Re: [Haskell-cafe] sendfile leaking descriptors on Linux?
I believe this might be causing problems for me with patch-tag. me: Like mightybyte, I run my app in a shell loop that will just restart it after a crash. But every once in a while it won't restart because of the busy socket and I need to do a manual restart, killing multiple processes (usually 2). the error I get is: bind: resource busy (Address already in use) This sounds like what is being discussed. I have some questions. the server will eventually run out of file descriptors How can I tell if I am running out of file descriptors? If I use the portable build of the sendfile package, everything works fine for hours and hours of this happening. What is the portable build and how do I use it? Do you eventually have troubles here as well? What is the reason to use the linux native build then? speed / more connections per second? thanks for your help! 2010/2/4 Bardur Arantsson s...@scientician.net: Hi all, I've been using the sendfile package off Hackage, but it seems to be leaking file descriptors when using the Linux-native build. What's happening in my specific case is the following: 1) client requests a range of a file 2) server starts sending the range 3) client disconnects before receiving the whole file This happens over and over with the client requesting different ranges of the file (so the client does make progress). If I use the portable build of the sendfile package, everything works fine for hours and hours of this happening. If I use the Linux-native build of the sendfile package, the server will eventually run out of file descriptors. According to lsof the files that are being kept open are the data files being sent in 2) above. This is on GHC 6.10.x (Ubuntu). Is anyone else seeing this? Anyone got any idea what's going on? Cheers, ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Linear programming in Haskell
Louis Wasserman wrote: Yo, Man, I'd never used FFI before, but it's really not as scary as I'd feared. The FFI is fantastic. We can even use C higher order functions with normal Haskell function arguments... :) I've implemented a more comprehensive interface to GLPK's simplex solver and -- rather importantly, for my own needs -- its MIP solver. This doesn't depend on hmatrix, and in fact, it doesn't require any matrix or vector manipulation at all -- linear functions are specified as a straight-up Data.Map from an arbitrary variable type to their coefficients. I like your interface, it is very complete and user friendly. (I used hmatrix because of (my own) laziness, to take advantage of some utilities, but of course it is not really required.) Thanks for your work! Alberto The library is now available as glpk-hs on hackage. Example: import Data.LinearProgram.LPMonad import Data.LinearProgram import Data.LinearProgram.GLPK objFun :: LinFunc String Int objFun = linCombination [(10, x1), (6, x2), (4, x3)] lp :: LP String Int lp = execLPM $ dosetDirection Max setObjective objFun leqTo (varSum [x1, x2, x3]) 100 leqTo (10 *^ var x1 ^+^ 4 * x2 ^+^ 5 *^ var x3) 600 -- c *^ var v, c * v, and linCombination [(c, v)] are all equivalent. -- ^+^ is the addition operation on linear functions. leqTo (linCombination [(2, x1), (2, x2), (6, x3)]) 300 varGeq x1 0 varBds x2 0 50 varGeq x3 0 setVarKind x1 IntVar setVarKind x2 ContVar main = print = glpSolveVars mipDefaults lp This requires GLPK to be installed, like below. Louis Wasserman wasserman.lo...@gmail.com mailto:wasserman.lo...@gmail.com http://profiles.google.com/wasserman.louis On Wed, Feb 24, 2010 at 4:07 AM, Alberto Ruiz ar...@um.es mailto:ar...@um.es wrote: I have uploaded to hackage an interface to the simplex algorithm based on GLPK. It is a very early version, it will probably have lots of problems. In the future I would like to add support for integer variables (MIP). Any suggestion is welcome. This is an example taken from glpk-utils: http://code.haskell.org/hmatrix/packages/glpk/examples/simplex3.hs Documentation: http://perception.inf.um.es/~aruiz/hmatrix-glpk/ http://perception.inf.um.es/%7Earuiz/hmatrix-glpk/ Installation: $ sudo apt-get install libglpk-dev $ cabal update $ cabal install hmatrix-glpk If hmatrix is not installed we also need $ sudo apt-get install libgsl0-dev liblapack-dev I hope it is useful, Alberto Erik de Castro Lopo wrote: Alberto Ruiz wrote: I think that GSL does not include linear programming solvers, but in the GSL home page there is a reference to the GLPK package: http://www.gnu.org/software/glpk/glpk.html I have not used it, but it would be very nice to have a simple Haskell interface to GLPK (or other similar library) in hmatrix or as a separate package. I will take a look at this. I used GLPK many years ago and I found it excellent. Erik ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] hmake and hat
Hi, Am Donnerstag, den 25.02.2010, 20:48 -0800 schrieb Jason Dusek: Can you provide a link to something describing the error? Well, here are our patches to make hmake compile: http://patch-tracker.debian.org/package/hmake/3.14-2 And building hat results in the attached buildlog. Greetings, Joachim -- Joachim nomeata Breitner Debian Developer nome...@debian.org | ICQ# 74513189 | GPG-Keyid: 4743206C JID: nome...@joachim-breitner.de | http://people.debian.org/~nomeata dpkg-buildpackage -rfakeroot -D -us -uc dpkg-buildpackage: setze CFLAGS auf Standardwert: -g -O2 dpkg-buildpackage: setze CPPFLAGS auf Standardwert: dpkg-buildpackage: setze LDFLAGS auf Standardwert: dpkg-buildpackage: setze FFLAGS auf Standardwert: -g -O2 dpkg-buildpackage: setze CXXFLAGS auf Standardwert: -g -O2 dpkg-buildpackage: Quellpaket hat dpkg-buildpackage: Quellversion 2.05+rerolled-7 dpkg-buildpackage: Quellen geÀndert durch Arjan Oosting ar...@debian.org dpkg-buildpackage: Host-Architektur amd64 fakeroot debian/rules clean dh_testdir dh_testroot rm -f configure-stamp build-stamp [ ! -f lib/x86_64-Linux/config ] || /usr/bin/make clean Warning: package(s) packedstring not available (according to ghc-pkg) MkProg: Can't find module IO in user directories . src/hattrans src/compiler98 Or in installed libraries/packages at /usr/lib/ghc-6.10.4/base-3.0.3.1 /usr/lib/ghc-6.10.4/base-4.1.0.0 /usr/lib/ghc-6.10.4/haskell98-1.0.1.0 /home/jojo/.cabal/lib/base64-string-0.1/ghc-6.10.4 Asked for by: src/hattrans/HatTrans.hs Fix using the -I, -P, or -package flags. Stop - hmake dependency error. Warning: package(s) packedstring not available (according to ghc-pkg) MkProg: Can't find module IO in user directories . src/hattrans src/compiler98 Or in installed libraries/packages at /usr/lib/ghc-6.10.4/base-3.0.3.1 /usr/lib/ghc-6.10.4/base-4.1.0.0 /usr/lib/ghc-6.10.4/haskell98-1.0.1.0 /home/jojo/.cabal/lib/base64-string-0.1/ghc-6.10.4 Asked for by: src/hattrans/HatTrans.hs Fix using the -I, -P, or -package flags. Stop - hmake dependency error. make[1]: Entering directory `/tmp/hat-2.05+rerolled' cd src/hattrans; /usr/bin/make clean make[2]: Entering directory `/tmp/hat-2.05+rerolled/src/hattrans' rm -f /tmp/hat-2.05+rerolled/targets/x86_64-Linux/obj/hattrans/*. *. *.hi *.hc make[2]: Leaving directory `/tmp/hat-2.05+rerolled/src/hattrans' cd src/hatlib; /usr/bin/make clean make[2]: Entering directory `/tmp/hat-2.05+rerolled/src/hatlib' rm -rf /tmp/hat-2.05+rerolled/targets/x86_64-Linux/obj/hatlib/ghc6 Hat/PreludeBuiltin.hs Hat/PreludeBasic.hs Hat/IOBuiltinTypes.hs Hat/SystemBuiltinTypes.hs Hat/TimeBuiltinTypes.hs Hat/DirectoryBuiltinTypes.hs Hat/Char.hs Hat/Ratio.hs Hat/List.hs Hat/Complex.hs Hat/Numeric.hs Hat/Maybe.hs Hat/Monad.hs Hat/Ix.hs Hat/Array.hs Hat/IO.hs Hat/System.hs Hat/Locale.hs Hat/CPUTime.hs Hat/Time.hs Hat/Random.hs Hat/Directory.hs Hat/Data/IORef.hs Hat/Data/Bool.hs Hat/Data/Char.hs Hat/Data/Complex.hs Hat/Data/Ix.hs Hat/Data/Array.hs Hat/Data/Bits.hs Hat/Data/Either.hs Hat/Data/Maybe.hs Hat/Data/FiniteMap.hs Hat/Data/List.hs Hat/Data/Ratio.hs Hat/Data/Set.hs Hat/Data/Tuple.hs Hat/Data/PackedString.hs Hat/Data/Monoid.hs Hat/Data/Word.hs Hat/Control/Monad.hs Hat/Control/Monad/Fix.hs Hat/Control/Arrow.hs Hat/Control/Monad/Identity.hs Hat/Control/Monad/Trans.hs Hat/System/CPUTime.hs Hat/System/Cmd.hs Hat/System/Random.hs Hat/System/Directory.hs Hat/System/Environment.hs