope: type variable `ix' In function types, using new type variables (i.e. type variables not bound by the type class header) implicitly means universal quantification over these variables. Why is this disallowed in type families inside type classes? Thanks, Martijn.

3 3

Ord k => Data.Set.Set k -> (k->v) -> Data.Map.Map k v
by Matthias Görgens 02 Dec '09

02 Dec '09

I feel that Data.Set and Data.Map should be working together more closely. For example you can already get the keyset of a Map, but the `other way' is not built-in. I mean a function with a signature like Ord k => Data.Set.Set k -> (k->v) -> Data.Map.Map k v You can implement it in O(n): > assoc :: (a->b) -> [a] -> [(a,b)] > assoc f = map (\x -> (x, f x)) > mapToMap :: Ord k => (k -> v) -> Data.Set.Set k -> Data.Map.Map k v > mapToMap f = Data.Map.fromAscList . assoc f . Data.Set.toAscList The name assoc alludes to the assoc-lists of Lisp.

1 0

Building network package on Windows
by Iavor Diatchki 02 Dec '09

02 Dec '09

Hi, I have been trying to build the package "network" from hackage (version 2.2.1.3) on Windows Vista, and I could really use some help. Building on the command line, or under cygwin completely failed (command line due to cabal not being able to execute something---possibly configure---although it would not say; cygwin first due to lack of gcc, which is tested but, apparently, the outcome ignored, and after gcc was installed some incompatibility with the header files which were detected but reported unusable). I managed to build the library under MinGW with msys without serious obstacles. I can also build my package against the result and all is well. Unfortunately, if I try to use my package to build an executable application I get a linker error, reporting a missing symbol during linking: C:\Users\diatchki\AppData\Roaming\cabal\network-2.2.1.3\ghc-6.10.3/libHSnetwork-2.2.1.3.a(Socket.o):fake:(.text+0xb014): undefined reference to `getnameinfo' collect2: ld returned 1 exit status Now, "getnameinfo" is present in the header files, and it is also defined in the library ws2_32.a which is being passed to GHC so I am not sure what is going on. Any ideas? Searching the web suggests that the problem may be somehow related to the standard calling conventions but I don't really understand. Also, if I understand correctly, this functionality is related to IPv6 support, which I do not need at the moment, so it would be great if it could be easily disabled in some way. Any ideas, suggestion, workarounds, etc. would be greatly appreciated, -Iavor PS: I am using GHC 6.10.3

5 9

KiCS (Curry to Haskell interpreter) problem
by Bernd Brassel 02 Dec '09

02 Dec '09

> I am playing around with KiCS and I have a strange problem, when I > evaluate a goal the variable bindings are not displayed, I see only > the value of the expression. The idea is, that you decide yourself which bindings you want to see. For example, you write (let x free in (not x,x)) to get (True,False) More? (False,True) More? No more solutions > I would have contacted the author but his email in not in the haskell > cabal file. Thanks for the hint. I have added my mail now and will do another update in a few weeks. Meanwhile, anyone interested in the system can contact me directly under "bbr at informatik.uni-kiel.de" Thanks for your interest in kics! Bernd

1 0

buildExpressionParser prefix recursion
by Warren Harris 02 Dec '09

02 Dec '09

I was wondering why Parsec's buildExpressionParser doesn't allow prefix expressions to be handled recursively, e.g. given a prefix "!" operation, it seems that "!!a" could parse without requiring parentheses ("!(!a)"). Is there an easy way to extend it? (I have a rich expression grammar I'd like to parse which includes several prefix forms, and it would be a shame to have to abandon buildExpressionParser since it is so concise and convenient.) Thanks, Warren

1 0

university courses on type families/GADTs?
by Tom Schrijvers 02 Dec '09

02 Dec '09

Hello Haskell Cafe, I was wondering whether there are any universities that teach about Haskell type families or GADTs? Thanks, Tom -- Tom Schrijvers Department of Computer Science K.U. Leuven Celestijnenlaan 200A B-3001 Heverlee Belgium tel: +32 16 327544 e-mail: tom.schrijvers(a)cs.kuleuven.be url: http://www.cs.kuleuven.be/~toms/ Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm

3 2

Patches and forks for GHC 6.12
by Robin Green 02 Dec '09

02 Dec '09

Some packages will need modifications to build or work with GHC 6.12 (in some cases, just modifications to the .cabal file). I've created this wiki page to track work people have done on that which hasn't yet been included into official packages or repositories: http://haskell.org/haskellwiki/Patches_and_forks_for_GHC_6.12 Please add any ones you know of (or just email me and I'll add them for you). -- Robin

1 0

Trying to sort out multiparameter type classes and their instances
by Jeremy Fitzhardinge 02 Dec '09

02 Dec '09

I'm playing around with some types to represent a game board (like Go, Chess, Scrabble, etc). I'm using a type class to represent the basic Board interface, so I can change the implementation freely: class Board b pos piece where -- Update board with piece played at pos play :: b pos piece -> pos -> piece -> b pos piece -- Query pos to get piece (Nothing if off board) at :: b pos piece -> pos -> Maybe piece -- Empty board empty :: b pos piece and a Position on the board is represented thus: class Position p where up :: p -> p down :: p -> p left :: p -> p right :: p -> p With a concrete implementation using a tuple: instance (Enum c,Enum r) => Position (c,r) where up = second pred down = second succ left = first pred right = first succ My initial Board is a function: position -> Maybe piece, but I'm having a hard time writing the instance for it. My first attempt is: instance Board (pos -> Maybe piece) pos piece where empty = \_ -> Nothing at = ($) play b pos piece = move where move pos' | pos' == pos = Just piece | otherwise = b pos' but ghci complains: board.hs:34:15: Kind mis-match Expected kind `* -> * -> *', but `pos -> Maybe piece' has kind `*' In the instance declaration for `Board (pos -> Maybe piece) pos piece' Playing around with parentheses on the instance line got various similar messages, but I couldn't get anything to work. What am I missing here? One thing that strikes me is that "Board (pos -> Maybe piece) pos piece" has a lot of redundancy, and I'm wondering if I'm defining the Board type class wrong in the first place. Given that the "b" type parameter necessarily defines the position and pieces, I tried using dependent types: class Board b | b -> pos, b -> piece where ... but this complains "Not in scope: type variable `pos'/`piece'", so I tried sprinkling some existential types around as well: class (forall pos piece. b pos piece) => Board b | b -> pos, b -> piece where ... (with and without the dependent types) but this just complains "malformed class assertion". As you can probably tell, I'm just thrashing around trying things without really understanding what's going on here, so I'm hoping someone can give me some useful pointers. J

2 4

Re: [Haskell-cafe] university courses on type families/GADTs?
by Manuel M T Chakravarty 02 Dec '09

02 Dec '09

Tom Schrijvers wrote, > I was wondering whether there are any universities that teach about Haskell type families or GADTs? I do in my course "Language-based Software Safety" (both TFs and GADTs). It's an advanced, research-oriented course for 4th year undergraduate and for postgraduate students. (It wasn't offered last year and this year, but will be again offered next year.) Manuel

1 0

Re: [Haskell] ANN: NoSlow - Microbenchmarks for array libraries
by Henning Thielemann 01 Dec '09

01 Dec '09

On Fri, 27 Nov 2009, Roman Leshchinskiy wrote: > You can get more information (including the ugly tables) from my blog > > http://unlines.wordpress.com/2009/11/27/noslow Btw. storablevector supports 'zip' using http://hackage.haskell.org/package/storable-tuple but you may also use 'zipWith' with an atomar result type for testing. I found that the dependency on Cairo and GTK originates from criterion. I have to build criterion with 'cabal install -f-chart criterion'. Unfortunately I cannot do 'cabal fetch criterion' in order to look into the cabal file, before the dependencies are resolved ...

