March 2013 - Haskell-Cafe

my knucleotide fast on 64 bit but extremely slow on 32 bit?
by Branimir Maksimovic 29 Mar '13

29 Mar '13

I have posted previous knucleotide program, it is fast on 64 bit butvery slow on 32 bit.I cannot install 32 bit ghc to test it so I can only guess is thatcause is use of Int64 for hash and HashMap array indexing.What bothers me is that it that much slower , and I guessthat array indexing of 64 bit int where native index is 32 bitis culprit.Am I right? If I make Int index of array it will be fast on32 bit platform too?I cannot imagine that hashing to 64 bit int is culprit ratherarray indexing as I guess that every array accessrequires indirect access on 32 bit platform? 64 bit :http://benchmarksgame.alioth.debian.org/u64q/program.php?test=knucleotide&lang=ghc&id=119.80 secs32 bit:http://benchmarksgame.alioth.debian.org/u32q/program.php?test=knucleoti… secs ;( I really like Haskell, but Im frustrated as how inpredictable it is.Im programming from 1983' but found Haskell as much more difficultto program efficiently than C++.In C++ whatever lousy program I write it performs decently,but not so in Haskell. It requires much deeper knowledge thanc++.

2 2

introducing Maybe at managing level
by luc taesch 29 Mar '13

29 Mar '13

I was looking for some link introducing the way FP/ Haskell handles errors and Exceptions. This is for a non FP Guy, and ideally withought scaring them with Monads and category theory :-). for the background : the guy said : As I mentioned in another thread in banking (in particular) it is the exception processing that often dominates the functionality of a system - as the core concept is generally very straightforward. Developing for "exception handling" (not in a Java/C++ sense) is a tricky thing - as the exception don't necessarily shout themselves out - and are often why we have large misunderstood legacy systems which are hard to replace.

4 3

Re: [Haskell-cafe] warp and http-conduit on concurrent threads on windows
by Lars Kuhtz 29 Mar '13

29 Mar '13

That resolved the issue. Thanks for the hint. Lars On 2013-03-28 18:17, Joey Adams wrote: > Try upgrading to the latest network package. There's a bug prior to > version 2.4.1.0 where 'connect' blocks other threads from running, > because the underlying FFI call was marked "unsafe". > > Somewhat unrelated, but if your program needs to run for long periods > of time (e.g. a server), you're in for a nasty surprise: GHC > currently > doesn't support interrupting network I/O on Windows. After about a > thousand hung connections, the program will run out of address space > and hang. A partial workaround is to use socket timeouts; see > setHandleTimeouts in the network-socket-options package.

1 0

Remove redundancy with Template Haskell
by Corentin Dupont 29 Mar '13

29 Mar '13

Hi Café, I have a function that looks like this: call :: SomeFunction -> String -> SomeState The string is actually the representation of the function passed in parameter. It is stored in the state for documentation. So a call looks like that: call (\a -> putStrLn a) "\a -> putStrLn a" There is a clear redundancy here, how could I remove it with Template Haskell? I cannot figure out... Thanks! Corentin

3 3

warp and http-conduit on concurrent threads on windows
by Lars Kuhtz 28 Mar '13

28 Mar '13

Hi, I'd like to know what is wrong with the following program on windows8 (GHC 7.4.2, 32bit): {-# LANGUAGE OverloadedStrings #-} {-# LANGUAGE ScopedTypeVariables #-} module Main where import Control.Concurrent.Async import qualified Control.Exception as E import Network.HTTP.Conduit import Network.HTTP.Types import Network.Wai import Network.Wai.Handler.Warp query port = E.catch (simpleHttp ("http://haskell.org:" ++ show port) >>= print . take 10 . show) (\(e :: HttpException) -> print $ "caught: " ++ show e) listen = run 8080 $ \_ -> return $ responseLBS ok200 [] "abc" main = do withAsync (query 12345) $ \a -> do withAsync listen $ \b -> do wait a wait b I compile the program with "ghc --make -threaded Main.hs" and run it as "./Main +RTS -N". On POSIX systems this works as expected. Even if the failing "query" runs in a forever loop the "listen" thread responds promptly to requests. On windows the "listen" thread seems blocked by the failing "query" thread. Sometimes the query returns (relatively) prompt. But sometimes (about a third of all runs) it takes very long (about 20 sec). Also, sometimes it returns with "Connection timed out (WSAETIMEDOUT)", sometimes with "getAddrInfo: does not exist (error 11003)", and sometimes just with "FailedConnectionException". The fact that the "listen" thread is blocked seems to contradict the following quote form the documentation of Control.Concurrent: -- Quote from Control.Concurrent -- Using forkOS instead of forkIO makes no difference at all to the scheduling behaviour of the Haskell runtime system. It is a common misconception that you need to use forkOS instead of forkIO to avoid blocking all the Haskell threads when making a foreign call; this isn't the case. To allow foreign calls to be made without blocking all the Haskell threads (with GHC), it is only necessary to use the -threaded option when linking your program, and to make sure the foreign import is not marked unsafe. -- End Quote -- By the way: using withAsyncBound instead of withAsync seems to improve (but not completely solve) the issue. Thanks, Lars