Hat/System/Exit.hs Hat/System/Locale.hs Hat/System/Time.hs Hat/System/IO/Unsafe.hs Hat/System/IO.hs Hat/System/IO/Error.hs Hat/System/Console/GetOpt.hs Hat/Debug/Trace.hs Hat/Test/QuickCheck.hs Hat/Text/Show.hs Hat/Text/Read.hs Hat/Text/PrettyPrint/HughesPJ.hs Hat/Text/PrettyPrint.hs Hat/Text/ParserCombinators/Parsec/Char.hs Hat/Text/ParserCombinators/Parsec/Combinator.hs Hat/Text/ParserCombinators/Parsec/Error.hs Hat/Text/ParserCombinators/Parsec/Expr.hs Hat/Text/ParserCombinators/Parsec/Pos.hs Hat/Text/ParserCombinators/Parsec/Prim.hs Hat/Text/ParserCombinators/Parsec.hs Hat/Foreign/Ptr.hs Hat/Foreign/StablePtr.hs Hat/Foreign/ForeignPtr.hs Hat/Foreign/Storable.hs Hat/Foreign/Marshal/Error.hs Hat/Text/ParserCombinators/Parsec/Perm.hs Hat/Control/Monad/Reader.hs Hat/Control/Monad/Writer.hs Hat/Control/Monad/State.hs Hat/Control/Monad/RWS.hs Hat/Control/Monad/Cont.hs Hat/Control/Monad/Error.hs Hat/Control/Monad/List.hs Hat/Prelude.hs Hat/*.hi Hat/Hat.hx \ PreludeBuiltin.hx PreludeBasic.hx IOBuiltinTypes.hx SystemBuiltinTypes.hx TimeBuiltinTypes.hx DirectoryBuiltinTypes.hx Char.hx Ratio.hx List.hx Complex.hx Numeric.hx Maybe.hx Monad.hx Ix.hx Array.hx IO.hx System.hx Locale.hx CPUTime.hx Time.hx Random.hx Directory.hx Data/IORef.hx Data/Bool.hx Data/Char.hx
Re: [Haskell-cafe] sendfile leaking descriptors on Linux?
On Feb 26, 2010, at 04:28 , Thomas Hartman wrote: me: Like mightybyte, I run my app in a shell loop that will just restart it after a crash. But every once in a while it won't restart because of the busy socket and I need to do a manual restart, killing multiple processes (usually 2). the error I get is: bind: resource busy (Address already in use) This is on application restart? It's not out of file descriptors, it's just the system keeping the socket around (netstat will show it in TIME_WAIT, or possibly in a shutdown negotiation state such as LAST_ACK, FIN_WAIT, etc. TCP lacks *reliable* socket shutdown negotiation). You want to configure the socket with SO_REUSEADDR before trying to bind it (setSocketOption socket ReuseAddr 1). As to the portable version of sendfile, it's because not all systems offer a sendfile() system call. Linux and *BSD do, and can use the native implementation; the portable version emulates sendfile() when it doesn't exist, at the price of additional CPU usage/system load (sendfile() having been created specifically to reduce system load in the common case for web servers). -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu electrical and computer engineering, carnegie mellon universityKF8NH PGP.sig Description: This is a digitally signed message part ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] problem loading package unix-2.3.2.0 while installing hackages on OpenSuse
I have OpenSuse 11.2 running on VMWare and I have recently installed ghc 6.10.4 from the RPM from OpenSuse. Now I want to install cabal. I have downloaded the required packages from hackage and I try to run: runhaskell Setup configure on any of these packages. With option -v I can see an error while linking unix-2.3.2.0. I am no expert on OpenSuse yet. Does anyone know what's happening here? linux-nrvh:/home/gmi/Download/cabal/zlib-0.5.2.0 # runhaskell -f /usr/bin/ghc -v Setup configure --with-compiler=ghc-6.10.4 ... Loading package filepath-1.1.0.2 ... linking ... done. Loading package unix-2.3.2.0 ... Setup: command line: can't load .so/.DLL for: dl (libdl.so: cannot open shared object file: No such file or directory) -- View this message in context: http://old.nabble.com/problem-loading-package-unix-2.3.2.0-while-installing-hackages-on-OpenSuse-tp27717114p27717114.html Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Re: Proper round-trip HughesPJ/Parsec for Doubles?
Andy Gimblett schrieb: Hi Christian, [...] It may make sense to use something like readMaybe (which is missing in the Prelude) instead of read to allow the parser to fail more nicely. It seems to be kicking up reasonable errors as it is, e.g.: *Main parse aFloat 2e-h Left (line 1, column 4): unexpected h expecting digit yes, this is fine, because you reject -h, but suppose it was passed to read (due to a programming error). I haven't seen any uncaught exceptions propagating, if that's what you're worried about...? Yes, read will always work for you. But you could use a parser *almost* as simple as many1 $ oneOf NaInfity+-.eE0123456789 and ask readMaybe if the parsed String can be read as Double. So here's what I have now: float' :: TokenParser st - GenParser Char st Double float' t = do n - maybeChar - spaces fs - choice [symbol t NaN, symbol t Infinity, do whole - many1 digit frac - option $ do char '.' ds - many1 digit return $ '.' : ds ex - option $ do choice [char 'e', char 'E'] s - maybeChar +- ds - many1 digit return $ concat [e, s, ds] return $ concat [whole, frac, ex] ] whiteSpace t return $ read $ n ++ fs where maybeChar :: String - GenParser Char st String maybeChar as = option (choice (map char as) = \a - return [a]) I would omit handling of spaces (that's a separate lexing step). It's enough to be able to parse those numbers, that are possible results of show (for round-trip). symbol t could be replaced by (try . string) in order to get rid of the TokenParser (that I don't like). Spaces following an initial minus sign are quite unusual and rather indicate that the sign does not belong to number, but that the sign is a separate operation. You can also break it immediately before do, which I think is sometimes more clear. If not an extra space is added following do this leads to an odd indentation of at least one line. I'm curious: which line in the above is indented oddly? Oh, wait: you don't mean odd as in strange, do you? You mean odd as in not even? So, e.g. the spaces line starts at column 5? What's wrong with that? Right, again a matter of taste. Cheers Christian P.S. below is my parser for tptp numbers (as comparison). It rejects leading zeros, but allows an initial + sign. fmap read real would work if the input does not start with + (or is NaN or Infinity). -- | does not allow leading zeros natural :: Parser String natural = string 0 | many1 digit decimal :: Parser String decimal = do s - option $ string + | string - ds - natural return $ s ++ ds real :: Parser String real = do d - decimal f - option $ do p - char '.' n - many1 digit return $ p : n e - option $ do x - char 'e' | char 'E' g - decimal return $ x : g return $ d ++ f ++ e ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] problem loading package unix-2.3.2.0 while installing hackages on OpenSuse
Am Freitag 26 Februar 2010 12:52:36 schrieb gmi: I have OpenSuse 11.2 running on VMWare and I have recently installed ghc 6.10.4 from the RPM from OpenSuse. Now I want to install cabal. I have downloaded the required packages from hackage and I try to run: In principle, just unpacking the cabal-install package and running bootstrap.sh is far far easier. runhaskell Setup configure on any of these packages. With option -v I can see an error while linking unix-2.3.2.0. I am no expert on OpenSuse yet. Does anyone know what's happening here? linux-nrvh:/home/gmi/Download/cabal/zlib-0.5.2.0 # runhaskell -f /usr/bin/ghc -v Setup configure --with-compiler=ghc-6.10.4 ... Loading package filepath-1.1.0.2 ... linking ... done. Loading package unix-2.3.2.0 ... Setup: command line: can't load .so/.DLL for: dl (libdl.so: cannot open shared object file: No such file or directory) You seem to not have libdl.so, which is needed by the unix package. Can you confirm this by running $ locate libdl (or $ ls /usr/lib | grep libdl $ ls /lib | grep libdl if you don't have [the rights to run] locate)? If that doesn't show something like /lib/libdl-2.9.so /lib/libdl.so.2 /usr/lib/libdl.a /usr/lib/libdl.so , you'll have to install glibc-devel, which provides /usr/lib/libdl.a and /usr/lib/libdl.so (the others are provided by glibc, but I can't imagine that you haven't that installed). And while you're in YaST, installing software, grab almost every *-devel package you see, unless you're sufficiently sure that you will not need that one - installing one more devel-package a week becomes annoying pretty soon. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Re: do we need types?
