March 2015 - Haskell-Cafe

C2HS parse error on Windows ...
by aditya siram 07 Mar '15

07 Mar '15

Hi all, c2hs runs into the following error on all of my .chs files, but only on Windows: c:\\program files\\haskell platform\\2014.2.0.0\\mingw\\bin\\../lib/gcc/x86_64-w64-mingw32/4.6.3/../../../../x86_64-w64-mingw32/include/stddef.h:20: (column 11) [ERROR] >>> Syntax error ! The symbol `__attribute__' does not fit here. My c2hs version is: C->Haskell Compiler, version 0.23.1 Snowbounder, 31 Oct 2014 build platform is "x86_64-mingw32" <1, True, True, 1> This is the command being run: "c:\Users\deech\AppData\Roaming\cabal\bin\c2hs.exe" "--cpp=c:\Program Files\Haskell Platform\2014.2.0.0\mingw\bin\gcc.exe" "--cppopts=-E" "--cppopts=-D__GLASGOW_HASKELL__=708" "--cppopts=-Dmingw32_BUILD_OS=1" "--cppopts=-Dx86_64_BUILD_ARCH=1" "--cppopts=-Dmingw32_HOST_OS=1" "--cppopts=-Dx86_64_HOST_ARCH=1" "--cppopts=-I./c-src" "--cppopts=-I./" "--cppopts=-IC:/MinGW/include" "--cppopts=-IC:\Users\deech\Downloads\fltkhs-0.1.0.1/c-src" "--cppopts=-IC:\Users\deech\Downloads\fltkhs-0.1.0.1" "--include=dist\build" "--cppopts=-IC:\Program Files\Haskell Platform\2014.2.0.0\lib\process-1.2.0.0\include" "--cppopts=-IC:\Program Files\Haskell Platform\2014.2.0.0\lib\directory-1.2.1.0\include" "--cppopts=-IC:\Program Files\Haskell Platform\2014.2.0.0\lib\time-1.4.2\include" "--cppopts=-IC:\ProgramFiles\Haskell Platform\2014.2.0.0\lib\Win32-2.3.0.2\include" "--cppopts=-IC:\Program Files\Haskell Platform\2014.2.0.0\lib\bytestring-0.10.4.0\include" "--cppopts=-IC:\Program Files\Haskell Platform\2014.2.0.0\lib\base-4.7.0.1\include" "--cppopts=-IC:\Program Files\Haskell Platform\2014.2.0.0\lib\integer-gmp-0.5.1.0\include" "--cppopts=-IC:\Program Files\Haskell Platform\2014.2.0.0\lib/include" "--output-dir=dist\build" "--output=Graphics\UI\FLTK\LowLevel\HorValueSlider.hs" "src\Graphics\UI\FLTK\LowLevel\HorValueSlider.chs" A little Googling turned up this thread: http://haskell.1045720.n5.nabble.com/BUG-string-h-FC4-x86-64-td3196369.html which suggested I pass -D__attribute__(A)= to c2hs's opts. I tried it and it led to other errors. I didn't pursue any further because it's an old thread. I'm not sure what stddef.h is and I'm unfamiliar with Windows. Any help is appreciated. Thanks! -deech

2 3

Imperative graph algorithm, (tail?) recursion, and speed
by Jeffrey Brown 06 Mar '15

06 Mar '15

Suppose I have a general (not a tree) directed graph, and I would like to find all the "pure descendents" of a node N -- that is, all nodes M for which (1) N is an ancestor of M, and (2) every partial or maximal sequence of predecessors starting from M either consists entirely of descendents of N, or else includes N itself. I have an imperative algorithm (below) for doing that. I want to know how hard it would be to implement in Haskell. It's complex, but basically it sets up a few lists like "members of this list have had all their descendents processed", "members of this list have been processed but their descendents have not been", "members of this list have not been processed at all", etc. Then I iterate through the graph, moving nodes from one collection to the other until the "to be processed" list is empty. I would like a function Graph -> Graph that does that and does not need to rely on a monad. The only way I can think of involves a few mutually recursive functions, each of which passes not only the original graph of interest, but all the lists described in the previous paragraph, around to the others. I don't understand Haskell's evaluation strategy well enough to anticipate the result, but my vague sense is that a lot of copies of the same data would be floating around, ruining the speed of it. Python code for the algorithm: def descendedOnlyFrom(gset): "All Gnodes n for which every ancestry of n includes a Gnode in gset (which can be a set or a list)." # For efficiency improvement ideas and verbose comments, # see treeSearch/all.py/node.descendedOnlyFrom if 1: # variables # "pure" = "descended only from the calling Glist" pe = set(gset) # determined Pure, yet to Explore children of pf = set() # determined Pure, Finished exploring children of pb = set(gset) # determined Pure, Both kinds: pe | pf ud = set() # purity UnDetermined # descended from root, but might have parents outside of pb udp = {} # "Parents of the UnDetermined" # This is a dictionary, from gnodes to sets of gnodes. # The other four variables are sets. while pe: while pe: k = 1 i = pe.pop(); pf.add(i) for c in set(i._children()): if c in pb | ud: continue # If already handled, do not repeat. if set(c._parents()) <= pb: pe.add(c); pb.add(c) else: ud.add(c); udp[c] = set(c._parents()) - pb for i in ud: ipf = udp[i] & pb # (i)'s (p)arents newly (f)ound in pb if ipf: udp[i] -= ipf if udp[i]==set(): ud.remove(i); del( udp[i] ) pe.add(i); pb.add(i) break return pb