3 5

module export question
by Sean McLaughlin 01 Dec '09

01 Dec '09

Say I have the following module: ---------------------------------------- module A ( T(T) , t , val ) where data T = T { t :: Int } val :: T val = T 7 ---------------------------------------- When I use A with the following imports, I don't expect this to work, but it does: import qualified A import A(T(..)) main = putStrLn $ show $ t A.val The problem is that I explicitly didn't export 't' as an element of T (by not writing T(..)). Am I just misunderstanding how exports work? I couldn't figure out what the correct behavior should be by looking at the 98 report. Sean

4 3

lazy logging (was: Are there standard idioms for lazy, pure error handling?)
by Evan Laforge 01 Dec '09

01 Dec '09

This is only peripherally related, but I also have a lot of list functions that can possibly be an error, but usually processing can continue. So they tend to return [Either Error Result]. I have another function thus: -- A foldr version is not lazy enough and overflows the stack. partition_either [] = ([], []) partition_either (x:xs) = let (ls, rs) = partition_either xs in case x of Left l -> (l:ls, rs) Right r -> (ls, r:rs) I was a little surprised I couldn't find this in the standard library... or maybe it is? Anyway, the trick is logging the errors while incrementally processing the non-errors. Logging them first forces the whole thing. Mixing the logging in with the processing forces the processing to be in IO and is not as nice as having it separate. If the next processing step is also pure, the errors must be propagated. 'map (fmap f)' should do it, as long as the errors are all the same type. It's still not ideal because the logs happen in a bunch at the end, and are still taking up memory meanwhile. The only thing I can think of is the dreaded lazy IO... plain unsafePerformIO wouldn't work because nothing is forcing it. It would be sort of an output version of getContents, but it would have to have some magic to know when a thunk has been evaluated without forcing it. But hey, the debugger does it! And the RTS already happily logs things in the middle of pure computation, so this would just be giving programmers access to it... so it's not *so* crazy :)

2 2

Fwd: Timing and Atom
by Lee Pike 01 Dec '09

01 Dec '09

[Tom -- resending my reply---I forgot to post to the list.] Neil, could you provide a reference (more on the practical side than the theory side) for the Stochastic Process Algebras you mention? And is there an embedding in Haskell? :) Lee Begin forwarded message: From: Lee Pike <leepike(a)gmail.com> Date: November 30, 2009 10:14:33 PM PST To: Tom Hawkins <tomahawkins(a)gmail.com> Subject: Re: Timing and Atom Tom, Thanks a lot for the detailed reply! > If you need greater timing resolution than what is provided by the > main loop, then I think the only option is to reference a hardware > counter. Right, of course. Use a hardware counter if you're not using a RTOS. That was the quick answer I wasn't considering. :) > I never considered running Atom generated functions in an asynchronous > loop until you posted your "Atomic Fibonacci Server" example Yeah, I always think it's fun to use synchronous languages asynchronously. Usually people think of trying to implement synchrony from asynchronous languages. > But there is a wide field of applications where this approach would > work just > fine -- not having to worry about meeting a hard loop time would > certainly simplify the design. I think it's a trade-off, right? You don't worry as much about time, but you have to worry about doing synchronization using handshakes or similar protocols. I know folks who make fault-tolerant systems generally prefer synchronized systems since asynchrony gets really hard once you start considering the possibility of random faults (See, for example, <http://embedded.eecs.berkeley.edu/giotto/> or _Real-Time Systems_ by Hermann Kopetz). > Could some form of static analysis be applied to provide the timing > guarantees? Of the asynchronous system (e.g., the "Fib. server")? In something like that, it should function under any possible schedule because of the handshakes (I should model-check it to prove that). Did I misunderstand your point? Thanks again for answering my question! Lee

1 0

KiCS (Curry to Haskell interpreter) problem
by Pasqualino "Titto" Assini 01 Dec '09

01 Dec '09

Hi, I am playing around with KiCS and I have a strange problem, when I evaluate a goal the variable bindings are not displayed, I see only the value of the expression. The same expression evaluated in pakcs (another curry interpreter) displays the bindings correctly. Is this a known bug? I would have contacted the author but his email in not in the haskell cabal file. Incidentally, is anyone else using KiCS ? Any comments/remarks ? Best, titto

1 0

Great Programs to Read?
by Michael Lesniak 01 Dec '09

01 Dec '09

Hello, In terms of "to become a great programmer, you need to read great programs"[1] what are "great" programs written in Haskell (for your personal definition of great), which source code is freely available on hackage or somewhere else on the net? I'm personally also interested in your definitions of great; for me, a great programs is defined by one of * good and well-written documentation (literate Haskell helps a lot) * novel ideas to use functional programming * elegance * showing how functional programming can ease tasks that are difficult to achieve in an imperative style Maybe we should create a Page on haskell.org (which I would do if I had write-access) mirroring the pages [2,3]? Kind regards, Michael [1] http://c2.com/cgi/wiki/Wiki?ReadGreatPrograms [2] http://c2.com/cgi/wiki/Wiki?GreatProgramsToRead [3] http://c2.com/cgi/wiki/Wiki?ProgramsToRead -- Dipl.-Inf. Michael C. Lesniak University of Kassel Programming Languages / Methodologies Research Group Department of Computer Science and Electrical Engineering Wilhelmshöher Allee 73 34121 Kassel Phone: +49-(0)561-804-6269

6 6

Re: [Haskell-cafe] Function composition questions from a newbie
by newbie2009 01 Dec '09

01 Dec '09

I am a newbie. Consider this code: square x = x * x add3 x y z = x + y + z leledumbo wrote: > > > what about (square . add3) 1 2? > > It doesn't work since add3, when curried (arguments of square "blended" > with add3's) with 1 argument becomes: > > add3 :: Num a => a -> a -> a > > which is a function that accepts 2 arguments and it's not compatible with > square's a. > Thank you so much for your clarification. To understand better, I tried some definitions 1) composition = (square . add3) 5 -- illegal as you had explained 2) composition x = (square . add3) 1 2 3 -- legal, but why? what is the significance of x 3) composition x = (square . add3) x -- legal, but why? 4) composition x y = (square . add3) x y -- legal, but why? 5) composition x = (square . add3) x 3 -- legal, but why? I dont understand why the above functions are legal, and what arguments they need. Could you please give an example of how to invoke it? -- View this message in context: http://old.nabble.com/Function-composition-questions-from-a-newbie-tp265702… Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