A type is, well, a type. A type class is a relation between types. The confusion probably comes from OO programming where (interfaces) describe unary relations (= one parameter type classes). (begin rant) I wouldn't easily give up algebraic data types just because OO programmers don't seem know about them - so the re-invent them and call it composite pattern... ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] do we need types?
I'd say we don't really need subclasses. I mean, what's the difference:
class Eq a where (==) :: a - a - Bool
instance Eq a = Eq (Maybe a) where
Nothing == Nothing = True
Just x == Just y = x == y
_ == _ = False
sort :: Eq a = [a] - [a]
or
data Eq a = Eq {eq :: a - a - Bool}
eqMaybe :: Eq a - Eq (Maybe a)
eqMaybe e = Eq {eq = eqM} where
eqM Nothing Nothing = True
eqM (Just x) (Just y) = eq e x y
eqM _ _ = False
sort :: Eq a - [a] - [a]
Replacing classes with types, we only lose one thing: the compiler won't deduce
the right instances for us. I'll trade it for the ability to abstract over
them. After all, we CAN deduce the right
instances by hand, it's just a finite amount of work (not very big, in my
experience).
Pasqualino Titto Assini wrote:
Hi, just a silly question (or maybe more than one):
In Haskell we have data types (Integer,[a],...) as well as type
classes (Num, Ord...).
But, if we have type classes do we still need types?
Why shouldn't the objects that we process be defined only by their
'interfaces' (assuming that a type class is a kind of interface)?
Maybe the real question is: are type classes a more primitive concept
than data types?
And if so, in a language that had only type classes what would a data
declaration like the following map to:
data List a = Cons a (List a) | Nil
And what about pattern matching? Would that still be possible, and
what form would it take?
And finally, would having only type classes make the type system any simpler?
Thanks,
titto
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] do we need types?
s/subclasses/classes/
Sorry for the confusion.
Miguel Mitrofanov wrote:
I'd say we don't really need subclasses. I mean, what's the difference:
class Eq a where (==) :: a - a - Bool
instance Eq a = Eq (Maybe a) where
Nothing == Nothing = True
Just x == Just y = x == y
_ == _ = False
sort :: Eq a = [a] - [a]
or
data Eq a = Eq {eq :: a - a - Bool}
eqMaybe :: Eq a - Eq (Maybe a)
eqMaybe e = Eq {eq = eqM} where
eqM Nothing Nothing = True
eqM (Just x) (Just y) = eq e x y
eqM _ _ = False
sort :: Eq a - [a] - [a]
Replacing classes with types, we only lose one thing: the compiler won't
deduce the right instances for us. I'll trade it for the ability to
abstract over them. After all, we CAN deduce the right
instances by hand, it's just a finite amount of work (not very big, in
my experience).
Pasqualino Titto Assini wrote:
Hi, just a silly question (or maybe more than one):
In Haskell we have data types (Integer,[a],...) as well as type
classes (Num, Ord...).
But, if we have type classes do we still need types?
Why shouldn't the objects that we process be defined only by their
'interfaces' (assuming that a type class is a kind of interface)?
Maybe the real question is: are type classes a more primitive concept
than data types?
And if so, in a language that had only type classes what would a data
declaration like the following map to:
data List a = Cons a (List a) | Nil
And what about pattern matching? Would that still be possible, and
what form would it take?
And finally, would having only type classes make the type system any
simpler?
Thanks,
titto
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] problem loading package unix-2.3.2.0 while installing hackages on OpenSuse
Thanks a lot. It was all about installing *-devel things. -- View this message in context: http://old.nabble.com/problem-loading-package-unix-2.3.2.0-while-installing-hackages-on-OpenSuse-tp27717114p27718263.html Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Re: do we need types?