4 6

SPDX and hackage
by Mike Meyer 06 Mar '15

06 Mar '15

I stumbled across the SPDX (https://spdx.org/) Possibly the hackage license list should be expanded to incorporate this in some way? Or at least the OSI-approved entries?

2 1

Haskell on RedHat?
by Mario Blažević 06 Mar '15

06 Mar '15

What is the recommended route for installing a recent GHC on RedHat Enterprise Linux? All I could find on the official haskell.org Web pages is a GHC 7.8.4 repository for Fedora 21. Also, are there any potential difficulties with deploying a Haskell executable compiled on an Ubuntu system to RedHat EL, or vice versa? Thank you.

4 4

Beginner
by Sarthak Shetty 06 Mar '15

06 Mar '15

Hi everyone, this is Sarthak Shetty from India. I'm a CSE student who is somewhat comfortable with Java, C, Python and R. I would like to learn Haskell and start contributing right away. Can someone help me get started?

4 3

Lifting Writer to ListT
by Nicola Gigante 06 Mar '15

06 Mar '15

Hi all, Suppose I have this monad stack: type MyMonad = ListT (Writer W) Where ListT is done right (from the list-t package). What I want is a nondeterministic computation where each alternative logs to the same writer, and indeed the type of runWriter . toList is MyMonad a -> ([a], w) so I think I got it right. What I would like to do is to write an instance of MonadWriter for this monad, but I'm stuck with the implementation of listen. Just using lift won't work, and indeed I see that MonadWriter lifting instances for other monads (e.g MaybeT), are a bit convoluted. Could someone explain me a little bit what should I do (what listen should do in a ListT in the first place) Thank you :) Nicola

2 2

is it possible to implement Functor for ByteString and Text
by silvio 05 Mar '15

05 Mar '15

I have recently heard that some people want to burn bridges (introducing Foldable and Traversable to Prelude) and I've been wondering if it was possible somehow allow Text and Bytestring like containers to make use of those functions. Something along the lines of import qualified Data.ByteString as BS newtype ByteString' a = ByteString' BS.ByteString type ByteString = ByteString' Word8 instance (ByteString' a ~ ByteString' Word8) => Functor (ByteString') where fmap f (ByteString' bs) = ByteString' $ BS.map f bs Or if DataContexts worked as you would expect. newtype (Word8 ~ a) => ByteString' a = ByteString' BS.ByteString However I couldn't find a solution and I was just wondering if it is possible. P.S. Using GADTS it does actually work for Foldable, but only because it doesn't have to output any ByteStrings. It doesn't work for Functor for instance. data ByteString' a where ByteString' :: BS.ByteString -> ByteString' Word8 type ByteString = ByteString' Word8 instance Foldable ByteString' where foldr f ini (ByteString' bs) = BS.foldr f ini bs Silvio

7 9

Am I using Parsec correctly?
by Cody Goodman 05 Mar '15

05 Mar '15

-- trying to parse the text below using Parsec: -- ACT I. -- SCENE I. An open place. -- [An open place. Thunder and lightning. Enter three Witches.] -- FIRST WITCH. -- When shall we three meet again -- In thunder, lightning, or in rain? -- Here's the code I have import Text.Parsec import Control.Applicative import Data.Traversable (sequenceA) section s = sequenceA [string s, string " ", many1 letter] `endBy` char '.' act = section "ACT" scene = section "SCENE" main = do parseTest act "ACT I.\n" parseTest scene "SCENE I." -- this returns: -- λ> main -- [["ACT"," ","I"]] -- [["SCENE"," ","I"]] -- Am I using Parsec correctly so far? -- After this I want to match acts (many1 act) and scenes (many1 scene) and I believe I see how to do that, just wanting to make sure I'm on the right track.

