- Haskell-Cafe - Haskell.org

Re: [Haskell-cafe] Embedded scripting Language for haskell app
by Felipe Lessa 18 Aug '10

18 Aug '10

On Tue, Aug 17, 2010 at 7:07 PM, Christopher Done <chrisdone(a)googlemail.com> wrote: > Sadly this is true. I went ahead and tested this to confirm; compiled > mueval (which uses hint), copied the executable to a virtual machine > and it required the GHC package repo among other GHC-related > libraries. > > The size is indeed a problem. But how much? How does this compare to Lua et al? IIRC, the Haskell Platform installer for Windows has around 70 MiB. So, if you want a simple installer to include in your installer, be prepared to have another 70 MiB. I said a dozen mebibytes (12 MiB) because I think the bare minimum needs to have at least this size to have something useful for an app, but this isn't backed up by anything. Lua, on the other hand, is embedded in the executable and weights less than 200 KiB, probably much less than a typical Haskell executable. I don't know about other interpreters. I just did a quick look at some executables I had in hand (compiled by GHC 6.12.3) and the smallest one weights at 2.3 MiB. Hmmm, let me see something: $ du -hs /usr/lib/ghc-6.12.3/ 588M /usr/lib/ghc-6.12.3/ Wow! That was unexpected. =) $ du -sm /usr/lib/ghc-6.12.3/* | sort -n | tail -n 5 23 /usr/lib/ghc-6.12.3/template-haskell-2.4.0.1 24 /usr/lib/ghc-6.12.3/ghc 77 /usr/lib/ghc-6.12.3/Cabal-1.8.0.6 78 /usr/lib/ghc-6.12.3/base-4.2.0.2 265 /usr/lib/ghc-6.12.3/ghc-6.12.3 $ du -sm /usr/lib/ghc-6.12.3/ghc-6.12.3/* | sort -n | tail -n 5 1 /usr/lib/ghc-6.12.3/ghc-6.12.3/ZipDataflow.p_hi 30 /usr/lib/ghc-6.12.3/ghc-6.12.3/libHSghc-6.12.3-ghc6.12.3.so 43 /usr/lib/ghc-6.12.3/ghc-6.12.3/HSghc-6.12.3.o 56 /usr/lib/ghc-6.12.3/ghc-6.12.3/libHSghc-6.12.3.a 111 /usr/lib/ghc-6.12.3/ghc-6.12.3/libHSghc-6.12.3_p.a Ah, so this one file alone takes 111 MiB =). But 30 MiB of libHSghc.so probably will need to be included. $ xz -9 < /usr/lib/ghc-6.12.3/ghc-6.12.3/libHSghc-6.12.3-ghc6.12.3.so | wc -c 4507512 So using one of the best generic compression algorithms available, the size of one of the biggest libraries that a program using hint may need comes down to 4.5 MiB. Probably if someone is careful enough to include only what really is necessary, the program installer will be at most 10 MiB and will need at most 50 MiB on disk. I think this is doable, but a lot more than Lua; hint can't be used on set-top boxes =). > Sandboxing and import whitelisting with hint is also a no-brainer -- > see mueval and http://tryhaskell.org/. I also implement import > filtering here with haskell-src-exts: > http://github.com/chrisdone/haskell-json/blob/master/haskell-json.hs#L137 > So import whitelisting is trivial. Sweet, I had forgotten about mueval. And wouldn't it be nice if we could hand haskell-src-exts's parse tree directly into GHC? Cheers! =) -- Felipe.

3 3

Which Mail editor or mail chain do you use ?
by Luc TAESCH 18 Aug '10

18 Aug '10

Gentlemen, Slightly off topic, but when reading this list , I was ( happily ) surprised when reading this list, about the quality of typing, including redaction,paragraph, greeks and footnote ref like , in just a (pure) text list ... baskell[1] seems interesting. And there's hslua[2]. Can one use hint[3] like this ? Thanks [1] baskell - http://hackage.haskell.org/package/baskell [2] hslua - http://hackage.haskell.org/package/hslua [3] hint - http://hackage.haskell.org/package/hint Is this some kind of idiom that developped over time and habit ( of people reading papers) , particularly educated, and in need for others recognition, or born out of using an (arche)typical environment/ toolchain Markdown - latex like ? May I ask you how you redact your answers and which toolchain you are using ? Thanks for quenching my curiosity Luc

4 3

Question about memory usage
by Tako Schotanus 18 Aug '10

18 Aug '10

I was reading this article: http://scienceblogs.com/goodmath/2009/11/writing_basic_functions_in_has.php And came to the part where it shows: > fiblist = 0 : 1 : (zipWith (+) fiblist (tail fiblist)) Very interesting stuff for somebody who comes from an imperative world of course. But then I read that "Once it's been referenced, then the list up to where you looked is concrete - the computations *won't* be repeated." and I started wondering how that works. Because this seems to mean that functions could have unknown (to the caller) memory requirements. How does one, programming in Haskell, keep that in check? And when does that memory get reclaimed? Cheers, -Tako

11 16

Re: [Haskell-cafe] Re: String vs ByteString
by Tako Schotanus 17 Aug '10

17 Aug '10

On Tue, Aug 17, 2010 at 13:29, Ketil Malde <ketil(a)malde.org> wrote: > Tako Schotanus <tako(a)codejive.org> writes: > > >> Just like Char is capable of encoding any valid Unicode codepoint. > > > Unless a Char in Haskell is 32 bits (or at least more than 16 bits) it > con > > NOT encode all Unicode points. > > And since it can encode (or rather, represent) any valid Unicode > codepoint, it follows that it is 32 bits (and at least more than 16 > bits). > > :-) > > (Char is basically a 32bit value, limited valid Unicode code points, so > it corresponds to UCS-4/UTF-32.) > > Yeah, I tried looking it up but I could find the technical definition for Char, but in the end I found that "maxBound" was "0x10FFFF" making it basically 24 bits :) I know for example that Java uses only 16 bits for its Chars and therefore can NOT give you all Unicode code points with a single Char, with Strings you can because of the extension points. -Tako

3 2

Re: [Haskell-cafe] Re: String vs ByteString
by Tom Harper 17 Aug '10

17 Aug '10

> I agree, Data.Text is great. Unfortunately, its internal use of UTF-16 > makes it inefficient for many purposes. In the first iteration of the Text package, UTF-16 was chosen because it had a nice balance of arithmetic overhead and space. The arithmetic for UTF-8 started to have serious performance impacts in situations where the entire document was outside ASCII (i.e. a Russian or Arabic document), but UTF-16 was still relatively compact, compared to both the UTF-32 and String alternatives. This, however, obviously does not represent your use case. I don't know if your use case is the more common one (though it seems likely). The underlying principles of Text should work fine with UTF-8. It has changed a lot since its original writing (thanks to some excellent tuning and maintenance by bos), including some more efficient binary arithmetic. The situation may have changed with respect to the performance limitations of UTF-8, or there may be room for it and a UTF-16 version. Any takers for implementing a UTF-8 version and comparing the two? > A large fraction - probably most - textual data isn't natural language > text, but data formatted in textual form, and these formats are > typically restricted to ASCII (except for a few text fields). > > For instance, a typical project for me might be 10-100GB of data, mostly > in various text formats, "real" text only making up a few percent of > this. The combined (all languages) Wikipedia is 2G words, probably less > than 20GB. > > Being agnostic about string encoding - viz. treating it as bytes - works > okay, but it would be nice to allow Unicode in the bits that actually > are text, like string fields and labels and such. Is your point that ASCII characters take up the same amount of space (i.e. 16 bits) as higher code points? Do you have any comparisons that quantify how much this affects your ability to process text in real terms? Does it make it too slow? Infeasible memory-wise?