A type class is a relation between types. Yes, but not officially, just de facto: http://hackage.haskell.org/trac/haskell-prime/ticket/90 Best, Maurício ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: GUI (was: Re: [Haskell-cafe] DLL on Windows)
Am Mittwoch 17 Februar 2010 12:17:10 schrieb Jeremy O'Donoghue: You're probably correct about the dependencies. I have never tried to compile wxHaskell against GHC 6.12.1 I'm waiting for Haskell Platform to be released to make the required changes since (working primarily on Windows) I just don't have time to create a complete GHC 6.12 installation with most of the HP libraries (some of which are a pain to get working on Windows). wxHaskell will support GHC 6.12.x within a couple of days of release of a suitable Haskell Platform. Regards Jeremy Beating a dead horse, but today I decided to see whether I could install wxHaskell, all it took was a few edits in some .cabal files and one in wxcore's Setup to make it all build with 6.12.1. Still, it's annoying because you have to unpack everything, edit and cabal install in the unpacked directories. Now since I did that, I could upload new verxions of wxdirect, wxcore and wx to hackage, so others could do a simple cabal install wx. But I would only do that if I get a go-ahead from somebody authorised (since Jeremy uploaded the previous versions, he'd be qualified, I think). Or I could send the edits to Jeremy and let him do the uploading, whatever is preferred. On Tue, 16 Feb 2010 17:55 -0800, Thomas DuBuisson thomas.dubuis...@gmail.com wrote: On Tue, Feb 16, 2010 at 3:48 PM, Henk-Jan van Tuyl hjgt...@chello.nl wrote: On Tue, 16 Feb 2010 18:57:20 +0100, Neil Mitchell ndmitch...@gmail.com wrote: I used to recommend Gtk2hs over wxHaskell for GUI development as there was always a version that worked on Windows with the latest GHC release. I think I might have to switch back to recommending C# for GUI development... The latest revision of wxHaskell can be compiled relatively easy, so that wxHaskell can be used immediately for the latest version of GHC. Exciting allegation, but it doesn't quite check out with GHC 6.12.1 + cabal 1.8.0.2 and cabal-install 0.8 (see below). I would expect this issue to be easily resolved by specifying containers = 0.3.* in the wxdirect package, but cabal is misbehaving (for me) and stating wxdirect needs reinstalled with a supposedly newer version of containers, 0.2.0.1. Thomas [to...@mavlo ~]$ cabal install wx Resolving dependencies... Downloading containers-0.2.0.1... Configuring containers-0.2.0.1... Preprocessing library containers-0.2.0.1... Building containers-0.2.0.1... Data/IntMap.hs:182:7: Could not find module `Data.Data': It is a member of the hidden package `base'. Perhaps you need to add `base' to the build-depends in your .cabal file. Use -v to see a list of the files searched for. cabal: Error: some packages failed to install: containers-0.2.0.1 failed during the building phase. The exception was: ExitFailure 1 wx-0.12.1.2 depends on containers-0.2.0.1 which failed to install. wxcore-0.12.1.2 depends on containers-0.2.0.1 which failed to install. wxdirect-0.12.1.1 depends on containers-0.2.0.1 which failed to install. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Generalizing nested list comprehensions
Hi all, If this question sounds a bit noob-ish, that would be because I am one - so apologies in advance! I have functions that look something like these: f1 :: [a] - [b] f1 xs = [foo [x1, x2] | x1 - xs, x2 - bar x1, baz x1 /= baz x2] f2 :: [a] - [b] f2 xs = [foo [x1, x2, x3] | x1 - xs, x2 - bar x1, x3 - bar x2, baz x1 /= baz x2, baz x1 /= baz x3, baz x2 /= baz x3] f3 :: [a] - [b] f3 xs = [foo [x1, x2, x3, x4] | x1 - xs, x2 - bar x1, x3 - bar x2, x4 - bar x3, baz x1 /= baz x2, baz x1 /= baz x3, baz x1 /= baz x4, baz x2 /= baz x3, baz x1 /= baz x4] Note that for the purposes of this discussion it does not matter what foo, bar and baz do. Now what I want to do is write a generalized function fn based on the pattern set by f1, f2 and f3 such that: fn :: Int - [a] - [b] and: (fn 1 xs) == f1 xs (fn 2 xs) == f2 xs (fn 3 xs) == f3 xs (fn 25 xs) == f25 xs - obviously if I were to implement f25 as nested list comprehensions it would be ridiculously tedious! Any ideas how I can implement fn? Thanks, Ishaaq -- View this message in context: http://old.nabble.com/Generalizing-nested-list-comprehensions-tp27719199p27719199.html Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Generalizing nested list comprehensions
Am Freitag 26 Februar 2010 15:52:15 schrieb Ishaaq Chandy: Hi all, If this question sounds a bit noob-ish, that would be because I am one - so apologies in advance! I have functions that look something like these: f1 :: [a] - [b] f1 xs = [foo [x1, x2] | x1 - xs, x2 - bar x1, baz x1 /= baz x2] f2 :: [a] - [b] f2 xs = [foo [x1, x2, x3] | x1 - xs, x2 - bar x1, x3 - bar x2, baz x1 /= baz x2, baz x1 /= baz x3, baz x2 /= baz x3] f3 :: [a] - [b] f3 xs = [foo [x1, x2, x3, x4] | x1 - xs, x2 - bar x1, x3 - bar x2, x4 - bar x3, baz x1 /= baz x2, baz x1 /= baz x3, baz x1 /= baz x4, baz x2 /= baz x3, baz x1 /= baz x4] Note that for the purposes of this discussion it does not matter what foo, bar and baz do. Now what I want to do is write a generalized function fn based on the pattern set by f1, f2 and f3 such that: fn :: Int - [a] - [b] and: (fn 1 xs) == f1 xs (fn 2 xs) == f2 xs (fn 3 xs) == f3 xs (fn 25 xs) == f25 xs - obviously if I were to implement f25 as nested list comprehensions it would be ridiculously tedious! Any ideas how I can implement fn? The easy way is recursion: import Control.Monad (guard) fn n xs = do x1 - xs let b1 = baz x1 contFn n [] [b1] x1 contFn 0 acc _ xn = [foo (reverse (xn:acc)] contFn n acc bs xi = do xj - bar xi let bj = baz xj guard (all (/= bj) bs) contFn (n-1) (xi:acc) (bj:bs) xj Now, there is probably a frighteningly elegant way to do it with foldM or somesuch, but I don't see it at the moment :( Thanks, Ishaaq ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: Function to detect duplicates
| Am Freitag 26 Februar 2010 00:57:48 schrieb Rafael Gustavo da Cunha Pereira | Pinto: | | There is a single 10 digit number that: | | 1) uses all ten digits [0..9], with no repetitions | 2) the number formed by the first digit (right to left, most | significant) is divisible by one | 3) the number formed by the first 2 digits (again right to left) is | divisible by two | 4) the number formed by the first 3 digits is divisible by three | and so on, until: | 11) the number formed by the first 10 digits (all!) is by 10 Since Ishaaq Chandy just posted about how to generalize nested list comprehensions, I thought this was an interesting way to approach this. First a couple of simple helper functions: val = foldl (\x y - x*10+y) 0 divides d n = n `mod` d == 0 So you could solve it using a set of list comprehensions: solutions = [[x1,x2,x3,x4,x5,x6,x7,x8,x9,x10] | x1 - [0..9] , x2 - [0..9], divides 2 $ val [x1,x2] , x3 - [0..9], divides 3 $ val [x1,x2,x3] , x4 - [0..9], divides 4 $ val [x1,x2,x3,x4] , x5 - [0..9], divides 5 $ val [x1,x2,x3,x4,x5] , x6 - [0..9], divides 6 $ val [x1,x2,x3,x4,x5,x6] , x7 - [0..9], divides 7 $ val [x1,x2,x3,x4,x5,x6,x7] , x8 - [0..9], divides 8 $ val [x1,x2,x3,x4,x5,x6,x7,x8] , x9 - [0..9], divides 9 $ val [x1,x2,x3,x4,x5,x6,x7,x8,x9] , x10 - [0] , length (nub [x1,x2,x3,x4,x5,x6,x7,x8,x9,x10]) == 10 ] This is a nicely declarative way to do it, and a pretty clear way to formulate the original problem statement. But it's a bit tedious with all the repetitions, so you would rather recurse to make it more general. Since list comprehensions are just a different way to work in the list monad (where | becomes 'guard'), I managed to come up with this: solve :: [Int] - [[Int]] solve prefix = do let l = length prefix if l == 10 then return prefix else do x - [0..9] let r = prefix++[x] guard (divides (l+1) (val r) nub r == r) solve r -k (PS: I'm happy to hear any comments regarding style or other issues) -- If I haven't seen further, it is by standing in the footprints of giants ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: GUI (was: Re: [Haskell-cafe] DLL on Windows)
Hi Daniel, On Fri, 26 Feb 2010 15:31 +0100, Daniel Fischer daniel.is.fisc...@web.de wrote: Am Mittwoch 17 Februar 2010 12:17:10 schrieb Jeremy O'Donoghue: You're probably correct about the dependencies. I have never tried to compile wxHaskell against GHC 6.12.1 [snip] Beating a dead horse, but today I decided to see whether I could install wxHaskell, all it took was a few edits in some .cabal files and one in wxcore's Setup to make it all build with 6.12.1. Still, it's annoying because you have to unpack everything, edit and cabal install in the unpacked directories. Now since I did that, I could upload new verxions of wxdirect, wxcore and wx to hackage, so others could do a simple cabal install wx. Thanks for doing this: it's great news for everyone who is on GHC 6.12.x. But I would only do that if I get a go-ahead from somebody authorised (since Jeremy uploaded the previous versions, he'd be qualified, I think). You're right - I'm probably as qualified as anyone! I've found that if a Cabal build works for one person, it's pretty reliable for all (this is one of the joys of Cabal and (to a lesser extent) the Haskell Platform. I'm quite happy for you to upload updates to Cabal. I'll pull them next week and make sure that they get merged to the wxHaskell darcs repo. If there are any reports of issues with the uploaded version (which I doubt) I'll look into them, so you're not taking on any long-term support burden. Or I could send the edits to Jeremy and let him do the uploading, whatever is preferred. I have no strong preference - the only thing I want to be sure of is that the darcs repo stays closely in sync with whatever is in Cabal. If you are OK with doing the upload, that's great. Otherwise I'll be very happy to do it based on your patches. Best regards Jeremy -- Jeremy O'Donoghue jeremy.odonog...@gmail.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] do we need types?
