November 2011 - Haskell-Cafe

Strange GC timings
by Artyom Kazak 12 Nov '11

12 Nov '11

Hello! The following program executes 1.5 seconds on my computer: -----------------------CODE BEGIN------------------------- module Main where import Data.Array.IArray main = print (answers ! 1000000) nextAns :: (Int, Int, Float) -> (Int, Int, Float) nextAns (a, n, r) = if r2 > 1 then (a+1, n+2, r2) else (a+1, n+3, r3) where a' = fromIntegral a n' = fromIntegral n r2 = r * (a'/(a'+1))**n' * (n'+1)*(n'+2)/(a'+1)^2 r3 = r2 * (n'+3) / (a'+1) answers :: Array Int Int answers = listArray (1, 1000000) (map snd3 $ iterate nextAns (1, 2, 2)) where snd3 (a, b, c) = b ------------------------CODE END-------------------------- From these 1.5 seconds, 1 second is spent on doing GC. If I run it with "-A200M", it executes for only 0.5 seconds (total). Which is more interesting, when I use UArray instead of Array, it spends only 0.02 seconds in GC, but total running time is still 1.5 seconds. Why are... these things?

2 1

Label macro expansion bug In HList.
by aditya siram 12 Nov '11

12 Nov '11

Hi all, I am exploring OOHaskell and ran into some compilation issues with some of the samples. I hope this is the right place to report it. For example "OCamlTutorial.hs" generates the following error: ../samples/OCamlTutorial.hs:98:3: Multiple declarations of `foo' Declared at: ../samples/OCamlTutorial.hs:54:1 ../samples/OCamlTutorial.hs:98:3 Failed, modules loaded: OOHaskell, Dynamic, Print, DeepNarrow, Nominal, New. This is happening because the "label" macro on line 98: $(label "varX") doesn't expand correctly and clashes with the "foo" function on line 54: foo f = f # field1 The "label" function is found in the Data.HList.MakeLabels module and should work this way according to the docs: > runQ (label "test") >>= putStrLn . pprint data Test deriving (Data.Typeable.Internal.Typeable) test = Data.HList.FakePrelude.proxy :: Data.HList.FakePrelude.Proxy Test But I get: data Foo_0 deriving (Data.Typeable.Internal.Typeable) foo_1 = Data.HList.FakePrelude.proxy :: Data.HList.FakePrelude.Proxy Foo_0 This still doesn't cleanly explain why "OCamlTutorial.hs" is failing since it seems to be generating unique datatypes and functions by suffixing them with a number, but it seems to point to the problem. I'm hoping someone who knows HList better might understand it better. Thanks, -deech

2 8

Data.Vector.Unboxed
by kaffeepause73 11 Nov '11

11 Nov '11

Hello, quick question about unboxed Vectors : Is it possible to create an unboxed vector of unboxed vector ? : > import qualified Data.Vector.Unboxed as V > type UnboxedNestedVextor = V.Vector (V.Vector Int) Alternatively I would have to use: > import qualified Data.Vector.Unboxed as V > import qualified Data.Vector as VB type UnboxedNestedVextor = VB.Vector (V.Vector Int) Is there a rule of thumb how much quicker Unboxed Vectors are ? Cheers Phil -- View this message in context: http://haskell.1045720.n5.nabble.com/Data-Vector-Unboxed-tp4977289p4977289.… Sent from the Haskell - Haskell-Cafe mailing list archive at Nabble.com.

7 8

Usage question with TemplateHaskell and Generic
by Magicloud Magiclouds 11 Nov '11

11 Nov '11

Hi, I'd like to simpler the work of deriving MyClass. And I have two ways to do: TemplateHaskell "$(derivingMyClass)", or Generic "deriving (MyClass)". Since I need to get the type name in the deriving, then I met this question: If I have "data A b = C b", then with TemplateHaskell, the type would be "VarT b", which means at compile time, I cannot get the exact type, so the type name would be "b". So I wonder if this could be resolved by TemplateHaskell, or Generic is the only choice. PS: I have not tried to do this in Generic. -- 竹密岂妨流水过山高哪阻野云飞

2 2

Re: [Haskell-cafe] Ridiculously slow FFI, or cairo binding?
by Eugene Kirpichov 11 Nov '11