3 3

Help mixing pure and IO code
by papa.eric＠free.fr 01 Dec '09

01 Dec '09

Hi haskell helpers, Learning haskell, I wanted to explore how to write pure code and then add some IO on top of it, keeping the main code pure. The idea was to write a very simple two-player game, then define some strategies to play it that do not involve IO, and finally use strategies involving Random or IO ("ask the user"). I failed to reuse the pure code, and the only solution I found was to rewrite most things for IO. Here is my attempt in literate haskell, it is quite short, I hope someone will be kind enough to tell me what I have missed... Thanks for any answer! The game is: each player in turn chooses a number, and wins if this number has already been chosen twice (the third occurrence wins). > import Debug.Trace > import Control.Monad Choosing a number is a "move", a game is all moves played so far. > type Move = Int > type Game = [Move] > > newGame :: Game > newGame = [] Game updating rule, Nothing for a winning move (the game stops) else Just the ongoing game. > update :: Game -> Move -> Maybe Game > update game move = if move `wins` game then Nothing else Just (move:game) > where move `wins` game = length (filter (== move) game) == 2 Let's give a name to both players, mainly for tracing purposes. > data Player = Player1 | Player2 deriving Show > > myTrace :: Player -> Move -> (a -> a) > myTrace player move = trace ((show player) ++ " plays " ++ (show move)) A strategy looks at the game and proposes the next move, and may involve state. > class Strategy a where > proposeNext :: Game -> a -> (Move, a) The "game engine" takes two strategies and returns the winner, tracing what is going on. runGame starts from a new (empty) game, runGame' updates an ongoing game and keeps track of who is Player1 or Player2. > runGame :: (Strategy a1, Strategy a2) => a1 -> a2 -> Player > runGame x y = runGame' (Player1, x) (Player2, y) newGame > > runGame' :: (Strategy a1, Strategy a2) => > (Player, a1) -> (Player, a2) -> Game -> Player > runGame' (px, x) (py, y) game = -- x and y are strategies > let (move, x') = proposeNext game x > follows = case update game move of > Nothing -> px -- winning move > Just nextgame -> runGame' (py, y) (px, x') nextgame > in myTrace px move $ follows An example of a pure strategy that plays a fixed list of numbers, then zeroes. > data Fixed = Fixed [Move] deriving Show > > instance Strategy Fixed where > proposeNext game s = case s of > Fixed [] -> (0, Fixed []) > Fixed (x:xs) -> (x, Fixed xs) Now you may run runGame (Fixed [1,2,3]) (Fixed [3,2,1,3]) and Player2 will win. Now I want the user to be asked for moves. This one works quite well, asking *once* the user for a list of moves. > askFixed :: String -> IO Fixed > askFixed name = liftM Fixed askio > where askio = putStr (name ++ ", pick a list of numbers: ") >> readLn I could easily reuse runGame: > runGameSingleIO :: (Strategy a1, Strategy a2) => IO a1 -> IO a2 -> IO Player > runGameSingleIO = liftM2 runGame And this works: runGameSingleIO (askFixed "Joe") (askFixed "Jack") Now I want each user to be asked for moves repeatedly until there is a win. This was my first try, using an infinite list and hoping lazyness would work. > askUntil :: String -> IO Fixed > askUntil name = liftM Fixed (sequence $ repeat askio) > where askio = putStr (name ++ ", pick a number: ") >> readLn However it does not work, if I evaluate: liftM2 runGame (askUntil "Joe") (askUntil "Jack") then Joe is indefinitely asked for his move... Now, I supposed that IO [Move] was maybe (when triggered) a "unitary action" returning the whole list with no lazyness, and that I would rather need to have IO attached to individual moves [IO Move], so that they could be triggered independently. Am I right, or is it still subtler? Anyway, I imagined I needed to start from > data FixedIO = FixedIO [IO Move] And that's where I failed. I could not define any instance of Strategy that could be turned to a (IO Strategy) and reuse runGame, and use [IO Move] somewhere. I could not bring the inner IOs of [IO Move] to be syhthetised in front of IO Strategy. Was it possible? All I was able to contrive is a new definition of Strategy and runGame for the IO case. This seems very awkward. Here it is: > class StrategyIO a where > proposeNextIO :: Game -> a -> IO (Move, a) > > runGameIO :: (StrategyIO a1, StrategyIO a2) => a1 -> a2 -> IO Player > runGameIO x y = runGameIO' (Player1, x) (Player2, y) newGame > > runGameIO' :: (StrategyIO a1, StrategyIO a2) => > (Player, a1) -> (Player, a2) -> Game -> IO Player > runGameIO' (px, x) (py, y) game = > do (move, x') <- proposeNextIO game x > case update game move of > Nothing -> return px > Just nextgame -> runGameIO' (py, y) (px, x') nextgame Then FixedIO could be made a strategy: > instance StrategyIO FixedIO where > proposeNextIO game s = case s of > FixedIO [] -> return (0, FixedIO []) > FixedIO (x:xs) -> liftM2 (,) x (return $ FixedIO xs) > > askFixedIO :: String -> FixedIO > askFixedIO name = FixedIO (repeat askio) > where askio = putStr (name ++ ", pick a number: ") >> readLn and this works: runGameIO (askFixedIO "Joe") (askFixedIO "Jack") However, I wonder how to do it reusing the "pure" versions, runGame and Strategy? Thanks for anyone that has followed so far! Eric

8 9

Re: Timing and Atom
by Tom Hawkins 01 Dec '09

01 Dec '09

On Tue, Dec 1, 2009 at 2:24 AM, Lee Pike <leepike(a)gmail.com> wrote: > Tom, > > I have a (hopefully) easy question about timing and Atom in the use-case > where you're handling all your own scheduling without relying on a RTOS > (where you get preemption). Suppose I want a rule to fire every 2ms. > Naturally, I'd want to set a timer. But timers in Atom are based on the > __global_clock, and the rate of the __global_clock is not a constant > time-reference: it depends on the rule complexity of each rule as well as > when rules are scheduled. > > So do you try to do do a WCET analysis for Atom-generated code running > without an RTOS and tweak periods/timers based on that to get regular > timing. This is a hard problem if you have a lot of periods that are > relatively prime (I'm assuming at Eaton, this is rarely the case). In our application, the RTOS runs the Atom generated function in a foreground task at 1KHz. It turns out this is the only task the RTOS has to schedule. So in our case, __global_clock is accurate to about 1ms. We only use an RTOS because we're stuck with a poor set of development tools. This scheduling could have been easily implemented by polling a hardware counter. This has the added benefit of easily checking for execution overruns: void main() { unsigned long long nextTime = 0; while (1) { while (hardwareCounter < nextTime); nextTime = hardwareCounter + delta; atomFunction(); if (hardwareCounter > nextTime) reportExecutionOverrun(); } } If you need greater timing resolution than what is provided by the main loop, then I think the only option is to reference a hardware counter. But this will only allow you to tweak a control law or a signal processing algorithm. It won't give you execution timing guarantees. This one limitation of Atom: everything has to be scheduled as a function of a base period. The trick is to select the right period. Too course grained and computations will take too long. Too fine grained and you'll spend a lot of effort breaking up computations into several tiny rules just to meet timing. It's the same tradeoff ASIC designers have when selecting a clock rate. I never considered running Atom generated functions in an asynchronous loop until you posted your "Atomic Fibonacci Server" example (http://leepike.wordpress.com/2009/05/05/an-atomic-fibonacci-server-explorin…) I'm curious how such a system would behave if it referenced a hardware clock to enable and disable rules. I can see how this could be problematic for hard real-time, safety critical stuff. But there is a wide field of applications where this approach would work just fine -- not having to worry about meeting a hard loop time would certainly simplify the design. Could some form of static analysis be applied to provide the timing guarantees?