This reminds me of an email posted to this list long ago
by Luke Palmer, describing a use of records-as-interfaces
in Agda.
--
Jason Dusek
-- Forwarded message --
From: Luke Palmer lrpal...@gmail.com
Date: 2009/12/29
Subject: Re: [Haskell-cafe] Alternatives to type classes.
To: Jason Dusek jason.du...@gmail.com
Cc: haskell haskell-cafe@haskell.org
On Tue, Dec 29, 2009 at 6:22 PM, Jason Dusek jason.du...@gmail.com wrote:
Consider the real numbers. They are a group. We have an
identity element `0', inverses and closure under the associative
operation `+'.
Group+ = (+, 0, -1 * _)
They are another group, too -- the group with `*':
Group* = (*, 1, 1 / _)
Ignoring 0 for sake of discussion.
This seems like a real problem with the whole notion of
typeclasses -- we can't really say a set/type is its
extension with some new operations.
One road to go on this is to make every extension of the set
with new ops a different type; but that seems really horribly
inconvenient. I wonder what approaches have been tried here?
I consider typeclasses a happy notational medium. They are not
perfect, they miss some cases, but they are pretty good.
For full generality at the expense of some verbosity, I like Agda's
solution pretty well. Agda allows you to open a record into a
scope.
record Group (a : Set) where
field
_+_ : a - a - a
-_ : a - a
0 : a
conj : {a : Set} - Group a - a - a - a
conj g x y = x + y + (-x)
where open g
Maybe I even got the syntax right :-P
The cool thing is that you can use this for the invariant-keeping
property of typeclasses, too. Eg. Data.Map relies on the fact that
there is at most one Ord instance per type. By parameterizing the
module over the Ord record, we can do the same:
record Ord (a : Set) where ...
module MapMod (a : Set) (ord : Ord a) where
Map : b - Set
Map = ...
insert : {b : Set} - a - b - Map b - Map b
insert = ...
...
So we have the liberty of being able to use different Ord instances,
but different Ord instances give rise to different Map types, so we
can not violate any invariants.
You can do something similar in Haskell using an existential type,
although it is very inconvenient:
data Ord a = ...
data MapMod map a b = MapMod { empty :: map a b, insert :: a - b -
map a b - map a b, ... }
withMap :: Ord a - (forall map. MapMod map a b - z) - z
withMap ord f = f ( {- implement MapMod here, using ord for ordering }- )
Then you could use maps on different Ords for the same type, but they
could not talk to each other.
Some syntax sugar could help the Haskell situation quite a lot.
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Testing and module export lists
On Feb 24, 2010, at 20:51 , Ivan Miljenovic wrote: On 25 February 2010 11:24, Don Stewart d...@galois.com wrote: Seriously?? Doesn't that break the module system? Maybe I misunderstood it; all I know is that Curt Sampson says he uses this kind of stuff for testing purposes by not having to export functions. See the -fwarn-unused-binds section at http://www.haskell.org/ghc/docs/latest/html/users_guide/options-sanity.html As I read that, the leading underscore doesn't export anything; it just suppresses any unused warning for the symbol (which is consistent with the other uses of leading underscore in warnings; it's also consistent with binding to _ to throw a result away). -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu electrical and computer engineering, carnegie mellon universityKF8NH PGP.sig Description: This is a digitally signed message part ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] ErrorT instance of MonadPlus
Hello, ErrorT instance of MonadPlus define that in case of fail both arguments of mplus a value of Error will be the Error of second computation: m `mplus` n = ErrorT $ doa - runErrorT mcase a of Left _ - runErrorT nRight r - return (Right r) Could it be changed in this way: m `mplus` n = ErrorT $ doa - runErrorT mcase a of Left e - do b - runErrorT n case b of Left e' - return $ Left (e `eplus` e') r - return r r - return r where `eplus` could be placed in class Error a: eplus :: a - a - a eplus x y = y -- default implementation In this case we could combine Errors more flexible. Best regards, Dmitry ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: Proper round-trip HughesPJ/Parsec for Doubles?
real :: Parser String real = do d - decimal f - option $ do p - char '.' n - many1 digit return $ p : n Just to throw two bits in here, this is the only style that doesn't require leaning on the space bar and squinting to line things up, doesn't require any fancy editor support, and looks fine with proportional fonts. It also allows you to move lines around with 'dd' and 'p' in vi :) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] vector to uvector and back again
Great, thanks! On Thu, Feb 25, 2010 at 10:29 PM, Dan Doel dan.d...@gmail.com wrote: On Friday 26 February 2010 12:13:56 am Bryan O'Sullivan wrote: Dan, do you think you might be releasing your port of uvector-algorithms to vector any time soon? I've ported mwc-random to use vector, and I'd like to move statistics (which needs uvector-algorithms) and criterion (ditto) too. I don't want to hold anything up, so I've released the port to vector. It's available on hackage as vector-algorithms 0.3: http://hackage.haskell.org/package/vector-algorithms It's mostly a straight port, so not much new to learn. - There's no .Array in the module names anymore. - The Schwartzian transform combinators are gone from Data.Vector.Algorithms.Combinators, because I haven't decided on the best way to handle those yet (the existing implementation won't work on all vectors in the MVector class). Hope that isn't a problem. - There's also a new module D.V.A.Search, which so far implements a couple variations on binary search. It was something I was starting before the switch to vector, so it isn't complete yet. There are some moderate performance regressions on some of the algorithms, but nothing major. Also, the optimizer in 6.12 seems to get very confused when working with IO as the PrimMonad in question, resulting in significantly worse performance. So, I'd recommend sticking with ST, or at least making sure the algorithms are called in ST, with stToIO. HEAD is better on both these fronts, so things should get better in the future. Let me know if there are any issues.* -- Dan * P.S. I just noticed I left the .cabal recommending -O2 and -fvia-c -optc-O3. That's obviously not current since the .cabal is set to compile with -Odph and with the NCG. I'll amend that in a later version. :) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] do we need types?
On Fri, Feb 26, 2010 at 04:23:52PM +0300, Miguel Mitrofanov wrote:
I'd say we don't really need subclasses. I mean, what's the difference:
class Eq a where (==) :: a - a - Bool
instance Eq a = Eq (Maybe a) where
Nothing == Nothing = True
Just x == Just y = x == y
_ == _ = False
sort :: Eq a = [a] - [a]
or
data Eq a = Eq {eq :: a - a - Bool}
eqMaybe :: Eq a - Eq (Maybe a)
eqMaybe e = Eq {eq = eqM} where
eqM Nothing Nothing = True
eqM (Just x) (Just y) = eq e x y
eqM _ _ = False
sort :: Eq a - [a] - [a]
Replacing classes with types, we only lose one thing: the compiler won't
deduce the right instances for us. I'll trade it for the ability to abstract
over them. After all, we CAN deduce the right
instances by hand, it's just a finite amount of work (not very big, in my
experience).
But then we would lose the invarient that there is a unique pairing
between a type and a given class. for instance, you would no longer be
able to implement things like Set and Map,
For instance if you called the two following functions with different
ord arguments, you would suddenly break all the invarients of what 'Set'
means.
insert :: Ord a - a - Set a - Set a
member :: Ord a - a - Set a - Bool
The unique correspondence between types and classes (i.e. no local
instances) is a main _feature_ of type classes. Often when people think
they need local instances, they are just applying type classes when they
should be using a different idiom, such as the one you mention.