11 Nov '11

Hi, The actual thanks for tracking this down go to Vincent Hanquez for finding that we're doing a lot of gmp calls (and for making me aware of ltrace), and to Felipe Lessa for finding that this is caused by poor code generated for cFloatConv :) Your changes look identical to those that I made in my copy. On Thu, Nov 3, 2011 at 11:41 AM, Axel Simon <Axel.Simon(a)in.tum.de> wrote: > Hi Eugene, > > > On 02.11.2011, at 17:34, Eugene Kirpichov wrote: > > Heh. >> >> Guess what! >> A simple {-# INLINE cFloatConv #-} helped to the same extent! >> >> Axel, I think this change should be pretty easy to incorporate, and it >> probably makes sense to inline all other functions in Types.chs too. >> >> > I've added INLINE pragmas to all these odd c2hs marshalling functions. > Could you pull and check if I've done it correctly? > > Thanks for tracking this down! > > Axel > > > Would you like me to send the trivial darcs patch or the gtk2hs team >> will take care of this? >> >> On Wed, Nov 2, 2011 at 7:29 PM, Felipe Almeida Lessa >> <felipe.lessa(a)gmail.com> wrote: >> >>> On Wed, Nov 2, 2011 at 11:24 AM, Jean-Marie Gaillourdet >>> <jmg(a)gaillourdet.net> wrote: >>> >>>> Hi Eugene, >>>> >>>> did you try using the SPECIALIZE pragma? It is part of the Haskell 98 >>>> and Haskell 2010 specifications. >>>> >>> >>> I don't think it's going to make any difference, as the core already >>> have an specialized poor version. See my first e-mail. >>> >>> -- >>> Felipe. >>> >>> >> >> >> -- >> Eugene Kirpichov >> Principal Engineer, Mirantis Inc. http://www.mirantis.com/ >> Editor, http://fprog.ru/ >> > > -- Eugene Kirpichov Principal Engineer, Mirantis Inc. http://www.mirantis.com/ Editor, http://fprog.ru/

1 1

How can I get rid of this link error?
by Amy de Buitléir 11 Nov '11

11 Nov '11

I wrote a library called AmysGeometry. The only modules in it (so far) are: Amy/Geometry/ThreeD.hs Amy/Geometry/UnitSphere.hs The library compiles just fine, and I've been using it with other programs for a while. Then yesterday, I started getting the following error. Linking dist/build/add-centroids/add-centroids ... /usr/local/lib/AmysGeometry-1.0/ghc-7.0.3/libHSAmysGeometry-1.0.a(UnitSphere.o): In function `s3c6_info': (.text+0x4557): undefined reference to `__stginit_AmysGeometryzm1zi0_AmyziGeometry_' /usr/local/lib/AmysGeometry-1.0/ghc-7.0.3/libHSAmysGeometry-1.0.a(UnitSphere.o): (.data+0x3f0): undefined reference to `AmysGeometryzm1zi0_AmyziGeometry_DZCLocation_static_info' collect2: ld returned 1 exit status cabal: Error: some packages failed to install: I think this started because I recompiled some stuff with profiling. So I wiped out my current GHC and Haskell installation, re-installed it, and rebuilt AmysGeometry. But I'm still getting this error when I try to link with it. I would be grateful for any advice on how to diagnose this and fix it. Oh, and "ghc-pkg check" doesn't report any errors.

2 2

Consuming anyToken with parsing with derivatives (derp) library in Hackage?
by Ahn, Ki Yung 11 Nov '11

11 Nov '11

I was playing with derp recently, and many of the Char and Combinator of Parsec seem to be easily definable from derp. However, I haven't yet figured out a natural way to define anyToken or satisfy combiator of Parsec using derp. There is a way of course, to list all the unicode characters and combine them with <|>, which is extremely inefficient, even if you limit yourself that satisfy combinator be compile time template haskell function, it still seems a bit silly. For things like space, it seems reasonable to list all take the sum of them, since there are handful of them. But things like consume anything that is not a space ought to be defined as a complement or negation of the space parser. In general CFG is not closed under negation, but most of the time what we want to take a negation is really for regular fragment of the language, like consuming anything that is not a linebreak to lex/parse one line comments. In Matt's Standford talk, he mentioned some vision about going beyond CFG and he mentioned negation, but in the library implementation I didn't find that. Is it already possible with some trick? Is there a standard way to define such combinators with derp? Or, is it the case that there a way to do it in the scheme library but not implemented in Haskell derp package? I read through the scheme implementation too but there didn't seem to be a combinator for tacking negations either. Ki Yung Ahn

1 0

graphviz: dotizeGraph and graphToGraph not adding positions
by Max Rabkin 10 Nov '11

10 Nov '11

My understanding of the documentation for Data.GraphViz.dotizeGraph and graphToGraph is that they should add position attributes to a graph. But they always seem to return graphs with empty attribute lists. What am I doing wrong in the following tiny example? > dotizeGraph nonClusteredParams (insNode (0, "Blah") $ empty :: Gr String ()) 0:([],"Blah")->[] I expected the first list to contain at least a Pos attribute. I have written a wrapper around /usr/share/dot which shows the program is outputting positional information. --Max

2 1

Easiest to use NoSQL storage with Haskell?
by dokondr 10 Nov '11

10 Nov '11

Hi, What Haskell package to work with NoSQL storage is both mature and easiest to use? I need persistent storage for simple key/value lists (not complex JSON docs). CouchDB and Cassandra seems to be overkill for my needs. What about Riak, MongoDB, Voldemort, etc. ? Thanks! Dmitri.

4 7

timezone-series, timezone-olson dependencies
by Ben Gamari 10 Nov '11

10 Nov '11

Is there a reason why the current version of the timezone-series and timezone-olson packages depend on time<1.3? With time 1.4 being widely used at this point this will cause conflicts with many packages yet my tests show that both packages work fine with time 1.4. Could we have this upper bound bumped to 1.5? Cheers, - Ben

2 1