2 2

nice pattern from simple converging series
by Christopher Howard 28 Mar '13

28 Mar '13

I made a nice little pattern from just a few lines of Haskell (mostly code comments) using gloss. It is very kindergarten in terms of mathematical art, but the idea was to illustrate that in a pretty short amount of time, and a small amount code, you could easily translate simple math concepts into geometry on your screen. http://frigidcode.com/code/haskell-series-circles/ -- frigidcode.com

1 0

Parallel Data.Vector.generate?
by Myles C. Maxfield 28 Mar '13

28 Mar '13

Hello all. I'm using the Data.Vector.generate function with a complicated creation function to create a long vector. Is it possible to parallelize the creation of each element? Alternatively, if there was something like parMap for vectors, I suppose I could pass id to Data.Vector.generate and use parMap. Does something like this exist? Thanks, Myles C. Maxfield

2 2

duplicate type signatures
by Heinrich Hördegen 28 Mar '13

28 Mar '13

Dear all, why is this allowed (GHCi, version 7.4.1): main :: IO () main = do let a, a :: Int a = 5 print a but this not: main :: IO () main = do let a :: Int a :: Int a = 5 print a Is there a deeper sens or is it just a little bit inconsistent? Heinrich

2 1

Re: [Haskell-cafe] [extension]syntactic sugar for maps
by Eli Frey 28 Mar '13

28 Mar '13

> Sorry, I forgot to explain (probably because I'm too used to it). I am referring to a syntax for easy creation of maps. Something equivalent to lists: > > to build a list: [ 1, 2, 3] > to build a map; { 1, "one", 2, "two", 3, "three"} > > Without it I am always forced to use fromList. This looks like something to use records for, or in any case something where association list performance is not an issue. If you just want to store some configuration-like structure and pass it around, a record is great for this. You might find where in other languages you would simply leave a key "null", in Haskell you can just fill it with a Nothing. Maps (hash or binary-tree) really pay off when they are filled dynamically with massive numbers of associations. I find when I am ending up in this scenario, I am generating my map programatically, not writing it as a literal. Sometimes people even write maps simply as functions and not even as a data-structure. > myMap char = case char of > 'a' -> 1 > 'b' -> 2 > 'c' -> 3 Perhaps you could describe a situation you are in where you are wanting this, and we could see if there is something you can do already that is satisfying and solves your problem. On Wed, Mar 27, 2013 at 12:59 PM, Eli Frey <eli.lee.frey(a)gmail.com> wrote: > > http://hackage.haskell.org/trac/ghc/wiki/OverloadedLists comes to mind. > > This assumes you can turn ANY list into a thing. Maps only make sense to > be constructed from association list. If I've got a [Char], how do I make > a map form it? > > > On Wed, Mar 27, 2013 at 12:56 PM, Nicolas Trangez <nicolas(a)incubaid.com>wrote: > >> On Wed, 2013-03-27 at 21:30 +0200, Răzvan Rotaru wrote: >> > I am terribly missing some syntactic sugar for maps (associative data >> > structures) in Haskell. I find myself using them more than any other >> > data >> > structure, and I think there is no big deal in adding some sugar for >> > this >> > to the language. I could not find out whether such an extension is >> > beeing >> > discussed. If not, I would like to propose and extension. Any help and >> > suggestions are very welcome here. Thanks. >> >> http://hackage.haskell.org/trac/ghc/wiki/OverloadedLists comes to mind. >> >> Nicolas >> >> >> _______________________________________________ >> Haskell-Cafe mailing list >> Haskell-Cafe(a)haskell.org >> http://www.haskell.org/mailman/listinfo/haskell-cafe >> > >

2 1

[extension]syntactic sugar for maps
by Răzvan Rotaru 27 Mar '13

27 Mar '13

Hi, I am terribly missing some syntactic sugar for maps (associative data structures) in Haskell. I find myself using them more than any other data structure, and I think there is no big deal in adding some sugar for this to the language. I could not find out whether such an extension is beeing discussed. If not, I would like to propose and extension. Any help and suggestions are very welcome here. Thanks. Also related to the topic: 1/ Is there a list of proposals for extensions to Haskell that has currently been accepted in the new standard? I have not found one on Haskell' ( http://hackage.haskell.org/trac/haskell-prime/query?status=new&status=assig…) but it looks outdated. No mentions to the new standard! 2/ I have seen somewhere a statement that a new language standard will be published yearly. Didn't happen until now. Is there even a new standard on the way? Cheers, Răzvan

4 4