John
--
John Meacham - ⑆repetae.net⑆john⑈ - http://notanumber.net/
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] sendfile leaking descriptors on Linux?
Thanks, I altered my top level request handler as follows mysmartserver conf h stateProxy = do socket - bindPort conf I added (setSocketOption socket ReuseAddr 1) here Should this be added in a comment, or even in function code, in Happstack.Server.SimpleHTTP? What are the tradeoffs, when would you *not* want ot use ReuseAddr?) webserverTid - forkIO $ simpleHTTPWithSocket socket conf h putStrLn . ( starting happs server ++ ) = time control - startSystemState stateProxy -- start the HAppS state system putStrLn . ( happs state started ++ ) = time waitForTermination killThread webserverTid stateShutdown control I can't replicate the error reliably so I won't know if it actually fixed the problem for a while, therefore just leaving this cookie crumb trail in case others may find it helpful. 2010/2/26 Brandon S. Allbery KF8NH allb...@ece.cmu.edu: On Feb 26, 2010, at 04:28 , Thomas Hartman wrote: me: Like mightybyte, I run my app in a shell loop that will just restart it after a crash. But every once in a while it won't restart because of the busy socket and I need to do a manual restart, killing multiple processes (usually 2). the error I get is: bind: resource busy (Address already in use) This is on application restart? It's not out of file descriptors, it's just the system keeping the socket around (netstat will show it in TIME_WAIT, or possibly in a shutdown negotiation state such as LAST_ACK, FIN_WAIT, etc. TCP lacks *reliable* socket shutdown negotiation). You want to configure the socket with SO_REUSEADDR before trying to bind it (setSocketOption socket ReuseAddr 1). As to the portable version of sendfile, it's because not all systems offer a sendfile() system call. Linux and *BSD do, and can use the native implementation; the portable version emulates sendfile() when it doesn't exist, at the price of additional CPU usage/system load (sendfile() having been created specifically to reduce system load in the common case for web servers). -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu electrical and computer engineering, carnegie mellon university KF8NH ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Redirecting output
Is there any way I can temporarily reassign stdout? In particular, will this be handled well by the FFI? I'm hoping there's maybe something like PHP's output buffering thingy http://php.net/manual/en/book.outcontrol.php. I'm *binding* (not doing a process call) to an external library which outputs directly to stdout, but whose output I'd like to use... Louis Wasserman wasserman.lo...@gmail.com http://profiles.google.com/wasserman.louis ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] sendfile leaking descriptors on Linux?
On Feb 26, 2010, at 14:43 , Thomas Hartman wrote: in Happstack.Server.SimpleHTTP? What are the tradeoffs, when would you *not* want ot use ReuseAddr?) In the modern world, any server socket probably should use SO_REUSEADDR. About the only modern use for the default would involve servers launched as a result of a port-knock or similar. -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu electrical and computer engineering, carnegie mellon universityKF8NH PGP.sig Description: This is a digitally signed message part ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] ANNOUNCE: wxHaskell for ghc-6.12.1
Hi all, there's a new suite of wxHaskell on hackage, http://hackage.haskell.org/package/wxdirect-0.12.1.2 http://hackage.haskell.org/package/wxcore-0.12.1.3 http://hackage.haskell.org/package/wx-0.12.1.3 If you have ghc-6.12.1 and wxWidgets-2.8, a simple cabal install wx should give you wx GUI goodness. It built fine on my openSUSE, and I don't see any dependencies which look like they weren't unproblematically cross- platform. Unfortunately I was a little too restrictive with the constraints on containers, so it doesn't work with ghc-6.10, sorry. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] EDSL's using Filet-O-Fish of Barrelfish project
Hi Kashyap, What would be your recommendation on how to get a grasp of FoF Well, you can read the PLOS paper first. Also, FoF is a literate code, so hopefully the learning curve will be less steeper. You should also take a look at the Fugu compiler literate code and, in less extend, the Hamlet compiler. They are sitting in ./tools/. would the tic-tac-toe sample help - could you share that please? Tic-tac-toe in the caps system was really nothing but a hack (for a demo). There is no pedagogical value in staring at it: it is just the file ./capabilities/caps.hl extended with approx. 4800 capabilities! Reading the official caps.hl is much more interesting, and, if you really want to, you will quickly see how you can implement your favorite board game in there. But don't! :-) Regards, -- Pierre-Evariste DAGAND ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Redirecting output
On Feb 26, 2010, at 14:44 , Louis Wasserman wrote:
Is there any way I can temporarily reassign stdout? In particular,
will this be handled well by the FFI? I'm hoping there's maybe
something like PHP's output buffering thingy. I'm *binding* (not
doing a process call) to an external library which outputs directly
to stdout, but whose output I'd like to use...
That way lies pain. It can be done, but if it breaks it will break
rather thoroughly and leave you stranded. Untested (aside from
compilation), no warranty, etc.:
module Redirect (withRedirectedStdout) where
import System.IO
import System.Posix.IO
-- usage: withRedirectedStdout destinationFileName $ yourIOActionHere
-- this might work with Haskell I/O, if you're lucky; don't bet on it
-- stick to FFI calls that use raw I/O
withRedirectedStdout :: FilePath - IO a - IO a
withRedirectedStdout f io= do
hFlush stdout
-- style note: should be stdOutput but that doesn't work for the
binds
-- and I'm not sure it's worth doing those right. maybe if
someone wants this
-- for the library I'll clean it up
old - dup 1
closeFd 1
-- this will pattern-match fail if we don't get stdout back; if
you manage to trigger it,
-- you get to keep both pieces. (you probably closed stdin,
which is usually a
-- recipe for disaster, or at minimum extreme confusion)
1 - openFd f WriteOnly (Just 0600) (defaultFileFlags {noctty =
True, trunc = True, nonBlock = False})
res - io
closeFd 1
1 - dup old
closeFd old
return res
Actually, I bet this fails because your FFI function uses C/C++ stdio
and doesn't flush it. That will probably require bringing in C
stdio's fflush(stdout), which will be even more painful.