8 12

putStrLn "ß"
by Henk-Jan van Tuyl 17 Aug '10

17 Aug '10

L.S., When I try putStrLn "ß" (Eszett (sharp S)) in WinGhci, the interpreter seems to have disappeared; ctrl-C gives the message "Interrupted" in a separate window, but no new prompt. When I start GHCi in a shell, the interpreter displays: Prelude> print "ß" "\223" Is this a known issue? Other experiments with WinGhci: >print "αγδ" "a?d" (Trying to print alpha, gamma, delta.) >print "β" (Trying to print beta) interpreter hangs again. Details: - WinGhci 1.0.2 - GHCi, version 6.12.3 - Windows XP Regards, Henk-Jan van Tuyl -- http://Van.Tuyl.eu/ http://members.chello.nl/hjgtuyl/tourdemonad.html --

4 3

ANN: primes-0.2.0.0
by Sebastian Fischer 16 Aug '10

16 Aug '10

Hello, I have uploaded new versions of the primes package to Hackage: http://hackage.haskell.org/package/primes Version 0.2.0.0 significantly improves the memory requirements (and as a result also the run time) compared to 0.1.1. The run time is now almost linear and when streaming an infinite list of primes the memory requirements are almost constant. I think, primes is now the most efficient prime generator on Hackage. The interface has been generalised to support arbitrary Integral types and not just Integers and I have added two convenience functions: isPrime :: Integral int => int -> Bool primeFactors :: Integral int => int -> [int] These functions can be used for numbers with moderately large prime factors but are impractical when passing RSA numbers [1] because they simply use trial division [2]. The performance improvements have also been implemented in version 0.1.1.1 which provides the old interface. As the only API changes (of 0.2) are additions and the generalisation to Integral types, switching directly to 0.2 should be straightforward. The implementation is similar to Melissa O'Neil's code on the Haskell Wiki [3] which is one of the most efficient (pure) prime generators (without an upper bound) in Haskell that I know of. My goal was to simplify her implementation without sacrificing performance significantly. The performance of the prime package now almost matches the performance of Melissa's code. Judge for yourself whether you find the implementation more readable: http://github.com/sebfisch/primes/blob/0.1.1.1/Data/Numbers/Primes.hs The 0.2 tarball contains (source, not binary) programs that can be used to measure the performance of my implementation using Criterion (runtime.hs) and checking the memory requirements when generating large primes (memory.hs). Cheers, Sebastian [1]: http://en.wikipedia.org/wiki/RSA_numbers [2]: http://en.wikipedia.org/wiki/Trial_division [3]: http://www.haskell.org/haskellwiki/Prime_numbers -- Underestimating the novelty of the future is a time-honored tradition. (D.G.)

1 0

expression problem
by Patrick Browne 16 Aug '10

16 Aug '10

Hi, The expression problem [1] can be described as the ability to add new variants (either constructors or methods) to a data type without changing the existing code. The Haskell and OO language issues are well described at [1] It seems that the expression problem does not exist in Maude[2]. My question is: Using basic and easy parts of the Haskell, does Haskell currently handle the expression problem. In other words I would like to avoid demanding and sophisticated coding techniques or non-standard Haskell extensions (i.e. it should be easy for the average programmer to use). If there is a simple solution could you please point me to an example. Thanks, Pat [1] Dr. Ralf Laemmel http://channel9.msdn.com/shows/Going+Deep/C9-Lectures-Dr-Ralf-Laemmel-Advan… [2]http://lists.cs.uiuc.edu/pipermail/maude-help/2010-January/000336.html This message has been scanned for content and viruses by the DIT Information Services E-Mail Scanning Service, and is believed to be clean. http://www.dit.ie

2 1

Deprecated gtk2hs functions
by Alex Rozenshteyn 16 Aug '10

16 Aug '10

I recently started playing around with gtk2hs. I noticed that `onClicked`, `afterClicked`, etc. functions have been deprecated, presumably in favor of the `on` and `after` functions in the Glib signals module, but I couldn't find a collection of the appropriate signals to replace the functionality. Am I simply being blind? -- Alex R

3 4

A GHC error message puzzle
by Yitzchak Gale 16 Aug '10

16 Aug '10

The file "error_puzzle.hs" begins like this: main = do inp <- readFile "input" writeFile "output" $ process inp process :: String -> String When compiled with GHC 6.12.3 and run, it gives the following result: $ ./error_puzzle error_puzzle: output: hClose: illegal operation (handle is finalized) What is the cause of the error? Give a definition for process and an input file that reproduce this result. No "unsafe" functions like unsafePerformIO are allowed. Have fun! -Yitz

9 24

xemacs newbie question - haskell mode
by rgowka1 16 Aug '10

16 Aug '10

HI - I have been struggling to get the Xemacs recognize hs file. I have installed the Haskell-mode.. However I keep getting the message "File mode specification error: (wrong-number-of-arguments require 3)".. How do I go about fixing this?? Many thanks

2 1

Assorted AT fun and games
by Andrew Coppin 16 Aug '10

16 Aug '10

It all began with class (Vector (Point x)) => HasSpace x where type Point x :: * So far, so good. I was rather surprised that you _can_ write class (Complete (Completed x)) => Incomplete x where type Completed x :: * complete :: x -> Completed x I was almost as surprised to discover that you _cannot_ write class (HasSpace x, Complete (Completed x), HasSpace (Completed x), Point x ~ Point (Completed x)) => Incomplete where... It just means that every time you write an Incomplete instance, you might have to add the Point constraint manually. Which is mildly irritating. More worrying, adding Point foo ~ Point bar to an instance declaration causes GHC to demand that you turn on Undecidable Instances, for reasons beyond my comprehension. It's also interesting that when you write a class instance that has some constraint on it, and then you try to write a subclass instance, you still have to repeat the exact same constraint, even though the superclass instance declaration implies it. The only reason I can think of is that theoretically somebody could add a new superclass instance without the constraint. (Wouldn't that be an overlapping instance though?) And now things get *really* interesting. Consider this: data Foo x = Foo !x !(Point x) Surprisingly, GHC accepts this. This despite the rather obvious fact that "Point x" can exist if and only if "x" has a HasSpace instance. And yet, the type definition appears to say that "x" is simply an unconstrained type variable. Intriguing... Next, you can't derive Eq or Show for this type, but you *can* declare them manually: instance (Show x, Show (Point x)) => Show (Foo x) where show (Foo x px) = "Foo (" ++ show x ++ ") (" ++ show px ++ ")" Again, no hint of the fact that this will only work if we have HasSpace x. And yet GHC happily accepts this. I'm starting to think maybe I'm pushing the type system further than it can cope, and I should just completely redesign the whole thing...