2 1

Re: [Haskell-cafe] Hackage and Free Software
by Duncan Coutts 05 Mar '15

05 Mar '15

All, Hackage has of course always been for open source Haskell software (and has always rejected "AllRightsReserved" packages). Prompted by this recent discussion, Gershom, SPJ, and the hackage admins have come up with a few changes to make our current implicit open source policy a bit more explicit: * The hackage homepage will say "Hackage is the Haskell community's central package archive of open source software." * The signup and upload pages will have a new blurb on open source licenses. See below. * We will accept a patch along the lines of https://mail.haskell.org/pipermail/cabal-devel/2015-March/010019.html to continue to reject "AllRightsReserved" on hackage (though we may move the check from cabal into hackage). * And on a somewhat related topic: we will finish the discussion on hackage trustee guidelines (mostly about editing package metadata) and post them. https://github.com/haskell/hackage-server/commit/66a7acd125d486e55bb6674358… Here's the new blurb: Open source licenses The code and other material you upload and distribute via this site must be under an open source license. This is a service operated for the benefit of the community and that is our policy. It is also so that we can operate the service in compliance with copyright laws. The hackage operators do not want to be in the business of making judgements on what is and is not a valid open source license, but we retain the right to remove packages that are not under licenses that are open source in spirit, or that conflict with our ability to operate this service. (If you want advice, see the ones Cabal recommends.) The hackage operators do not need and are not asking for any rights beyond those granted by the open source license you choose to use. All normal open source licenses grant enough rights to be able to operate this service. In particular, we expect as a consequence of the license that: 1. we have the right to distribute what you have uploaded to other people 2. we have the right to distribute certain derivatives and format conversions, including but not limited to: * documentation derived from the package * alternative presentations and formats of code (e.g. html markup) * excerpts and presentation of package metadata * modified versions of package metadata Please make sure that you comply with the license of all code and other material that you upload. For example, check that your tarball includes the license files of any 3rd party code that you include. We hope this will be uncontroversial as it is just the status quo. Note that the hackage admins are not getting involved in a license debate, and we are not asking for any grant of rights (implicitly or explicitly) when you upload stuff. The open source license you use grants all the rights we need to be able to run the site. Duncan On Sat, 2015-02-28 at 19:27 +0200, fr33domlover wrote: > Hello haskellers! > > I would like to make a proposal regarding the license of software in Hackage. > > > > One of the major parts of the Haskell community infrastructure in the package > database server Hackage. As far as I know - please correct me if I'm wrong - > Hackage makes no restriction on the license of packages upload to it. But as a > community working in cooperation to make good software, the Haskell community > has embraced licenses like (L)GPL, MIT and BSD, which are free software > licenses. > > Actually the all-rights-reserved tag in Hackage [1] has only two packages > tagged by it - the dummy no-op project HNop, and another package whose COPYING > file contains a broken link and whose README says "BSD style license". > > > > Software freedom is an ethical basis for collaboration on making software > that's truly good and loayl to its users, and providing them control and > freedom to access and use their computing resources. > > It seems to me that the Haskell community is already enbracing this ethical > basis, but Hackage doesn't provide any guarantees and it means that you'd have > to check each package to be sure. By having that all-rights-reserved tag it > also in a way welcomes software that doesn't go by these rules - however it > seems that no packages do that even in the presence of the possibility. > > > > > I'd like to make a suggestion: have Hackage accept only packages released under > free software licenses. This is probably true for most/all packages there, but > making it official will both send a strong message to and from the community, > and provide people with the security and peace of mind, knowing their software > is free as in freedom. > > It is also possible that companies use Haskell to create proprietary software > using permissive-licensed libraries and tools from Hackage. I hope this isn't > true, but even if it is, this software isn't offered by Hackage and I hope its > existence doesn't affect the community's use of Hackage and free software. > > > > Would you consider to embrace free software officially, including in Hackage? > > > > Thanks for reading, > waiting to hear from the community and from haskell.org maintainers, > fr33domlover > > > > [1] http://hackage.haskell.org/packages/tag/all-rights-reserved > > --- > fr33domlover <http://www.rel4tion.org/people/fr33domlover> > GPG key ID: 63E5E57D (size: 4096) > GPG key fingerprint: 6FEE C222 7323 EF85 A49D 5487 5252 C5C8 63E5 E57D > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe

1 0

GHCi timing statistics
by David Rush 05 Mar '15

05 Mar '15

Hi all - Is it my imagination, or are the statistics printed by GHCi when :set +s has been enabled entirely spurious when SMP parallelism is turned on? I'm trying to determine if I am actually getting any improvement, and - both by the wall clock and by the reported statistics I don't seem to be; however, the reported statistics don't seem to line up very well with the wall clock time. An additional oddity that I've noted is that my algorithm's reference implementation - which simply can't run in parallel due to a foolishly introduced data dependency - still seems to kick 16 CPUs well into the 90%+ utilization range when I have +RTS -N. This code has no parallelism primitives in it whatsoever. I realize that I'm not being terribly rigorous in my description here, but that is partly because I am not sure how to report on this yet. I think my next step is going to be to build a standalone executable to check using the Unix time command. For reference I am running Haskell platform from Debian Jessie on a 16- core amd64 system with 29GiB or RAM available. This would appear to be GHC 7.6.3 :( - d

3 2