--
brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com
system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu
electrical and computer engineering, carnegie mellon universityKF8NH
PGP.sig
Description: This is a digitally signed message part
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: Function to detect duplicates
Am Freitag 26 Februar 2010 16:50:42 schrieb Ketil Malde: | Am Freitag 26 Februar 2010 00:57:48 schrieb Rafael Gustavo da Cunha | Pereira | | Pinto: | There is a single 10 digit number that: | | 1) uses all ten digits [0..9], with no repetitions | 2) the number formed by the first digit (right to left, most | significant) is divisible by one | 3) the number formed by the first 2 digits (again right to left) is | divisible by two | 4) the number formed by the first 3 digits is divisible by three | and so on, until: | 11) the number formed by the first 10 digits (all!) is by 10 Since Ishaaq Chandy just posted about how to generalize nested list comprehensions, I thought this was an interesting way to approach this. Yes. But it approaches the border, for 20 digits it would become annoying to type. First a couple of simple helper functions: val = foldl (\x y - x*10+y) 0 divides d n = n `mod` d == 0 So you could solve it using a set of list comprehensions: solutions = [[x1,x2,x3,x4,x5,x6,x7,x8,x9,x10] | x1 - [0..9] First digit can't be 0, so make it [1 .. 9]. Since you use the fact that the last digit must be the 0, pull all others from [1 .. 9]. , x2 - [0..9], divides 2 $ val [x1,x2] , x1 /= x2 , x3 - [0..9], divides 3 $ val [x1,x2,x3] , x3 `notElem` [x1,x2] -- etc. , x4 - [0..9], divides 4 $ val [x1,x2,x3,x4] , x5 - [0..9], divides 5 $ val [x1,x2,x3,x4,x5] , x6 - [0..9], divides 6 $ val [x1,x2,x3,x4,x5,x6] , x7 - [0..9], divides 7 $ val [x1,x2,x3,x4,x5,x6,x7] , x8 - [0..9], divides 8 $ val [x1,x2,x3,x4,x5,x6,x7,x8] , x9 - [0..9], divides 9 $ val [x1,x2,x3,x4,x5,x6,x7,x8,x9] , x10 - [0] , length (nub [x1,x2,x3,x4,x5,x6,x7,x8,x9,x10]) == 10 ] Doesn't look as nice, but early pruning saves a lot of work (in this case, for very small values of a lot). This is a nicely declarative way to do it, and a pretty clear way to formulate the original problem statement. A very direct translation :) But it's a bit tedious with all the repetitions, so you would rather recurse to make it more general. Since list comprehensions are just a different way to work in the list monad (where | becomes 'guard'), I managed to come up with this: solve :: [Int] - [[Int]] Not on a 32-bit system. Word would suffice there, but you don't know that in advance, so it'd be Int64 or Integer solve prefix = do let l = length prefix if l == 10 then return prefix else do x - [0..9] You can guard (x `notElem` prefix) here, or use x `notElem` prefix below, but don't use nub r == r when you know that only the new element may be duplicated. let r = prefix++[x] guard (divides (l+1) (val r) nub r == r) solve r -k (PS: I'm happy to hear any comments regarding style or other issues) I would make the length of the prefix a parameter of solve. It's admittedly less elegant, but all those calls to length hurt me :) Regarding style, I think I prefer solve prefix = case length prefix of 10 - return prefix l - do x - [0 .. 9] ... over the if-then-else. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: Function to detect duplicates
Daniel Fischer daniel.is.fisc...@web.de skrev: Am Freitag 26 Februar 2010 16:50:42 schrieb Ketil Malde: solutions = [[x1,x2,x3,x4,x5,x6,x7,x8,x9,x10] | x1 - [0..9] First digit can't be 0, so make it [1 .. 9]. Since you use the fact that the last digit must be the 0, pull all others from [1 .. 9]. Originally, I pulled from alternating odds (x1 - [1,3..9] etc) and evens, since this is fairly easy to deduce... I reverted this since the point was to use brute force. solve :: [Int] - [[Int]] Not on a 32-bit system. Word would suffice there, but you don't know that in advance, so it'd be Int64 or Integer Hm? The Ints are just individual digits here. I would make the length of the prefix a parameter of solve. I thought about generating a list with solutions for increasing lenghts, so that e.g. 'solve [] !! 10' would solve this particular problem. solve prefix = case length prefix of 10 - return prefix l - do x - [0 .. 9] ... over the if-then-else. Yes, much nicer. Thanks for the feedback! -k -- If I haven't seen further, it is by standing in the footprints of giants ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: Function to detect duplicates
Am Freitag 26 Februar 2010 21:34:28 schrieb Ketil Malde: Daniel Fischer daniel.is.fisc...@web.de skrev: Am Freitag 26 Februar 2010 16:50:42 schrieb Ketil Malde: solutions = [[x1,x2,x3,x4,x5,x6,x7,x8,x9,x10] | x1 - [0..9] First digit can't be 0, so make it [1 .. 9]. Since you use the fact that the last digit must be the 0, pull all others from [1 .. 9]. Originally, I pulled from alternating odds (x1 - [1,3..9] etc) and evens, since this is fairly easy to deduce... I reverted this since the point was to use brute force. Yes, but did you forget x10 or did you think that one was too obvious? solve :: [Int] - [[Int]] Not on a 32-bit system. Word would suffice there, but you don't know that in advance, so it'd be Int64 or Integer Hm? The Ints are just individual digits here. Yup. I didn't realise that you don't call val for the 10-digit number(s). If you also did x10 - [0 .. 9] and checked val [x1, x2, ..., x10] `mod` 10 == 0, it would overflow, that's what I was thinking of. I would make the length of the prefix a parameter of solve. I thought about generating a list with solutions for increasing lenghts, so that e.g. 'solve [] !! 10' would solve this particular problem. That's nice, but I think it'd be ugly with a DFS, much nicer with a BFS, like Rafael did. solve prefix = case length prefix of 10 - return prefix l - do x - [0 .. 9] ... over the if-then-else. Yes, much nicer. Thanks for the feedback! -k ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Redirecting output
Okay, okay, I'm convinced that trying to mess with it in C is easier than in
Haskell. =P
Louis Wasserman
wasserman.lo...@gmail.com
http://profiles.google.com/wasserman.louis
On Fri, Feb 26, 2010 at 2:14 PM, Brandon S. Allbery KF8NH
allb...@ece.cmu.edu wrote:
On Feb 26, 2010, at 14:44 , Louis Wasserman wrote:
Is there any way I can temporarily reassign stdout? In particular, will
this be handled well by the FFI? I'm hoping there's maybe something like
PHP's output buffering thingyhttp://php.net/manual/en/book.outcontrol.php.
I'm *binding* (not doing a process call) to an external library which
outputs directly to stdout, but whose output I'd like to use...
That way lies pain. It can be done, but if it breaks it will break rather
thoroughly and leave you stranded. Untested (aside from compilation), no
warranty, etc.:
module Redirect (withRedirectedStdout) where
import System.IO
import System.Posix.IO
-- usage: withRedirectedStdout destinationFileName $ yourIOActionHere
-- this might work with Haskell I/O, if you're lucky; don't bet on it
-- stick to FFI calls that use raw I/O
withRedirectedStdout :: FilePath - IO a - IO a
withRedirectedStdout f io= do
hFlush stdout
-- style note: should be stdOutput but that doesn't work for the binds
-- and I'm not sure it's worth doing those right. maybe if someone
wants this
-- for the library I'll clean it up
old - dup 1
closeFd 1
-- this will pattern-match fail if we don't get stdout back; if you
manage to trigger it,
-- you get to keep both pieces. (you probably closed stdin, which is
usually a
-- recipe for disaster, or at minimum extreme confusion)
1 - openFd f WriteOnly (Just 0600) (defaultFileFlags {noctty = True,
trunc = True, nonBlock = False})
res - io
closeFd 1
1 - dup old
closeFd old
return res
Actually, I bet this fails because your FFI function uses C/C++ stdio and
doesn't flush it. That will probably require bringing in C stdio's
fflush(stdout), which will be even more painful.
--
brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com
system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu
electrical and computer engineering, carnegie mellon universityKF8NH
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Re: do we need types?
This seems quite relevant: http://lambda-the-ultimate.org/node/3837 titto ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] sendfile leaking descriptors on Linux?