4 4

couchdb - newView command
by Andrew U. Frank 16 Aug '10

16 Aug '10

i use haskell couchdb 0.10 with more luck than others: it works for me quite well. I have been careful to convert to json strings separately from the calls to runCouchDB - so i could check that the values were correct. i sense there is a difficulty with 'newView' - it does not allow to update a viewset, but one has to select for each change a new viewset (design) name; which is sort of annoying. the newView operation should ptobably be called 'newViewset', because it sets multiple views in one design document. it would be nice, if that could not only be set, but also updated. (the full set at once) i would also recommend that the first parameter is of type DB and not String - or are there any particular reasons for this? i appreciate the package and i like it! andrew

2 1

hgettext and cabal
by abau 16 Aug '10

16 Aug '10

Hi, [1] describes how to use hgettext with cabal. When running "runhaskell Setup build" this works fine. But "cabal build" says: Not in scope: `__MESSAGE_CATALOG_DOMAIN__' Not in scope: `__MESSAGE_CATALOG_DIR__' [1] says: "__MESSAGE_CATALOG_DOMAIN__ and __MESSAGE_CATALOG_DIR__ are macro definitions, whose hold configured strings from the Cabal." Why "cabal build" does not expand these macros? Thank you. [1] http://www.haskell.org/haskellwiki/Internationalization_of_Haskell_programs

1 0

Re: philosophy of Haskell
by Patai Gergely 16 Aug '10

16 Aug '10

> No. As you say the world1 value is immutable, but that's not > contradictory. If between 'getLine' and 'print' something was done by a > concurrent thread, then that change to the world is captured by 'print'. Well, that's technically true, but it basically amounts to saying that the 'model' of IO is itself. If I see 'print x', I don't really think of an action that performs arbitrary side effects and prints x in the process (unless some error prevents even that). Some notion of compositionality is required, which would allow me to state how 'print x' contributes to the changes of the world in a way that doesn't depend on any context. Actions in the state monad are composable in that sense, while the same doesn't apply to the IO monad when explained in terms of state passing. Gergely -- http://www.fastmail.fm - The professional email service

2 1

Is bumping the version number evil, if it's not mandated by the PVP?
by Sebastian Fischer 15 Aug '10

15 Aug '10

Hello, I wonder whether (and how) I should increase the version number of a library when the API does not change but the implementation gets more efficient. Should I bump a.b.C or even a.B to signal that it's worth using the new version or should I bump only a.b.c.D such that packages that depend on a.b get installed with the new version automatically? When bumping only a.b.c.D, the new version is not installed as a dependency if the old version already is installed (unless the new version is explicitly demanded.) It seems bumping a.b.c.D has advantages for some users and disadvantages for others. Hence, I guess I should make a major version bump. Is it bad habit to make a major version bump if the API does not change? Maybe I should simply change the API too ;) Cheers, Sebastian -- Underestimating the novelty of the future is a time-honored tradition. (D.G.)

4 6

bug in Network.Browser
by Alexander Kotelnikov 15 Aug '10

15 Aug '10

Hello. Yesterday I hit a bug in Network.Browser: connection in the connection pool is not reused if you are connecting to a destination with qualified port (not 80 for HTTP or 80 but explicitly provided like http://www.google.com:80) The reason for the failure is quite trivial, but I had not found an easy fix which won't break anything. Problem is that in TCP.hs function isTCPConnectedTo :: HandleStream ty -> String -> IO Bool isTCPConnectedTo conn name = do v <- readMVar (getRef conn) case v of ConnClosed -> return False _ | map toLower (connHost v) == map toLower name -> catch (getPeerName (connSock v) >> return True) (const $ return False) | otherwise -> return False socket's hostname, (connHost v) is compared to hostname or hostname:port which is stored in name. Hope this message will reach this code maintainers Alexander

1 0

Help us test gio APIs!
by Andy Stewart 15 Aug '10

15 Aug '10

Hi all, I'm working on merge gio-branch (https://patch-tag.com/r/AndyStewart/gio-branch/home) to gtk2hs-0.12.0. GIO (http://library.gnome.org/devel/gio/stable/) is cross-platform APIs for file operation, we can use gio APIs develop file manager or similar application. If anyone want to use those APIs, help us test those APIs at https://patch-tag.com/r/AndyStewart/gio-branch/home, then we can release gio-0.12.0 quicker. A GIO file-manager screenshot at : http://www.flickr.com/photos/48809572@N02/4793031888/ Thanks! -- Andy

1 0

Removing polymorphism from type classes (viz. Functor) (Again)
by Alexey Karakulov 15 Aug '10

15 Aug '10

I was inspired by George Pollard's post<http://www.haskell.org/pipermail/haskell-cafe/2009-July/063981.html>at haskell-cafe and tried to implement the non-polymorphic Functor class ( I named it Functor' ). I changed some names and added reasonable constraints. type family NewPt f a class Functor' f where type Point f map ∷ (a ~ Point f, b ~ Point g, g ~ NewPt f b, Functor' g) ⇒ (a → b) → f → g I would like to be able to write: type OldPt f = NewPt f (Point f) class (f ~ OldPt f) ⇒ Functor' f ... but ghc says it's not implemented yet (version 6.12.1). However, it's not the main problem. Now I can write some instances: type instance NewPt [a] b = [b] instance Functor' [a] where type Point [a] = a map = fmap type instance NewPt ByteString a = ByteString instance Functor' ByteString where type Point ByteString = Word8 map = BS.map But I can't write instance for Set: type instance NewPt (Set a) b = Set b instance Ord a ⇒ Functor' (Set a) where type Point (Set a) = a map = Set.map ghci complains: Could not deduce (Ord a1) from the context (g ~ NewPt (Set a) a1, a1 ~ Point g, Functor' g) arising from a use of `Set.map' at ... The type of Set.map is Set.map :: (Ord a, Ord b) => (a -> b) -> Set a -> Set b (Ord a) is in the instance context, and what about b? Type of map for Set instance would be: original: map ∷ (a ~ Point f, b ~ Point g, g ~ NewPt f b, Functor' g) ⇒ (a → b) → f → g substitute: f → Set a, g → Set b map :: Functor' (Set b) ⇒ (a →b) →Set a →Set b (Ord b) must be deduced from (Functor (Set b)) but it doesn't. I don't know whether it's my mistake somewhere or ghc problem. (Sorry for my English, it's not perfect). -- All the best, Alexey

3 7

ANN: improve-0.0.8
by Tom Hawkins 15 Aug '10

15 Aug '10

ImProve [1] is a imperative programming language for high assurance applications. Using Yices [2], ImProve verifies programs will always adhere to assertion specifications, irrespective of program input. If it is possible for an assertion not to be upheld, ImProve will emit a counter example in the form of a program trace, which illustrates how the assertion could be violated. This release improves the means to annotate programs for debugging (see 'label'). Also, the model checker now includes a pass to remove irrelevant statements, thus speeding up verification and making counter example traces easier to read. At Eaton, we have started using ImProve on critical chunks of code in our hybrid vehicles. To date we have ImProve generated C successfully running in a new prototype vehicle. In addition to finding a handful of minor bugs, ImProve has revealed two specification ambiguities arising from multiple, conflicting requirements. Due to the nature of the conflict, it is highly unlikely that the defects would have been identified with conventional testing. -Tom [1] http://hackage.haskell.org/package/improve [2] http://yices.csl.sri.com/