2 1

ANNOUNCE: haskell-mode 2.7.0
by Svein Ove Aas 01 Dec '09

01 Dec '09

Haskell-mode, a major mode for writing Haskell code in Emacs, has been updated to version 2.7.0. This release contains mostly bugfixes, but also adds support for hlint highlighting via flymake. Additionally, haskell-mode now has a bug-tracker and mailing list. Mailing list: http://projects.haskell.org/cgi-bin/mailman/listinfo/haskellmode-emacs Bug tracker: http://trac.haskell.org/haskellmode-emacs/ Unless I am asked not to, minor (N.M.x) releases will only be announced on the mailing list from now on. Download location is, as always, http://projects.haskell.org/haskellmode-emacs/ The changelog is as follows: === Changes since 2.6.4 * fill-paragraph (M-q) now only affects comments, and correctly handles Haddock commentary. adaptive-fill-mode is turned off, as it was interfering. * Yet more unicode symbols * Better support for unicode encoding of haskell source files * mdo correctly indented * Indentation fixes, fixes to the fixes, and fixes to the fixes to the fixes * New command: M-x haskell-check, calls (by default) hlint on the current file. Also bound to C-c C-v. You can also use the flymake minor mode with this. -- Svein Ove Aas

4 4

Re: [Haskell-cafe] Function composition questions from a newbie
by leledumbo 01 Dec '09

01 Dec '09

> This doesnt. But why? Back to the definition of function composition: f . g is possible if g returns a value that's compatible with f's argument. Now, let's check the type of square and add3: square :: Num a => a -> a add3 :: Num a => a -> a -> a -> a (square . add3 1 2) is actually seen by the compiler / interpreter as (square . (add3 1 2)) due to precedence of . operator. So, what's the type of add3 1 2? add3 :: Num a => a -> a Hmm... seems compatible with square. what about (square . add3) 1 2? It doesn't work since add3, when curried (arguments of square "blended" with add3's) with 1 argument becomes: add3 :: Num a => a -> a -> a which is a function that accepts 2 arguments and it's not compatible with square's a. -- View this message in context: http://old.nabble.com/Function-composition-questions-from-a-newbie-tp265702… Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

1 0

Tips for a FFI-wrapped C library?
by Casey Hawthorne 30 Nov '09

30 Nov '09

I would suspect that the IO Monad is to unsafe, maybe a more stringent monad would be more appropriate. -- Regards, Casey

1 0

Re: [Haskell-cafe] Function composition questions from a newbie
by muad 30 Nov '09

30 Nov '09

newbie2009 wrote: > > I am a newbie. Consider this code: > > square x = x * x > add3 x y z = x + y + z > add x y = x + y > composition5 x = (square . add3) x > composition6 x y = (square . add3) x y > > 1) What arguments can i pass to composition5? Please give an example of > calling it. > 2) What arguments can i pass to composition6? Please give an example of > calling it. > 3) How can I write the function "add" to take any number of arguments? I > would like to call it either as > "add 1 2" or "add 1 2 3" etc > > Thank you. > > for (1) and (2) you can ask haskell what the types are, Prelude> :t composition5 composition5 :: (Num (a -> a -> a), Num a) => a -> a -> a -> a Prelude> :t composition6 composition6 :: (Num (a -> a -> a), Num a) => a -> a -> a -> a This is a worry, it wants binary operation to be a number but it is not! There must be an error in the program. Let us try and fix by changing them to composition5 x = (square . add x) composition6 x y = (square . add3 x y) y now we see: Prelude> :t composition5 composition5 :: (Num a) => a -> a -> a Prelude> :t composition6 composition6 :: (Num a) => a -> a -> a which is reasonable, Prelude> composition5 1 2 9 Prelude> composition6 3 4 121 As for question 3, one would be better to take a list of numbers and to add [1,2,3] instead of add 1 2 3. (Something similar to add 1 2 3 is possible in a couple different way (have a go without using any typeclasses!) but more advanced than one should like to bother with at this point). add [] = 0 add (x:xs) = x + add xs would be fine anyway. -- View this message in context: http://old.nabble.com/Function-composition-questions-from-a-newbie-tp265702… Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

1 0

haddock + class instances + hidden types
by Sean McLaughlin 30 Nov '09

30 Nov '09

Hi, Is there a way to make haddock not generate instance documentation for non-exported types? Right now for instance, if there is a class A in module A, and I have a hidden type B in module B and I have instance A.A B in module B, then in the documentation for module A, the (supposedly hidden) type B shows up as one of the instances. There is no hyperlink, but I don't think it should show up there. Thanks! Sean

2 1

Parsec - separating Parsing from Lexing
by Fernando Henrique Sanches 30 Nov '09

30 Nov '09

Hello. I'm currently implementing a MicroJava compiler for a college assignment (the implemented language was defined, the implementation language was of free choice). I've sucessfully implemented the lexer using Parsec. It has the type String -> Parser [MJVal], where MJVal are all the possible tokens. However, I don't know how to implement the parser, or at least how to do it keeping it distinguished from the lexer. For example, the MicroJava grammar specifies: Program = "program" ident {ConstDecl | VarDecl | ClassDecl} "{" {MethodDecl} "}". The natural solution (for me) would be: program = do string "program" programName <- identifier ... However, I can't do this because the file is already tokenized, what I have is something like: [Program_, identifier_ "testProgram", lBrace_, ...] for the example program: program testProgram { ... How should I implement the parser separated from the lexer? That is, how should I parse Tokens instead of Strings in the "Haskell way"? Fernando Henrique Sanches