Hello, It will be interesting to see if that makes any difference -- it shouldn't. In happstack-server we use 'listenOn'. According to the documentation listenOn already sets ReuseAddr: http://hackage.haskell.org/packages/archive/network/2.2.1.7/doc/html/Network.html#v%3AlistenOn A quick look at the source code confirms it is calling: setSocketOption sock ReuseAddr 1 It sounds to be like you are getting the socket in use error because you have 2 processes running. Seems like the first one hasn't really died, but the second one is started anyway. How is it that your loop starts a second server when the first one has not finished? - jeremy On Fri, Feb 26, 2010 at 1:43 PM, Thomas Hartman tphya...@gmail.com wrote: Thanks, I altered my top level request handler as follows mysmartserver conf h stateProxy = do socket - bindPort conf I added (setSocketOption socket ReuseAddr 1) here Should this be added in a comment, or even in function code, in Happstack.Server.SimpleHTTP? What are the tradeoffs, when would you *not* want ot use ReuseAddr?) webserverTid - forkIO $ simpleHTTPWithSocket socket conf h putStrLn . ( starting happs server ++ ) = time control - startSystemState stateProxy -- start the HAppS state system putStrLn . ( happs state started ++ ) = time waitForTermination killThread webserverTid stateShutdown control I can't replicate the error reliably so I won't know if it actually fixed the problem for a while, therefore just leaving this cookie crumb trail in case others may find it helpful. 2010/2/26 Brandon S. Allbery KF8NH allb...@ece.cmu.edu: On Feb 26, 2010, at 04:28 , Thomas Hartman wrote: me: Like mightybyte, I run my app in a shell loop that will just restart it after a crash. But every once in a while it won't restart because of the busy socket and I need to do a manual restart, killing multiple processes (usually 2). the error I get is: bind: resource busy (Address already in use) This is on application restart? It's not out of file descriptors, it's just the system keeping the socket around (netstat will show it in TIME_WAIT, or possibly in a shutdown negotiation state such as LAST_ACK, FIN_WAIT, etc. TCP lacks *reliable* socket shutdown negotiation). You want to configure the socket with SO_REUSEADDR before trying to bind it (setSocketOption socket ReuseAddr 1). As to the portable version of sendfile, it's because not all systems offer a sendfile() system call. Linux and *BSD do, and can use the native implementation; the portable version emulates sendfile() when it doesn't exist, at the price of additional CPU usage/system load (sendfile() having been created specifically to reduce system load in the common case for web servers). -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu electrical and computer engineering, carnegie mellon universityKF8NH ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Hackage statistics
Is there a simple way to discover the answer to questions like how many packages are on Hackage? or how many downloads per day does Hackage serve? Or is this information not posted anywhere? (I know there was a blog article a while back with some stats in it, but what about the statistics today?) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Hackage statistics
andrewcoppin: Is there a simple way to discover the answer to questions like how many packages are on Hackage? or how many downloads per day does Hackage serve? Or is this information not posted anywhere? (I know there was a blog article a while back with some stats in it, but what about the statistics today?) 501 users have uploaded 7202 versions of 1870 packages. http://hackage.haskell.org/cgi-bin/hackage-scripts/stats The download stats are computed occasionally, but I've not had time since Hackage moved to the new server. The last edition is here: In November, there were 287 package released on Hackage, bringing the total number of Haskell packages to 1720. We had a record 143 thousand downloads from Hackage in November, breaking the previous monthly record of 95 thousand set in October 2009, and bringing the total downloads from Hackage to 1.53M. http://haskellwebnews.wordpress.com/2009/12/05/whats-new-in-haskell-december-2009/ We average about 100k downloads a month (e.g. via cabal-install). Or 2-3 thousand downloads a day. More updates soon. -- Don ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: do we need types?
Am Freitag 26 Februar 2010 14:07:07 schrieb Johannes Waldmann: I wouldn't easily give up algebraic data types Nor I. Without easily defined ADTs, Haskell wouldn't be nearly so awesome. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Hackage statistics
Am Freitag 26 Februar 2010 23:57:05 schrieb Don Stewart: 501 users have uploaded 7202 versions of 1870 packages. http://hackage.haskell.org/cgi-bin/hackage-scripts/stats The download stats are computed occasionally, but I've not had time since Hackage moved to the new server. Would it be possible to automate that, add that info to the stats page and link to it from the front page? We average about 100k downloads a month (e.g. via cabal-install). I'm pretty sure cabal-install plays a BIG role in that. I wouldn't have half - what do I say, probably not a quarter - as many packages installed without it. A propos, a big thank you to all who contributed to it. Or 2-3 thousand downloads a day. More updates soon. -- Don ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Hackage redirect loop
Trying to access http://hackage.haskell.org/trac/hackage I land in a redirection loop no matter which page or which browser I try. It worked earlier today. Has anybody an idea what the heck might be going on? Is it just me? ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Hackage redirect loop
On Sat, Feb 27, 2010 at 01:56:34AM +0100, Daniel Fischer wrote: Trying to access http://hackage.haskell.org/trac/hackage I land in a redirection loop no matter which page or which browser I try. It worked earlier today. Has anybody an idea what the heck might be going on? Is it just me? Works fine for me. Good luck, -- Felipe. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Hackage redirect loop
Am Samstag 27 Februar 2010 02:15:17 schrieb Felipe Lessa: On Sat, Feb 27, 2010 at 01:56:34AM +0100, Daniel Fischer wrote: Trying to access http://hackage.haskell.org/trac/hackage I land in a redirection loop no matter which page or which browser I try. It worked earlier today. Has anybody an idea what the heck might be going on? Is it just me? Works fine for me. Dang! Oddly, it's only hackage, hackage.haskell.org/trac/ghc works fine, as does .../trac/hugs, haskell-prime, gtk2hs, the package database, all work flawlessly. Only the hackage trac doesn't. Good luck, -- Felipe. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] sendfile leaking descriptors on Linux?
Indeed, the error occurs when two processes are running at the same time. One process isn't serving anything, and the other is just looping, reported that it can't stop because the port is taken, and repeating ad infinitum. The loop I have is ulimit -v 15 while true; do echo starting patchtag: `date` ./dist/build/patchtagserver/patchtagserver --port=80 -- echo patchtag exited: `date` echo patchtag exited: `date` patch-tag.log rm -f _local/patchtagserver_state.lock The ulimit is so that it will die after consuming a reasonable amount of memory. (I seem to have a memory leak somewhere, but it takes many hours to get up to 150M.) Everything else looks pretty standard to me. It seems to work as designed most of the time, but occasionally will result in two processes. I'm baffled as to how this happens. If anybody has any idea, I would love to hear them. 2010/2/26 Jeremy Shaw jer...@n-heptane.com: Hello, It will be interesting to see if that makes any difference -- it shouldn't. In happstack-server we use 'listenOn'. According to the documentation listenOn already sets ReuseAddr: http://hackage.haskell.org/packages/archive/network/2.2.1.7/doc/html/Network.html#v%3AlistenOn A quick look at the source code confirms it is calling: setSocketOption sock ReuseAddr 1 It sounds to be like you are getting the socket in use error because you have 2 processes running. Seems like the first one hasn't really died, but the second one is started anyway. How is it that your loop starts a second server when the first one has not finished? - jeremy On Fri, Feb 26, 2010 at 1:43 PM, Thomas Hartman tphya...@gmail.com wrote: Thanks, I altered my top level request handler as follows mysmartserver conf h stateProxy = do socket - bindPort conf I added (setSocketOption socket ReuseAddr 1) here Should this be added in a comment, or even in function code, in Happstack.Server.SimpleHTTP? What are the tradeoffs, when would you *not* want ot use ReuseAddr?) webserverTid - forkIO $ simpleHTTPWithSocket socket conf h putStrLn . ( starting happs server ++ ) = time control - startSystemState stateProxy -- start the HAppS state system putStrLn . ( happs state started ++ ) = time waitForTermination killThread webserverTid stateShutdown control I can't replicate the error reliably so I won't know if it actually fixed the problem for a while, therefore just leaving this cookie crumb trail in case others may find it helpful. 2010/2/26 Brandon S. Allbery KF8NH allb...@ece.cmu.edu: On Feb 26, 2010, at 04:28 , Thomas Hartman wrote: me: Like mightybyte, I run my app in a shell loop that will just restart it after a crash. But every once in a while it won't restart because of the busy socket and I need to do a manual restart, killing multiple processes (usually 2). the error I get is: bind: resource busy (Address already in use) This is on application restart? It's not out of file descriptors, it's just the system keeping the socket around (netstat will show it in TIME_WAIT, or possibly in a shutdown negotiation state such as LAST_ACK, FIN_WAIT, etc. TCP lacks *reliable* socket shutdown negotiation). You want to configure the socket with SO_REUSEADDR before trying to bind it (setSocketOption socket ReuseAddr 1). As to the portable version of sendfile, it's because not all systems offer a sendfile() system call. Linux and *BSD do, and can use the native implementation; the portable version emulates sendfile() when it doesn't exist, at the price of additional CPU usage/system load (sendfile() having been created specifically to reduce system load in the common case for web servers). -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu electrical and computer engineering, carnegie mellon university KF8NH ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