1 0

Unwrapping long lines in text files
by michael rice 15 Aug '10

15 Aug '10

The program below takes a text file and unwraps all lines to 72 columns, but I'm getting an end of file message at the top of my output. How do I lose the EOF? Michael ====== unwrap.hs ====== main = do line <- getLine if null line then do putStrLn "" main else do printList (words line) 1 main printList :: [String] -> Int -> IO () printList [] _ = do putStrLn "" printList (w:[]) k = do if k+(length w) <= 72 then do putStrLn w else do putStrLn "" putStrLn w printList r@(w:ws) k = do if k+(length w) <= 72 then do putStr w putStr " " printList ws (k+(length w)+1) else do putStrLn "" printList r 1

7 14

Re: [Haskell-cafe] ATs vs FDs
by Antoine Latter 14 Aug '10

14 Aug '10

What's wrong with fun-deps? The associated type synonym syntax is prettier, but I didn't tknow that fun-deps were evil. Do you have any links? Take care, Antoine On Aug 14, 2010 5:19 AM, "Andrew Coppin" <andrewcoppin(a)btinternet.com> wrote: As I understand it, ATs were invented because FDs are "evil" and must never be used ever for any purpose. However, it doesn't seem to be possible to use ATs to do the same things that FDs can do. You can use ATs to write type functions, which take one type and return another type. This allows you to express type relationships in a very elegant way. However, what it does /not/ seem to allow you to do is express one-to-one relationships. For example, I'd like to be able to say that the next vector up from a Vector3 is a Vector4, and the next vector down is a Vector2. And I can say that. What I can't say is that the *only* next vector up is a Vector4. And thus, all my code is littered with ambiguous type warnings because although /currently/ there's only one class instance, somebody could come along some day and write another one. What am I missing? _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe(a)haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

3 2

ATs vs FDs
by Andrew Coppin 14 Aug '10

14 Aug '10

As I understand it, ATs were invented because FDs are "evil" and must never be used ever for any purpose. However, it doesn't seem to be possible to use ATs to do the same things that FDs can do. You can use ATs to write type functions, which take one type and return another type. This allows you to express type relationships in a very elegant way. However, what it does /not/ seem to allow you to do is express one-to-one relationships. For example, I'd like to be able to say that the next vector up from a Vector3 is a Vector4, and the next vector down is a Vector2. And I can say that. What I can't say is that the *only* next vector up is a Vector4. And thus, all my code is littered with ambiguous type warnings because although /currently/ there's only one class instance, somebody could come along some day and write another one. What am I missing?

3 5

Re: [Haskell-cafe] Is bumping the version number evil, if it's not mandated by the PVP?
by Sebastian Fischer 14 Aug '10

14 Aug '10

[CC'ing café again] On Aug 14, 2010, at 12:25 PM, Max Rabkin wrote: > On Sat, Aug 14, 2010 at 11:13 AM, Sebastian Fischer > <sebf(a)informatik.uni-kiel.de> wrote: >> Hello, >> >> I wonder whether (and how) I should increase the version number of >> a library >> when the API does not change but the implementation gets more >> efficient. >> >> Should I bump a.b.C or even a.B to signal that it's worth using the >> new >> version or should I bump only a.b.c.D such that packages that >> depend on a.b >> get installed with the new version automatically? > > My understanding is that the PVP only describes the *minimum* version > bump, not the maximum. There is a third option though: give the > updated version two version numbers, one with an a.b.c.D bump so that > reverse dependencies get the performance improvement, and one with an > a.b.C bump so that users see a new version as worthwhile. That's an interesting idea! In my case I'll probably bump version a.b.c.D with unchanged API but better performance and additionally release a new major version with a new API as well (as I planned to extend the API independently). Thanks! Sebastian -- Underestimating the novelty of the future is a time-honored tradition. (D.G.)

1 0

Fwd: [Haskell-cafe] ATs vs FDs
by Gábor Lehel 14 Aug '10

14 Aug '10

---------- Forwarded message ---------- From: Gábor Lehel <illissius(a)gmail.com> Date: 2010/8/14 Subject: Re: [Haskell-cafe] ATs vs FDs To: Andrew Coppin <andrewcoppin(a)btinternet.com> You're missing equality constraints for classes. In other words: class (a ~ Smaller (Bigger a), a ~ Bigger (Smaller a)) => Vector a where type Bigger a; type Smaller a FDs and TFs are equivalent in theory, however their implementation in GHC has shortcomings on (I believe) both sides. In this case, GHC unfortunately doesn't yet support equality constraints in class contexts, which you would need to do some of the things FDs can do. I believe there are workarounds for this in some cases, including one which uses plain type classes to encode not-quite-equality, and one where you just add the constraints to every method of the class instead, but I don't know the details. On Sat, Aug 14, 2010 at 12:19 PM, Andrew Coppin <andrewcoppin(a)btinternet.com> wrote: > As I understand it, ATs were invented because FDs are "evil" and must never > be used ever for any purpose. However, it doesn't seem to be possible to use > ATs to do the same things that FDs can do. > > You can use ATs to write type functions, which take one type and return > another type. This allows you to express type relationships in a very > elegant way. However, what it does /not/ seem to allow you to do is express > one-to-one relationships. > > For example, I'd like to be able to say that the next vector up from a > Vector3 is a Vector4, and the next vector down is a Vector2. And I can say > that. What I can't say is that the *only* next vector up is a Vector4. And > thus, all my code is littered with ambiguous type warnings because although > /currently/ there's only one class instance, somebody could come along some > day and write another one. > > What am I missing? > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe > -- Work is punishment for failing to procrastinate effectively. -- Work is punishment for failing to procrastinate effectively.

1 0

ANNOUNCE: Sifflet visual programming language, release 1.0!
by gdweber＠iue.edu 14 Aug '10

14 Aug '10

Sifflet 1.0 is now available on Hackage. Sifflet is a visual, functional programming language and support system for students learning about recursion. Sifflet programmers define functions by drawing diagrams, and the Sifflet interpreter uses diagrams to show how the function calls are evaluated. New in This Release ------------------- - Sifflet now has exporters to Scheme, Python, and Haskell and a small supporting "library" for each of those languages. (These exporters are intended for beginning students who are learning these languages, not for "industrial-strength" code production.) - Although the exporters apparently generate correct code, there are at present too many parentheses -- a few extras in Python, many extras in Haskell. (No extra parentheses in Scheme, of course.) - Sifflet has a new number type, Sifflet.Data.Number.Number, which allows numbers to be exact or inexact, almost as in Scheme. Originally intended for the Scheme exporter, it is now also used internally by Sifflet's interpreter. - The code is reorganized. There is now a separate library package, sifflet-lib, in addition to the sifflet package which provides only the executable. At present, the library's API is highly unstable. - Sifflet has a new parser for user inputs, made with Parsec, allowing more intuitive input of numbers (liberation from Haskell number syntax: for example, 0.0 can now be entered as 0.0, 0., or .0) and providing more intelligible error messages for some kinds of invalid input. - Sifflet requires Gtk2hs packages (gtk, glib, cairo) version 0.11; support for the 0.10 Gtk2hs packages is dropped. About Sifflet ------------- Sifflet is a visual, functional programming language intended as an aid for learning about recursion. * A picture explains Sifflet better than words: please see the screenshot showing how to evaluate 3!: http://mypage.iu.edu/~gdweber/software/sifflet/home.html * Features: - Visual editor. - Visual tracer/debugger which shows how recursive and other function calls are evaluated. To support active learning and avoid screen clutter, Sifflet displays only as much of the computation as the user requests. - A carefully crafted tutorial with 31 pictures and about as many pages. - Number, string, and list data types. - A small collection of primitive functions. - Runnable examples of compound functions. - (New!) Sifflet functions can be "exported" to Scheme, Python 3, and Haskell. * Download: http://hackage.haskell.org/package/sifflet-lib http://hackage.haskell.org/package/sifflet * Home page: http://mypage.iu.edu/~gdweber/software/sifflet/home.html * Tutorial: http://mypage.iu.edu/~gdweber/software/sifflet/doc/tutorial.html -- ___ ___ __ _ / _ \ / _ \| | | | Gregory D. Weber, Associate Professor / /_\// / | | | /\ | | Indiana University East / /_\\/ /__| | |/ \| | http://mypage.iu.edu/~gdweber/ \____/\_____/\___/\__/ Tel. (765) 973-8420; FAX (765) 973-8550