5 6

control-monad-failure and mtl
by Edward Z. Yang 30 Nov '09

30 Nov '09

Hello folks, I took advantage of Thanksgiving weekend to port my application to use Control.Monad.Failure, and learned (slightly painfully) that I still needed to pick some mechanism to instantiate my failure monads as. After the experience, I have three questions/comments: 1. Why isn't there an instance for Either in mtl? (There is one for Transformers. The error message left me very puzzled there: the docs clearly claimed the instance existed, and only a little source code diving elucidated the situation.) Copying the instance declaration from the transformers version seems to fix it. 2. I was having difficulty instantiating MonadFailure as an ErrorT for an arbitrary monad. Here is an example: {-# LANGUAGE PackageImports, FlexibleContexts #-} import "mtl" Control.Monad.Error import "mtl" Control.Monad.State import Control.Monad.Failure data MyError = MyError String instance Error MyError where strMsg = MyError type MyMonad = ErrorT MyError (State Integer) failureFunction :: MonadFailure MyError m => Integer -> m Integer failureFunction 0 = failure $ MyError "Cannot use zero" failureFunction n = return (n - 1) -- instantiate monadicFunction :: MyMonad Integer monadicFunction = failureFunction 23 Which results in the following error: failure.hs:19:18: No instance for (MonadFailure MyError (ErrorT MyError (State Integer))) arising from a use of `failureFunction' at failure.hs:19:18-35 Possible fix: add an instance declaration for (MonadFailure MyError (ErrorT MyError (State Integer))) In the expression: failureFunction 23 In the definition of `monadicFunction': monadicFunction = failureFunction 23 Which seems to contradict the documentation and source code, which states: Instances: [...] (Error e, Monad m) => MonadFailure e (ErrorT e m) How do I misunderstand? 3. In a motivating example, one of the goals of MonadFailure is to let us mix the error code of third-party modules into the generic failure mode. Control.Monad.Failure appears to give the machinery for instantiating a generic failure monad, but it doesn't have any facilities for the opposite direction: that is, marshalling a specific error form into the generic error form. Am I mistaken, and if not, would it be a welcome addition to the library? Cheers, Edward

2 2

peekCString & free memory
by El Barto 30 Nov '09

30 Nov '09

Hello, I get a segfault when I do str <- peekCString ptr free ptr return (Just str) But not when I don't free the C pointer. I also get the same behaviour with ByteString.packCString... Could you please tell me if the memory is correctly freed by GHC when I don't do it myself? And how can I specify a custom free function (i.e. xmlFree function in libxml2)? Maybe I should use a data type with two fields : the String/ByteString and the ForeignPtr to the CString? Regards

4 5

Mystery operator?
by michael rice 30 Nov '09

30 Nov '09

From: http://www.haskell.org/haskellwiki/Blow_your_mind#Polynomials -- splitting in two (alternating) -- "1234567" -> ("1357", "246") -- the lazy match with ~ is necessary for efficiency, especially enabling processing of infinite lists foldr (\a ~(x,y) -> (a:y,x)) ([],[]) This works but can't find (~) operator anywhere. Please explain or site a reference. Michael

6 10

Multi-versioning and compatibility classes.
by Jason Dusek 30 Nov '09

30 Nov '09

I have been looking at the Cabal roadmap (tickets marked for bottom) and notice that no one has mentioned multi-versioning or compatibility classes. Maybe these are bad ideas? Multi-versioning seems highly desirable. Essentially, it allows us to have the same qualified names be provided by different packages for different packages. This is desirable in the case of package versions which really are no longer the same, like Parsec 2 and Parsec 3; it also allows for the case where more than one package specifies the same qualified name because it is obvious (like `Text.JSON` or some such). It stands to reason that the "package environment" specified in the `.cabal` file is enough to tell us which package we really need to use. However, there are problems, too. Like what if I have one package that depends on `parsec-2.1.0.1` and another that depends on `parsec-2.1.0.0`? With multi-versioning, they both compile fine but then I try to use a parser from one with a combinator defined in the other: this fails to compile because: parsec-2.1.0.1:Text.ParserCombinators.Parsec.GenParser is not the same as: parsec-2.1.0.0:Text.ParserCombinators.Parsec.GenParser So now we need compatibility classes. The first package might specify a package dependency like: ~parsec-2.1.0.1 where we take the `~` to mean "requires something compatible with". The second package specifies a similar dependency: ~parsec-2.1.0.0 Then the maintainer of Parsec specifies that 2.1.0.1 is compatible with 2.1.0.0 and we can compile the second package, the one that require 2.1.0.0, with 2.1.0.1 since they are compatible and that is all it is asking for. Then we try to use these two packages together and everything works. -- Jason Dusek

1 0

Define a module in multiple files
by Ozgur Akgun 30 Nov '09

30 Nov '09

Dear Cafe, Is there a way of splitting the definition of a module into multiple files? Cheers, -- Ozgur Akgun

2 3

Re: Are there standard idioms for lazy, pure, error handling?
by John Lato 30 Nov '09

30 Nov '09

> From: Bas van Dijk <v.dijk.bas(a)gmail.com> > > On Mon, Nov 30, 2009 at 6:22 AM, John Millikin <jmillikin(a)gmail.com> wrote: >> ...I've considered two possible error handling modes... > > Regarding parsing, there's a third option: iteratees[1]. See [2] for a > motivation and description of iteratees. > I think it's interesting to note that this works with iteratees because they essentially do exactly what Malcolm suggested. That is, when an error occurs, it is referenced in the internal state of the iteratee, and processing is aborted. Iteratees are strict, but they allow for many of the same benefits of laziness, and can work well for large data sets. John

1 0

ANNOUNCE: graphviz-2999.7.0.0
by Ivan Lazar Miljenovic 30 Nov '09

30 Nov '09

I'm pleased to announce the latest version of my Haskell bindings to the Graphviz [1] suite of tools for visualising graphs. As usual, it is available from Hackage [2]. [1] http://graphviz.org/ [2] http://hackage.haskell.org/package/graphviz Changes in this release include: * Updated and extended functions to interact with the Graphviz tools. This now includes: - a better listing of available output types; - distinguishing file outputs from canvas outputs; - ability to automatically add the correct file extension to file outputs - Return any errors if calling Graphviz failed rather than just printing them to stderr * Improved Color support: - fixed ColorScheme values - explicitly named X11 colors - conversion to/from values from the colour library * Removed OrientationGraph; problems with distinguishing it when parsing from node-based Orientation values; its functionality is duplicated by Rotate. * By default, the generated Dot code is now no longer indented; you can now use the prettyPrint functions in Data.GraphViz to produce readable Dot code. * Added a testsuite; this is buildable by building with --flags=test. Numerous printing and parsing bugs were picked up with this. Plans for the next release: * Determine how to deal with encodings, since Graphviz uses UTF-8 by default; the hSetEncoding function in GHC-6.12 would be ideal, but backwards-incompatible. * Add a generic version of the types in Data.GraphViz.Types that would let people have more fine-grained control over the layout of their graphs by removing the ordering restriction used. Note that this will _not_ be the default, just an option for those that want it. -- Ivan Lazar Miljenovic Ivan.Miljenovic(a)gmail.com IvanMiljenovic.wordpress.com

1 0

License question
by jean-christophe mincke 30 Nov '09

30 Nov '09

Hello, I am writing a port to F# of some haskell standard libraries. I would like to publish them under an open source license and mention the origin of the initial code. www.haskell.org is under the simple permissive license. Does this license also cover the souce code available from that site? Is there any restriction about the license I might choose? Thak you Regards jcm

