
Luca, to use Control.Parallel, you need to download and install two packages, deepseq and parallel, from hackage.haskell.org. Most likely this will work with cabal, or you just download the two tarballs . The reason is, when packaging "parallel", this package has been removed from the GHC core libraries. BTW I am unsure whether this is at all clever, since it needs specific GHC support (at least for now - am I right here?) In addition, be informed that Control.Parallel.Strategies has been heavily restructured just last month (splitting it into deepseq and parallel is one of the changes, but not the most fundamental). If you want to try examples from GpH publications, you will certainly have some problems. parallel-1.x versions containing the original definitions are on hackage as well and should work for experiments. Cheers Jost Berthold PS: Loading Control.Parallel into ghci is a handy debugging procedure for the sequential parts of your program without any hassle, but if you do not compile and link your program, it will not use any parallelism (and run slow anyway).
From: Luca Ciciriello
Subject: RE: ANNOUNCE: GHC version 6.12.1 To: , Message-ID: Content-Type: text/plain; charset="iso-8859-1" Installed 6.12.1 on MacOS X 10.6Now I'm unable to load in GHCi of that modules containing "import Control.Parallel"I'm missing something? Luca
From: marco-oweber@gmx.de To: glasgow-haskell-users@haskell.org Date: Mon, 14 Dec 2009 15:24:48 +0100 Subject: RE: ANNOUNCE: GHC version 6.12.1
Excerpts from Luca Ciciriello's message of Mon Dec 14 15:12:45 +0100 2009:
I've the 6.10.4 version installed on my MacOS X 10.6 OS. Have I to uninstall this version of GHC before installing the Mac .pkg for the 6.12.1?
Hi Luca,
You should be able to get some hints by looking at where ghc is installed.
By default ghc puts its libraries into directory which contains the ghc release name. So libs can be installed at the same time. There are also ghc(i)-pkg-$GHC_VERSION executables. So it should be possible. However I don't know about Mac details.
But maybe someone else can give you a more accurate answer.
Sincerly Marc Weber

On Dec 14, 2009, at 14:04 , Jost Berthold wrote:
The reason is, when packaging "parallel", this package has been removed from the GHC core libraries. BTW I am unsure whether this is at all clever, since it needs specific GHC support (at least for now - am I right here?)
Only to the extent that the (standardized) primitives are only in GHC; the packages were released and tested with 6.10.x. -- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allbery@kf8nh.com system administrator [openafs,heimdal,too many hats] allbery@ece.cmu.edu electrical and computer engineering, carnegie mellon university KF8NH

Thanks to all. BTW, reading the new wiki library page I've noticed that I can use atomically, pseq, par, forkIO, etc, simply importing GHC.Conc Luca
CC: allbery@ece.cmu.edu; luca_ciciriello@hotmail.com; glasgow-haskell-users@haskell.org From: allbery@ece.cmu.edu To: berthold@Mathematik.Uni-Marburg.de Subject: Re: ANNOUNCE: GHC version 6.12.1 Date: Mon, 14 Dec 2009 14:25:21 -0500
On Dec 14, 2009, at 14:04 , Jost Berthold wrote:
The reason is, when packaging "parallel", this package has been removed from the GHC core libraries. BTW I am unsure whether this is at all clever, since it needs specific GHC support (at least for now - am I right here?)
Only to the extent that the (standardized) primitives are only in GHC; the packages were released and tested with 6.10.x.
-- brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allbery@kf8nh.com system administrator [openafs,heimdal,too many hats] allbery@ece.cmu.edu electrical and computer engineering, carnegie mellon university KF8NH
_________________________________________________________________ View your other email accounts from your Hotmail inbox. Add them now. http://clk.atdmt.com/UKM/go/186394592/direct/01/

After I installed GHC-6.12.1, the cabal-install program could not works normally. $ sudo cabal install cabal: failed to parse output of 'ghc-pkg dump' $ please help me.

On 15/12/09 08:20, flw wrote:
After I installed GHC-6.12.1, the cabal-install program could not works normally.
$ sudo cabal install cabal: failed to parse output of 'ghc-pkg dump' $
You need an updated version of cabal-install. As I understand it, version 0.8 has not been released yet, but you can get it from the darcs repository: darcs get http://darcs.haskell.org/cabal-install cd cabal-install ./bootstrap.sh Cheers, Simon

On Tue, Dec 15, 2009 at 06:41:06AM +0000, Luca Ciciriello wrote:
BTW, reading the new wiki library page I've noticed that I can use atomically, pseq, par, forkIO, etc, simply importing GHC.Conc
Note that that is not a supported interface, and may break in future releases. Thanks Ian

On 14/12/09 19:04, Jost Berthold wrote:
Luca,
to use Control.Parallel, you need to download and install two packages, deepseq and parallel, from hackage.haskell.org. Most likely this will work with cabal, or you just download the two tarballs .
cabal install parallel (assuming you have a working cabal-install)
The reason is, when packaging "parallel", this package has been removed from the GHC core libraries. BTW I am unsure whether this is at all clever, since it needs specific GHC support (at least for now - am I right here?)
In addition, be informed that Control.Parallel.Strategies has been heavily restructured just last month (splitting it into deepseq and parallel is one of the changes, but not the most fundamental). If you want to try examples from GpH publications, you will certainly have some problems. parallel-1.x versions containing the original definitions are on hackage as well and should work for experiments.
I think it would be a good idea to have a wiki page describing the differences between the API used in those papers and the current API. Off the top of my head, the main differences are: * import Control.Parallel.Strategies, not import Strategies * seq is now pseq * rnf is now rdeepseq * don't use demanding, sparking Better still would be a selection of recipes or patterns for parallelising Haskell code. Anyone feel like tackling this? (I realise I'm the obvious person to do it, but I'm deep in mid-hack on the parallel GC and don't want to get too distracted right now). Cheers, Simon
participants (6)
-
Brandon S. Allbery KF8NH
-
flw
-
Ian Lynagh
-
Jost Berthold
-
Luca Ciciriello
-
Simon Marlow