2 1

MOF and functional language paradigm
by Vasili I. Galchin 14 Aug '10

14 Aug '10

Hello, In my company (and I believe many, many companies) they use MOF heavily: http://en.wikipedia.org/wiki/Meta-Object_Facility MOF seems to be tightly coupled with object oriented program and has nothing to do with functional language approach?? Regards, Vasili

1 0

Compiled OpenGL program has DOS window?
by Eitan Goldshtrom 13 Aug '10

13 Aug '10

Hi. I'm working in Windows on an OpenGL application. I finally got everything working, doing all of my testing through the interpreter. When I finally compiled and ran the program from an executable for the first time I noticed that before creating my OpenGL window an empty DOS prompt popped up. I think I understand why the computer would want to do that, but I don't know how to make it stop. I've done some OpenGL in C++ and I remember having a similar problem. If I remember correctly the problem was fixed with a compiler flag. I'm using GHC and I looked through their compiler flags, but I didn't find anything that looked like it would deal with this. Anyone have any ideas? -Eitan

4 4

Beautiful Differentiation using VectorSpace
by Christopher Witte 13 Aug '10

13 Aug '10

Hi all, I'm trying to go through the paper Beautiful Differentiation by Conal Elliott and got a bit stuck. So I decided to download the VectorSpace package, which as far as I know implements the paper, and just try and play around with it. Unfortunately when I tried the simple examples at the beginning of the paper I got an error: Prelude Data.VectorSpace> :m Data.Derivative Prelude Data.Derivative> let f1 z = sqrt(3*sin z) Prelude Data.Derivative> f1 (D 2 1) Top level: Couldn't match expected type `Data.VectorSpace.Scalar a' against inferred type `Data.VectorSpace.Scalar (Data.VectorSpace.Scalar a)' NB: `Data.VectorSpace.Scalar' is a type function, and may not be injective <interactive>:1:0: No instances for (Data.MemoTrie.HasTrie (Data.Basis.Basis a), Floating (Data.VectorSpace.Scalar a), Data.VectorSpace.VectorSpace (Data.VectorSpace.Scalar a)) arising from a use of `f1' at <interactive>:1:0-9 Possible fix: add an instance declaration for (Data.MemoTrie.HasTrie (Data.Basis.Basis a), Floating (Data.VectorSpace.Scalar a), Data.VectorSpace.VectorSpace (Data.VectorSpace.Scalar a)) In the expression: f1 (D 2 1) In the definition of `it': it = f1 (D 2 1) <interactive>:1:8: No instance for (Num (a Data.LinearMap.:-* (a :> Data.VectorSpace.Scalar a))) arising from the literal `1' at <interactive>:1:8 Possible fix: add an instance declaration for (Num (a Data.LinearMap.:-* (a :> Data.VectorSpace.Scalar a))) In the second argument of `D', namely `1' In the first argument of `f1', namely `(D 2 1)' In the expression: f1 (D 2 1) Which was completely incomprehensible to me. I tried looking at the documentation but I couldn't find anything about actually using the library. Does anyone have any experience using this library and maybe have some examples of it's use? Thanks Chris Witte

1 0

Build failure for PortMidi on mac: incorrect Block.h
by Brian Victor 13 Aug '10

13 Aug '10

I'm trying to build PortMidi on the mac, and get the error shown below. It appears that an OSX header is trying to include a Block.h file and getting the GHC Block.h instead of the one it expects. I'm new to Haskell and even newer to cabal, so I'm not sure how to go about solving this. Incidentally, cabal, cabal-install, and the system PortMidi libraries are all up-to-date. %cabal install PortMidi Resolving dependencies... Configuring PortMidi-0.1.3... Preprocessing library PortMidi-0.1.3... Building PortMidi-0.1.3... [1 of 1] Compiling Sound.PortMidi ( Sound/PortMidi.hs, dist/build/Sound/PortMidi.o ) In file included from /System/Library/Frameworks/CoreServices.framework/Frameworks/CarbonCore.framework/Headers/FSEvents.h:28, from /System/Library/Frameworks/CoreServices.framework/Frameworks/CarbonCore.framework/Headers/CarbonCore.h:218, from /System/Library/Frameworks/CoreServices.framework/Frameworks/AE.framework/Headers/AE.h:20, from /System/Library/Frameworks/CoreServices.framework/Headers/CoreServices.h:21, from portmidi/pm_mac/pmmacosxcm.c:38:0: /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:51:0: error: expected specifier-qualifier-list before ‘StgPtr’ /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:108:0: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘bdescr’ /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:163:0: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘void’ /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:174:0: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘void’ /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:187:0: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘void’ /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:198:0: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘void’ /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:219:0: error: expected ‘)’ before ‘n’ /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:223:0: error: expected ‘)’ before ‘n’ /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:235:0: error: expected declaration specifiers or ‘...’ before ‘nat’ /Library/Frameworks/GHC.framework/Versions/610/usr/lib/ghc-6.10.4/include/Block.h:241:0: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘StgWord’ cabal: Error: some packages failed to install: PortMidi-0.1.3 failed during the building phase. The exception was: exit: ExitFailure 1 -- Brian

2 2

ANN: hswip-0.3 embedding swi-prolog in haskell
by Evgeny Tarasov 13 Aug '10

13 Aug '10

I would like to announce a small library hswip. It is an analogue of pyswip [1] python library in haskell. Hswip allows to embed swi-prolog [2] interpreter into your haskell applications in a convenient way. Let't look at an example: main = do prologInit ["-q","-nosignals"] prologCall "asserta(a(b,c))." prologCall "asserta(parent(pam, bob))." prologCall "asserta(parent(tom, bob))." prologCall "asserta(parent(tom, liz))." prologCall "asserta(parent(bob, ann))." prologCall "asserta(parent(bob, pat))." prologCall "asserta(parent(pat, jim))." res <- prologCall "parent(X,Y)." print res Output: [[("X","pat"),("Y","jim")],[("X","bob"),("Y","pat")],[("X","bob"),("Y","ann")],[("X","tom"),("Y","liz")],[("X","tom"),("Y","bob")],[("X","pam"),("Y","bob")]] The library still has alpha quality and there are many things missing including exception handling and BSD license. Any feedback is welcome! -Evgeny Tarasov [1] http://code.google.com/p/pyswip/ [2] http://www.swi-prolog.org/ [3] http://hackage.haskell.org/package/hswip-0.3

1 0

Twidge Short URL not working...
by arn-arn fernch 13 Aug '10

13 Aug '10

Hello, It's my first time using this mailing list thing so forgive me if I made any mistakes. I was using Twidge (http://wiki.github.com/jgoerzen/twidge/) and it's mentioned there is an available support in the Haskell Cafe. I only have a slight problem with Twidge. It's currently not shortening URLs by default though I can see in my .twidgerc this line "shortenurls: yes". Any things I should consider? Thanks!