4 5

hscolour + unicode
by Sean McLaughlin 30 Nov '09

30 Nov '09

Hello, Is there a way to get the source documentation from cabal haddock --hyperlink-source to show up with unicode characters preserved? Right now the unicode from my comments and source programs are being mangled. I tried the --haddock-option="--use-unicode" option, but it seems to only work on the haddock docs, not the sources that are hscoloured and linked. Thanks, Sean

1 0

Haskell Job Openings (Pune, IN)
by Tom Hawkins 30 Nov '09

30 Nov '09

We finally got some open reqs in Eaton's engineering center in Pune, India. Could involve a little Haskell programming, or a lot, depending on what you want to do. http://jobsearch.naukri.com/job-listings-Senior-Engineer-Automotive-Embedde… I'd be happy to forward on any resumes to the hiring manager. -Tom

1 0

Haskell Weekly News: Issue 141 - November 29, 2009
by jfredett＠gmail.com 30 Nov '09

30 Nov '09

--------------------------------------------------------------------------- Haskell Weekly News http://sequence.complete.org/hwn/20091129 Issue 141 - November 29, 2009 --------------------------------------------------------------------------- Welcome to issue 141 of HWN, a newsletter covering developments in the [1]Haskell community. Firstly, sorry for the late HWN, a turkey coma is equal parts gift and curse. This week, we have some very exciting news, a new Haskell standard -- Haskell 2010 -- was announced, including several small changes. Also, we have lots of new packages, a possible Boston-area Hackathon coming up soon, and some great discussion. I won't hold you back from skipping all that and just reading the funny quotes! Haskellers, your (belated) Haskell Weekly News! Announcements Haskell 2010. Simon Marlow [2]announced the new revision of Haskell, Haskell 2010. Part of the new, less monolithic Haskell Prime process, Haskell 2010 includes several changes to the Haskell Language, including support for nonstandard if-then-else syntax (particularly wrt `do` notation and indentation), pattern guards, and several other changes. See the post for details. Clutterhs 0.1. Matt Arsenault [3]announced Clutterhs, version 0.1. A set of bindings for Clutter, a GObject based library for creating 2.5D interfaces using OpenGL. Interesting experiences of test automation in Haskell? Automation of Software Test 2010. John Hughes [4]announced a 'heads up' for the Automation af Software Test 2010 workshop NoSlow - Microbenchmarks for array libraries. Roman Leshchinskiy [5]announced his benchmark suite for various array and list libraries. CMCS 2010: First call for papers. Alexandra Silva [6]announced a first call for papers for the Tenth International Workshop on Coalgebraic Methods in Computer Science, taking place 26-28 March 2010, in Paphos, Cyprus. GPCE'10 First Call for Papers. Bruno Oliveira [7]announced a first call for papers for the Ninth International Conference on Generative Programming and Component Engineering. GPCE 2010 October 10-13, in Eindhoven, The Netherlands. Call for Participation: TLDI'10. Andrew Kennedy [8]announced a call for participation in the 2010 ACM SIGPLAN Workshop on Types in Language Design and Implementation Deadline Extension: JSC Special Issue on Automated Specification and Verification of Web Systems. demis [9]announced an extension to the paper deadline for the JSC Special Issue on Automated Specification and Verification of Web Systems. VSTTE 2010: Verified Software -- Second Call for Papers. Gudmund Grov [10]announced the second call for papers for the Third International Conference on Verified Software: Theories, Tools, and Experiments GPipe-1.1.0 with greatly improved performance. Tobias Bexelius [11]announced a new version of the GPipe package, now with greatly improved performance. wumpus-core. Stephen T [12]announced a new version of the wumpus-core package, a library for drawing 2D vector pictures, supporting output to SVG and postscript. package-vt-0.1.3.3, Haskell Package Versioning Tool. Krzysztof Skrzetnicki [13]announced the first release of his automatic version tracking tool, package-vt. Elerea 1.1. Patai Gergely [14]announced a new version of Elerea, a simple pull-based FRP library. Elerea (and FRP in general) allow for stream oriented programming, typically done in a applicative style. mecha-0.0.4. Tom Hawkins [15]announced a new version of Mecha, a little constructive solid modeling language intended for doing 3D CAD. atom-0.1.2. Tom Hawkins [16]announced a new release of Atom, a DSL for designing hard realtime embedded software with Haskell. This release adds guarded division operations, a new scheduling constraint, and a new rule scheduling algorithm. Managing Cabal Dependencies using Nix and Hack-nix. Marc Weber [17]announced a package for dealing with Cabal dependencies on the Nix OS platform. Discussion haskell in online contests. vishnu [18]asked about using Haskell in online contests, and particularly dealing with the SPOJ tool for judging programs. Namespaces for values, types, and classes. Sebastian Fischer [19]suggested allowing a namespace separation between class-names and other language elements. I miss OO. Michael Mossey [20]lamented his desire for Object-oriented features in Haskell, this led to a interesting discussion about name punning and typeclasses. Haskell Hackathon in Boston January 29th-31st? Ravi Nanavati [21]proposed a potential Hackathon in this editor's favorite city, to be held the 29th to the 31st. Blog noise [22]Haskell news from the [23]blogosphere. Blog posts from people new to the Haskell community are marked with >>>, be sure to welcome them! * Dan Piponi (sigfpe): [24]Programming with impossible functions, or how to get along without monads.. * Neil Brown: [25]Graph Layout with Software Transactional Memory and Barriers (plus video!). * Ivan Lazar Miljenovic: [26]If wishes were tests, code would be perfect. * Adam Jones: [27]Lambda Calculus compiler, Part II: Wading in with arithmetic. * FP Lunch: [28]Implementing a Correct Type-Checker for the Simply Typed Lambda Calculus. * Manuel M T Chakravarty: [29]Haskell 2010. * Holumbus: [30]hayoo.info. * Paul Chiusano: [31]Perfect strictness analysis (Part 2). * Paul Chiusano: [32]Perfect strictness analysis (Part 1). * Paul Chiusano: [33]Optional laziness doesn't quite cut it. * Neil Brown: [34]Force-Directed Graph Layout with Barriers and Shared Channels. * Neil Mitchell: [35]Reviewing View Patterns. * JP Moresmau: [36]EclipseFP 1.109.0 is out!. * CS Design Lab, University of Kansas: [37]Special LAMBDA meeting. * Joachim Breitner: [38]arbtt now in Debian. * Neil Mitchell: [39]Haskell DLL's on Windows. * Conal Elliott: [40]Thoughts on semantics for 3D graphics. Quotes of the Week * command: lambdabot will say 'it is forever etched into my memory' and then forget the quote by tomorrow. * copumpkin: orbitz makes [##C++] especially brutal to make people appreciate #haskell more * mtnviewmark: [on the State monad] "If I had a sta-ate, I'd compute it in the morning.... I'd compute it in the evening.... All over this la-and!" * monochrom: dynamic type is poor man's dependent type, they are so poor they can't buy a theorem prover. * Veinor: just remember: you can't spell fundeps without fun. * Botje: i think it's too early to be sticking things into other things .. * Berengal: Haskell is made of sugar and spice and everything nice, plus the mysterious compound X * sproingie: sproingie enables IncoherentInstances and ends up with Sarah Palin in his living room About the Haskell Weekly News New editions are posted to [41]the Haskell mailing list as well as to [42]the Haskell Sequence and [43]Planet Haskell. [44]RSS is also available, and headlines appear on [45]haskell.org. To help create new editions of this newsletter, please see the information on [46]how to contribute. Send stories to jfredett . at . gmail . dot . com. The darcs repository is available at darcs get [47]http://patch-tag.com/r/jfredett/HWN2/pullrepo HWN2 . References 1. http://haskell.org/ 2. http://article.gmane.org/gmane.comp.lang.haskell.general/17638 3. http://article.gmane.org/gmane.comp.lang.haskell.general/17652 4. http://article.gmane.org/gmane.comp.lang.haskell.general/17649 5. http://article.gmane.org/gmane.comp.lang.haskell.general/17647 6. http://article.gmane.org/gmane.comp.lang.haskell.general/17646 7. http://article.gmane.org/gmane.comp.lang.haskell.general/17645 8. http://article.gmane.org/gmane.comp.lang.haskell.general/17644 9. http://article.gmane.org/gmane.comp.lang.haskell.general/17637 10. http://article.gmane.org/gmane.comp.lang.haskell.general/17636 11. http://article.gmane.org/gmane.comp.lang.haskell.general/17635 12. http://article.gmane.org/gmane.comp.lang.haskell.cafe/66796 13. http://article.gmane.org/gmane.comp.lang.haskell.cafe/66770 14. http://article.gmane.org/gmane.comp.lang.haskell.cafe/66751 15. http://article.gmane.org/gmane.comp.lang.haskell.cafe/66726 16. http://article.gmane.org/gmane.comp.lang.haskell.cafe/66722 17. http://article.gmane.org/gmane.comp.lang.haskell.cafe/66693 18. http://thread.gmane.org/gmane.comp.lang.haskell.cafe/66810 19. http://thread.gmane.org/gmane.comp.lang.haskell.cafe/66777 20. http://thread.gmane.org/gmane.comp.lang.haskell.cafe/66723 21. http://thread.gmane.org/gmane.comp.lang.haskell.cafe/66616 22. http://planet.haskell.org/ 23. http://haskell.org/haskellwiki/Blog_articles 24. http://blog.sigfpe.com/2009/11/programming-with-impossible-functions.html 25. http://chplib.wordpress.com/2009/11/26/graph-layout-with-software-transacti… 26. http://ivanmiljenovic.wordpress.com/2009/11/26/if-wishes-were-tests-code-wo… 27. http://blog.finiteimprobability.com/2009/11/25/lambda-calculus-compiler-par… 28. http://sneezy.cs.nott.ac.uk/fplunch/weblog/?p=291 29. http://justtesting.org/post/256283708 30. http://holumbus.fh-wedel.de/blog/?p=27 31. http://pchiusano.blogspot.com/2009/06/perfect-strictness-analysis-part-2.ht… 32. http://pchiusano.blogspot.com/2009/06/perfect-strictness-analysis-part-1.ht… 33. http://pchiusano.blogspot.com/2009/05/optional-laziness-doesnt-quite-cut-it… 34. http://chplib.wordpress.com/2009/11/24/force-directed-graph-layout-with-bar… 35. http://neilmitchell.blogspot.com/2009/11/reviewing-view-patterns.html 36. http://jpmoresmau.blogspot.com/2009/11/eclipsefp-11090-is-out.html 37. http://www.ittc.ku.edu/csdlblog/?p=20 38. https://www.joachim-breitner.de/blog/archives/352-arbtt-now-in-Debian.html 39. http://neilmitchell.blogspot.com/2009/11/haskell-dlls-on-windows.html 40. http://conal.net/blog/posts/thoughts-on-semantics-for-3d-graphics/ 41. http://www.haskell.org/mailman/listinfo/haskell 42. http://sequence.complete.org/ 43. http://planet.haskell.org/ 44. http://sequence.complete.org/node/feed 45. http://haskell.org/ 46. http://haskell.org/haskellwiki/HWN 47. http://patch-tag.com/r/jfredett/HWN2/pullrepo%20HWN2