1 0

universal quantification is to type instantiations as existential quantification is to what
by Joshua Ball 13 Aug '10

13 Aug '10

Hi, If I have a universally quantified type mapInt :: forall a. (Int -> a) -> [Int] -> [a] I can instantiate that function over a type and get a beta-reduced version of the type mapInt [String] :: (Int -> String) -> [Int] -> [String] (I'm borrowing syntax from Pierce here since I don't think Haskell allows me to explicitly pass in a type to a function.) This makes sense to me. The universal quantifier is basically a lambda, but it works at the type level instead of the value level. My question is... what is the analog to an existential type? mapInt :: exists a. (Int -> a) -> [Int] -> [a] (I don't think this is valid syntax either. I understand that I can rewrite this using foralls and a new type variable, doing something that looks like double negation, but the point of my question is to get an intuition for what exactly the universal quantifier means in the context of function application... if this even makes sense.) In particular: 1. If I can "instantiate" variables in a universal quantifier, what is the corresponding word for variables in an existential quantifier? 2. If type instantiation of a universally quantified variable is beta-reduction, what happens when existentially quantified variables are [insert answer to question 1]? 3. Is there any analogue to existential quantifiers in the value world? "Forall" is to "lambda" as "exists" is to what? Also (separate question), is the following statement true? forall T :: * -> *. (forall a. T a) -> (exists a. T a) If so, what does the implementation look like? (What function inhabits it, if it is interpreted as a type?) Josh "Ua" Ball

5 6

Re: Documentation for OpenGL?
by Eitan Goldshtrom 13 Aug '10

13 Aug '10

Woops, nevermind. I WAS looking in the wrong place. I should've known =P. Sorry for the unnecessary request. -Eitan PS. For anyone interested all of the documentation is in Graphics.Rendering.OpenGL.GL.PerFragment on hackage

1 0

Documentation for OpenGL?
by Eitan Goldshtrom 13 Aug '10

13 Aug '10

Hi everyone. I'm looking for something specific. I'm trying to figure out how to use the depthFunc StateVar and I can't find any official documentation for it. I did a ":t depthFunc" in ghci and got "depthFunc :: StateVar (Maybe ComparisonFunction)" so I looked up ComparisonFunction and couldn't find anything official on that either. Is there any documentation for it? Maybe I wasn't looking in the right places? Thanks in advance. -Eitan

1 0

How do you change the mouse position?
by Eitan Goldshtrom 13 Aug '10

13 Aug '10

In C++, maybe Java, I remember using a Robot to change the location of the mouse on the screen. My intention is to do something like in an FPS game where the mouse is always centered to make sure it doesn't run into the edges of the screen. How can I do that in Haskell? I'm using OpenGL, so if there is a utility that specifically works with OpenGL then I believe that would be the most helpful. -Eitan

2 2

ghc in macports
by Ozgur Akgun 13 Aug '10

13 Aug '10

Dear Cafe, I wonder who is maintaining the ghc package in macports, and what the current stategy of doing things is? http://www.macports.org/ports.php?by=name&substr=ghc (ghc 6.10.4) Personally, I'd like to use the macports version, if the ghc version there was resonably recent (having 2 versions, a stable and an edge could be a good idea?) Thanks, Ozgur

6 9

ADT patch/update
by Sergey Mironov 13 Aug '10

13 Aug '10

Hi Cafe. I am searching for materials on one data type-related problem. Suppose we have version control system storing values of user-defined algebraic data type. 'User' (actually, programmer) wants to store his/her data and later update it by applying patches. By patch I mean value of some (another) type representing one simplest update of stored value. Say, for tree type > data Tree a = Node (Tree a) (Tree a) | Leaf a one may write > {-# LANGUAGE TypeFamilies #-} > {-# LANGUAGE UndecidableInstances #-} > > class Storable a where > type Patch a :: * > update :: Patch a -> a -> a > > data TreeStep = Left | Right -- Like in zippers ? > type family Location a :: * > type instance Location (Tree a) = [TreeStep] > > instance (Storable a) => Storable (Tree a) where > type Patch (Tree a) = (Location (Tree a), Patch a) > update = ... -- Move to specific location and call node's update Patch here is coded explicitly but I guess there may be some generic way of it to calculate. Whether any similar work has been carried out? Will be glad to read comments! -- Thanks, Sergey

3 3

hackage.haskell.org cannot be accessed?
by Magicloud Magiclouds 12 Aug '10

12 Aug '10

Hi, Today I found out that I cannot access hackage.haskell.org. I have tried vpn/proxy to see if my network has something wrong. No luck. -- 竹密岂妨流水过山高哪阻野云飞

4 3

Workshop on ML 2010 - Call for Participation
by Matthew Fluet 12 Aug '10

12 Aug '10

The 2010 ACM SIGPLAN Workshop on ML http://www.cs.rit.edu/~mtf/ml2010 Baltimore, Maryland, United States Sunday, September 26, 2010 co-located with ICFP 2010 Call for Participation ML is a family of programming languages that includes dialects known as Standard ML, Objective Caml, and F#. The development of these languages has inspired a large amount of computer science research, both practical and theoretical. This workshop aims to provide a forum to encourage discussion and research on ML and related technology (higher-order, typed, or strict languages). The format of the 2010 Workshop on ML will be different than that of recent years, returning to a more informal model: a workshop with presentations selected from submitted abstracts but without published proceedings. We hope that this format will encourage the presentation of more exciting (if unpolished) research and deliver a more lively workshop atmosphere. Invited Speaker ~~~~~~~~~~~~~~~ Luke Hoban (Microsoft) -- Bringing F# to Visual Studio 2010 Program ~~~~~~~ 9:00 Invited Talk Visual F#: Bringing F# to Visual Studio 2010 Luke Hoban (Microsoft Research) 10:00 Break 10:30 Probabilistic programming using first-class stores and first-class continuations Oleg Kiselyov (FNMOC); Chung-chieh Shan (Rutgers University) 10:55 Effective progamming in ML Daan Leijen (Microsoft Research); Ross Tate (University of California, San Diego) 11:20 Discussion 11:30 First-class modules and composable signatures in Objective Caml 3.12 Alain Frisch (LexiFi); Jacques Garrigue (Nagoya University Graduate School of Mathematics) 11:55 First-class modules: hidden power and tantalizing promises Jeremy Yallop (Applicative Ltd); Oleg Kiselyov (FNMOC) 12:20 Discussion 12:30 Lunch break 14:00 Deriving a Typed Implementation for Coroutines in ML Konrad Anton (Universitat Freiburg); Peter Thiemann (Universitat Freiburg) 14:25 The Design Rationale for Multi-MLton Suresh Jagannathan (Purdue University); Armand Navabi (Purdue University); KC Sivaramakrishnan (Purdue University); Lukasz Ziarek (Purdue University) 14:50 Discussion 15:00 Mirage: high-performance ML kernels in the cloud Anil Madhavapeddy (University of Cambridge); Thomas Gazagnaire (INRIA Sophia Antipolis) 15:25 Hosting a Standard ML compiler in a Web Browser: Status Report Martin Elsman 15:50 Discussion 16:00 Break 16:30 A simple and effective method for assigning blame for type errors David MacQueen (University of Chicago) 16:55 The MetaOCaml files: Status report and research proposal Oleg Kiselyov (FNMOC); Chung-chieh Shan (Rutgers University) 17:20 Discussion 17:30 Closing Registration and Local Arrangements ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ * https://regmaster3.com/2010conf/ICFP10/register.php * http://www.icfpconference.org/icfp2010/local.html Program Chair ~~~~~~~~~~~~~ Matthew Fluet Rochester Institute of Technology Program Committee ~~~~~~~~~~~~~~~~~ Kathleen Fisher AT&T Labs Research Adam Granicz IntelliFactory Daan Leijen Microsoft Research Johan Nordlander Lulea University of Technology Sungwoo Park Pohang University of Science and Technology Daniel Spoonhower Google