1 0

FAD
by Ozgur Akgun 29 Nov '09

29 Nov '09

While searching for a design tool for Haskell, I've come across the PhD thesis of Dan Russell <http://www.cs.kent.ac.uk/pubs/2001/1152/> He also mentioned the tool in cafe some time ago: http://www.haskell.org/pipermail/haskell-cafe/2003-June/004606.html Does anyone in the cafe, or himself, direct me to the download link of this tool; so that I can try it out. Cheers, -- Ozgur Akgun

1 0

Partially applied functions
by Casey Hawthorne 29 Nov '09

29 Nov '09

You can pattern match on the right hand side of '|' in a list comprehension, since a list comprehension is just the list monad. Just changed a few things. Hopefully this answers the OP's question and any interested others. add :: Int -> Int -> Int add x y = x + y -- a list of partially applied functions adds = [add 3, add 5, add 7, add 3, add 5, add 8] -- an example usage of the list kP pred = map (\ f -> f 10 ) (addPs pred) -- Wanted to do things like this. -- add3s = filter (?) adds -- add3s = [add 3, add 3] -- addEvens = filter (?) adds --addEvens = [add 8] -- addPs num = [ x | x <- adds, x 0 == num ] -- Changed 'x' to 'f' indicating a function application. addPs pred = [ f | f <- adds, pred (f 0) ] -- Main> kP (==3) -- Main> kP even -- Regards, Casey

1 0

ghci breakpoints
by Sean McLaughlin 29 Nov '09

29 Nov '09

Hello, I'm having trouble setting breakpoints from ghci. I load the files and I can tell things are working correctly because I can run the program and list the locations where I want to set breakpoints. E.g. Prelude> :list Focus.focus 510 > focus :: forall s m. Class s m => Neg -> m [Foci] 511 > focus f = 512 > let ats = Path.atoms f But when I try to set a breakpoint I get Prelude> :break Focus.focus No breakpoints found at that location. Does anyone have ideas about what I'm doing wrong in this case? Do I need to load with some special options to have breakpoints work? Thanks, Sean

2 1

some help debugging ghci build on powerpc
by Brian Denheyer 29 Nov '09

29 Nov '09

I'm trying to build ghc so that ghci will be included under linux power-pc. The build dies here: ~/ghc6-6.10.4/rts$ /home/briand/ghc6-6.10.4/ghc/stage1-inplace/ghc -optc-O -optc-Wall -optc-W -optc-Wstrict-prototypes -optc-Wmissing-prototypes -optc-Wmissing-declarations -optc-Winline -optc-Waggregate-return -optc-I../includes -optc-I. -optc-Iparallel -optc-Ism -optc-DCOMPILING_RTS -optc-fomit-frame-pointer -optc-DNOSMP -optc-I../gmp/gmpbuild -optc-I../libffi/build/include -optc-fno-strict-aliasing -optc-w -H32m -O -optc-O2 -I../includes -I. -Iparallel -Ism -DCOMPILING_RTS -package-name rts -optc-DNOSMP -static -I../gmp/gmpbuild -I../libffi/build/include -I. -dcmm-lint -c StgCRun.c -o StgCRun.o /tmp/ghc21791_0/ghc21791_0.s: Assembler messages: /tmp/ghc21791_0/ghc21791_0.s:15:0: Error: junk at end of line, first unrecognized character is `@' /tmp/ghc21791_0/ghc21791_0.s:19:0: Error: junk at end of line, first unrecognized character is `@' That relates to this line(s): StgRun: .quad .StgRun,.TOC.@tocbase,0 .size StgRun,24 .globl StgReturn It _looks_ like it's grabbing the right stuff from StgCRun.c, i.e. it's really trying to compile for power pc. I was hoping that this was a bit of cruft and someone could direct me to a fix for the assembly portion so I can get a little farther along. Thanks, Brian P.S. Or somebody could also tell me if trying to have the latest ghc working on power pc is even worth working on.

2 1

namespaces for values, types, and classes
by Sebastian Fischer 29 Nov '09

29 Nov '09

Dear Café, types and values are in separate namespaces. So, we can write data Name = Name where the first `Name` is a type and the second is a value. However, classes are in the same namespace as types. We cannot write class Name a where ... instance Name Name where ... Class contexts and types appear in different contexts just like types and values (unless I'm missing something). I googled for the reason of this design choice but found nothing. Does anyone know why types and values are in separate namespaces but classes and types are not? Cheers, Sebastian -- Underestimating the novelty of the future is a time-honored tradition. (D.G.)

6 6

haskell in online contests
by vishnu 29 Nov '09

29 Nov '09

Ive just started learning haskell pretty recently and Ive been trying to solve some online contest problems as part of this exercise. However, Ive been having almost no success. For various reasons my answers almost always are too slow. I recently stumbled across this link which was quite useful http://www.haskell.org/haskellwiki/SPOJ. This helped me speed up some of my programs where input was slowing me down. But being a noob, to a large extent I don't even know why my programs are slow sometimes or how to tell what makes them slow. I've been attempting problems from www.codechef.com (which uses SPOJ) in actuality. Because I have an admin account I can actually compare my solution against others there (which are almost always in C/C++ or Java) to try and figure out if Im missing a trick. Recently the problem I picked up was http://www.codechef.com/problems/DDILEMMA/ and I worked through solutions that just don't seem to be fast enough. I looked at successful submissions in C++ and JAVA which seem to do mostly what I'm doing (ofcourse there are differences because those are imperative languages and I might be misunderstanding things.). I've got my program, test input that I generated, cost center analysis all up on this page. http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5120#a5137 I've been getting some significant help from the #haskell channel but unfortunately this hasn't helped me break the barrier I need to. So I was wondering if someone would be kind enough to help me understand the profiler output and help me understand how to improve performance in cases like this thanks

7 13

Re: [Haskell-cafe] haskell in online contests
by wren ng thornton 29 Nov '09

29 Nov '09

vishnu wrote: > wow ok, I'm not even able to see why they're different in the desugared > version They're different because case is strict binding (see caveat below) whereas let is lazy binding. If we say, let foo = big ugly computation in bar Then if foo isn't used in bar, it'll never be computed. (Even if foo occurs in bar, provided it occurs somewhere that won't be evaluated.) Whereas if we say, case big ugly computation of foo -> bar Then we're forced to do the computation, regardless of whether foo is used in bar. Actually, that's not entirely true. In this example, because foo is a variable, the binding will be lazy; whereas if we replaced foo with a (non-lazy) pattern then it'd be strict. Part of the complication in explaining the exact details is because both case and let can have patterns inside of the binding so the desugaring process has to continue recursively. Also there're some details about using let for defining recursive groups, which you can't do with case. Also Haskell's syntax conflates defining functions with doing pattern matching, so there's an extra step to separate functions using pattern matching into the definition binding vs the pattern matching. In any case, once you're done compiling you end up with a strict binding operator (kind of like case) and a lazy binding operator (kind of like let). When in doubt, try playing around with different definitions to see whether they break when you pass in 'undefined'. There are also a few papers out there on how the STG works under the covers. -- Live well, ~wren

1 0

Hoogle server error again
by michael rice 29 Nov '09

29 Nov '09

Just tried to use Hoogle and got: 500 Internal Server Error Michael

1 0

implementing recursive let
by Ben Franksen 28 Nov '09

28 Nov '09

I am trying to write an interpreter for a very simple untyped functional language. I have a problem with mutually recursive let expressions, for which my interpreter loops :( This is a code snippet from the eval function: > eval :: Expr -> Eval Value > eval (Let decls body) = mdo > let (names,exprs) = unzip decls > let updateEnv env = foldr (uncurry M.insert) env $ zip names values > (values,result) <- local updateEnv $ liftM2 (,) (mapM eval exprs) (eval body) > return result Module M is Data.Map, the environment is a simple map from strings to values. Values are defined as > data Value = Data String | Function (Value -> Eval Value) The Eval monad is defined as > newtype Eval a = Eval { > unEval :: ErrorT String (StateT Env (Writer [String])) a > } deriving ( > Monad, > MonadFix, > MonadWriter [String], -- for warnings & other messages > MonadState Env, > MonadError String > ) > instance MonadReader Env Eval where > ask = get > local tr act = do > s <- get > modify tr > r <- act > put s > return r When I test this with an extremely simple expression, something like "let x = 1 in x", the code above loops. I don't understand why, especially since in a previous version it worked. In the previous version I had a simpler monad stack that went > newtype Eval a = Eval { > unEval :: ReaderT Env (Writer [String])) a > } deriving ( > Monad, > MonadFix, > MonadWriter [String], > MonadReader Env > ) (Replacing reader with state was done so I can add definitions to the environment at runtime. The ErrorT provides for errors, such as application of a non-function.) Expressions not involving let work fine. Also, if I replace the above definition by one which does not allow recursion (not using mdo, evaluating the defining expressions before the variable gets added to the environment), then non-recursive let-expressions (like the simple example above) work just fine. I am out of ideas as to what causes this problem. Does the addition of ErrorT make my monad too strict? How else can I implement mutual recursion? Cheers Ben

3 10

Partially applied functions
by Casey Hawthorne 28 Nov '09

28 Nov '09

You can pattern match on the right hand side of '|' in a list comprehension, since a list comprehension is just the list monad. add :: Int -> Int -> Int add x y = x + y -- a list of partially applied functions adds = [add 3, add 5, add 7, add 3, add 5, add 8] -- an example usage of the list kP num = map (\ f -> f 10 ) (addPs num) -- add3s = filter (?) adds -- add3s = [add 3, add 3] -- addEvens = filter (?) adds --addEvens = [add 8] addPs num = [ x | x <- adds, x 0 == num ] -- Regards, Casey

1 0