1 0

Is there a pure Haskell gzip/bzip compression module out there?
by C K Kashyap 12 Aug '10

12 Aug '10

Hi All, I looked at Hackage and found a couple of pure Haskell modules for un-compression. Are there any for comression? I found Huffman compression - but what I need is something that I can uncompress with standard tools. -- Regards, Kashyap

3 3

A question about State Monad and Monad in general
by C K Kashyap 12 Aug '10

12 Aug '10

Hi, I looked at State Monad yesterday and this question popped into my mind. >From what I gather State Monad essentially allows the use of Haskell's do notation to "invisibly" pass around a state. So, does the use of Monadic style fetch us more than syntactic convenience? Again, if I understand correctly, in Mutable Arrays also, is anything getting modified in place really? If not, what is the real reason for better efficiency? -- Regards, Kashyap

14 26

Accepting and returning polyvariadic functions
by Will Jones 12 Aug '10

12 Aug '10

Hi all, I'm trying to write a function (I'll call it `vtuple' for lack of a better name) that returns a function that itself returns multiple arguments in the form of a tuple. For example: > {-# LANGUAGE FlexibleInstances #-} > {-# LANGUAGE FunctionalDependencies #-} > {-# LANGUAGE MultiParamTypeClasses #-} > f :: Int -> IO () > f = undefined > g :: Int -> Int -> IO () > g = undefined > h :: Int -> Int -> Int -> IO () > h = undefined vtuple f :: IO (Int -> (Int, ())) vtuple g :: IO (Int -> Int -> (Int, (Int, ()))) I've tried to type vtuple using a type class; my current effort is something like: > class VTuple ia ir a r | r -> a, a -> ia where > vtuple :: (ia -> ir) -> IO (a -> r) > instance VTuple Int (IO ()) Int (Int, ()) where > --vtuple :: (Int -> IO ()) -> IO (Int -> (Int, ())) > vtuple = undefined > instance VTuple ia ir a r > => VTuple Int (ia -> ir) Int (a -> (Int, r)) where > --vtuple :: (Int -> ia -> ir) -> IO (Int -> a -> (Int, r)) > vtuple = undefined But this is problematic, since arrows creep in: For one argument (fine): vtuple :: (Int -> IO ()) -> IO (Int -> (Int, ())) > vf :: IO (Int -> (Int, ())) > vf = vtuple f For two arguments (also fine): vtuple :: (Int -> Int -> IO ()) -> IO (Int -> Int -> (Int, (Int, ()))) > vg :: IO (Int -> Int -> (Int, (Int, ()))) > vg = vtuple g For three (noooo!): vtuple :: (Int -> Int -> IO ()) -> IO (Int -> Int -> (Int, (Int -> (Int32, (Int32, ()))))) And so on. I've thought about it and it seems impossible to solve this problem -- you keep needing to ``split'' the function type one arrow further on. Is this a job for Template Haskell or is there a solution I'm missing here? Note that I'd also like to use types other than Int, but I don't think this is the primary complication here (touch wood). Any help much appreciated, thanks, Will

4 5

Re: Haskell in Industry
by Tom Hawkins 12 Aug '10

12 Aug '10

Hi Eil, Sorry, your email got lost in my inbox. I hope you don't mind me copying haskell-cafe. > I saw a video of a presentation you gave at CUFP awhile back and was > hoping to ask you a couple of questions. > I'm currently a junior at UT Dallas and trying to figure out what I'm > going to do after finishing undergrad. For the past two years or so one > of my main interests has been language implementations. It's something > that I'd like to continue to work with. For a long time I thought I > wanted to stay in academia and do research, but recently I've realized > that I'm more interested in applying these advanced programming > languages to real-world problems, such as what you have done with the > Atom DSL. Good, we need more functional programmers actually solving real problems. But please put your skills to work in an industry other than investment banking. > > So, I'm curious about your experiences with functional programming in > industry. > Have you found it hard to get companies interested in Haskell? Yes, it is very hard. But it can also be very easy. The best approach is to find a job or industry that does not already have a large software component, but desperately needs one. This was the reason Haskell was successful at Eaton. Though Eaton is a huge company, this particular division had surprisingly few software developers, and even fewer managers who understood software development. In short, there was no one to question the move to Haskell. IMO, the biggest obstacle to Haskell are programmers -- especially shear numbers of them. > Is it difficult to hire good programmers for a less mainstream language? No. Every time I put a call out for resumes, I've gotten several candidates that would have fit the bill. The problem has been convincing my management to interview these leads. > Have you had any trouble training people to use Haskell? Yes. I find that out of 10 people I train, only about 2 pick it up and run with it. I'm starting to believe you are either wired for functional programming, or you're not. > Do many of the engineers using Haskell at Eaton have advanced degrees? Actually most of Eaton's Haskell programmers do not have advanced degrees; myself included. In fact one of our most successful programmers was an automotive technician in his previous job. This supports my hypothesis that your functional programming aptitude is determined at birth. Best of luck with the job hunt. If by chance Eaton starts hiring again, I'll give you a ping. -Tom

17 24

WGP Call for Participation
by Bruno Oliveira 12 Aug '10

12 Aug '10

====================================================================== CALL FOR PARTICIPATION WGP 2010 6th ACM SIGPLAN Workshop on Generic Programming Baltimore, Maryland, US Sunday, September 26th, 2010 http://osl.iu.edu/wgp2010 Collocated with the International Conference on Functional Programming (ICFP 2010) ====================================================================== Goals of the workshop --------------------- Generic programming is about making programs more adaptable by making them more general. Generic programs often embody non-traditional kinds of polymorphism; ordinary programs are obtained from them by suitably instantiating their parameters. In contrast with normal programs, the parameters of a generic program are often quite rich in structure; for example they may be other programs, types or type constructors, class hierarchies, or even programming paradigms. Generic programming techniques have always been of interest, both to practitioners and to theoreticians, and, for at least 20 years, generic programming techniques have been a specific focus of research in the functional and object-oriented programming communities. Generic programming has gradually spread to more and more mainstream languages, and today is widely used in industry. This workshop brings together leading researchers and practitioners in generic programming from around the world, and features papers capturing the state of the art in this important area. Program ------- * 09.00-10.00: Session 1 Chair: Marcin Zalewski * Welcome + PC chair report Bruno C. d. S. Oliveira and Marcin Zalewski * Outrageous but Meaningful Coincidences (Dependent type-safe syntax and evaluation) Conor McBride * 10.00-10.30: Tea/coffee * 10.30-12.30: Session 2 Chair: Shin-Cheng Mu * Scrap Your Zippers: A Generic Zipper for Heterogeneous Types Michael D. Adams * Generic Storage in Haskell Sebastiaan Visser and Andres Loeh * Generic Selections of Subexpressions Martijn van Steenbergen, José Pedro Magalhães and Johan Jeuring * 12.30-14.00: Lunch * 14.00-16.00: Session 3 Chair: Bruno C. d. S. Oliveira * Generic Multiset Programming for Language-Integrated Querying Fritz Henglein and Ken Friis Larsen * Algorithms for Traversal-Based Generic Programming Bryan Chadwick and Karl Lieberherr * Ad-hoc Polymorphism and Dynamic Typing in a Statically Typed Functional Language Thomas van Noort, Peter Achten and Rinus Plasmeijer * 16.00-16.30: Tea/coffee * 16.30-18.00: Session 4 Chair: Conor McBride * Reason Isomorphically! Ralf Hinze and Daniel James * Constructing Datatype-Generic Fully Polynomial-Time Approximation Schemes Using Generalised Thinning Shin-Cheng Mu, Yu-Han Lyu and Akimasa Morihata Workshop homepage: http://osl.iu.edu/wgp2010/ Registration link: https://regmaster3.com/2010conf/ICFP10/register.php Local arrangements: http://www.icfpconference.org/icfp2010/local.html

1 0

Can we come out of a monad?
by C K Kashyap 11 Aug '10

11 Aug '10

Hi, In the code here - http://hpaste.org/fastcgi/hpaste.fcgi/view?id=28393#a28393 If I look at the type of modifiedImage, its simply ByteString - but isn't it actually getting into and back out of the state monad? I am of the understanding that once you into a monad, you cant get out of it? Is this breaking the "monad" scheme? -- Regards, Kashyap

29 77

cabal haddock doesn't work for me on win7
by Ryan Ingram 11 Aug '10

11 Aug '10

I'm not sure what's going on, here's the error I get: C:\haskell\ListZipper-1.1.1.0>cabal haddock Running Haddock for ListZipper-1.1.1.0... Preprocessing library ListZipper-1.1.1.0... Warning: The documentation for the following packages are not installed. No links will be generated to these packages: Cabal-1.8.0.2, QuickCheck-2.1.0.3, Win32-2.2.0.1, array-0.3.0.0, base-3.0.3.2, base-4.2.0.0, bin-package-db-0.0.0.0, ffi-1.0, rts-1.0, bytestring-0.9.1.5, containers-0.3.0.0, directory-1.0.1.0, extensible-exceptions-0.1.1.1, filepath-1.1.0.3, ghc-6.12.1, ghc-binary-0.5.0.2, ghc-prim-0.2.0.0, hpc-0.5.0.4, integer-gmp-0.2.0.0, mtl-1.1.0.2, old-locale-1.0.0.2, old-time-1.0.0.3, pretty-1.0.1.1, process-1.0.1.2, random-1.0.0.2, syb-0.1.0.2, template-haskell-2.4.0.0, time-1.1.4 Warning: Data.List.Zipper: could not find link destinations for: GHC.Classes.Eq GHC.Show.Show Test.QuickCheck.Arbitrary.Arbitrary Test.QuickCheck.Arbitrary.CoArbitrary GHC.Bool.Bool Data.Maybe.Maybe haddock: internal Haddock or GHC error: E:\Program Files (x86)\Haskell\haddock-2.7.2\html\haddock.css: openFile: does not exist (No such file or directory) Haskell platform is installed on "C:", I have no idea where it's getting the "E:" from. I can specify the correct css location with --css, but then I get the same error looking for haskell_icon.gif. -- ryan

2 2

ANN improve-0.0.4
by Tom Hawkins 11 Aug '10

11 Aug '10

Hi, ImProve [1] is a little imperative DSL that compiles to C code. Intended for high assurance embedded applications, ImProve is also an infinite state, unbounded model checker. Meaning ImProve can verify assertions in a program will always be true. Here's an example: module Main where import Language.ImProve launchControl :: Stmt () launchControl = do presidentialAuthority <- input bool "presidentialAuthority" buttonPressed <- input bool "buttonPressed" launchMissiles <- bool' "launchMissiles" $ (presidentialAuthority ||. buttonPressed) assert "conditionsToLaunch" $ launchMissiles ==. (presidentialAuthority &&. buttonPressed) main :: IO () main = verify "yices" 20 launchControl Running this program yields: verifying conditionsToLaunch FAILED: disproved basis in k = 1 (see trace) On verification failures, ImProve will emit a counter example showing the violation, which is essentially a program trace showing how things could go wrong. In this case, the trace looks like this: initialize launchMissiles := false cycle 0 input presidentialAuthority <== true input buttonPressed <== false launchMissiles <== true assertion FAILED: conditionsToLaunch This basically means Barack was somehow able to launch on his own. (Man! I even voted for the guy!) Under the hood, ImProve uses Yices [2] for SMT solving, and a method of unbounded model checking known as k-induction. Don't expect too much from ImProve; it's a very restrictive language. It's basically three data types (bool, int, float), variables assignments, and 'if' statements. No procedures, no loops, no arrays, no pointers, etc. However, it can still do interesting things -- especially with Haskell combinators in play. Hope you find it useful. And as always, questions, comments, and feedback welcome. -Tom [1] http://hackage.haskell.org/package/improve [2] http://yices.csl.sri.com/

1 0

Is there any experience using Software Transactional Memory in substantial applications?
by Serguey Zefirov 11 Aug '10

11 Aug '10

Recently we discussed Haskell and especially types in Russian part of LiveJournal and of course we talk about STM. My opponent gave me that link: http://logicaloptimizer.blogspot.com/2010/06/so-microsofts-experiments-with… It says that performance with STM in Microsoft Research was more than horrible. I failed to find convincing counter-evidence on the web. Not for Haskell, even for Java/C#/C++. So I asking Haskell-cafe for clarification. Do anyone here have an experience with STM in computing-intensive tasks? Did it help there? What are the possible reasons for STM failure in MR?

3 2

Re: [Haskell-cafe] Can we come out of a monad?
by John Lato 11 Aug '10

11 Aug '10

> From: Alexey Khudyakov <alexey.skladnoy(a)gmail.com> > > On Fri, 30 Jul 2010 09:29:59 +0200 > Stefan Holdermans <stefan(a)vectorfabrics.com> wrote: > >> > No I think here we breaking out from _arbitrary_ monad. If monadic > function works for every monad then it must work for identity monad > too. Here is simplest form of purify function: > >> purify2 :: (forall m . Monad m => m a) -> a >> purify2 m = runIdentity m > > I wonder could this function be written without resorting to concrete > monad I don't find purify2 particularly helpful because I almost never want to break out of any arbitrary monad; I want to be able to break out of a specific monad without knowing which monad it is, that is: purify3 :: Monad m => m a -> a purify3 = undefined --the only possible definition... However, I just realized that something else is almost as good: evalCont :: Cont r a -> r evalCont k = runCont k id As Cont is the "Mother of all monads", it can be used to emulate the behavior of any other monad. If you had a library with instances of MonadReader, MonadWriter, MonadState, etc. (or operational equivalents) defined for the Cont monad, then you would have a purify function that works for all interesting monads (except IO, STM, and maybe ST), which is almost as good. John

5 9