- Glasgow-haskell-users

Re: bindist for Intel MacOS X 10.4 (Tiger) with static libs
by Thorkil Naur 18 Jan '08

18 Jan '08

Hello, On Tuesday 08 January 2008 15:07, Christian Maeder wrote: > Hi, > > I've succeeded in building a binary distribution that uses static > libraries for gmp and readline. libreadline.a, libncurses.a and libgmp.a > with corresponding header files are included. (For license issues ask > someone else.) On http://gmplib.org/ we find: "GMP is distributed under the GNU LGPL. This license makes the library free to use, share, and improve, and allows you to pass on the result. The license gives freedoms, but also sets firm restrictions on the use with non-free programs." I have not attempted to check whether your distribution fulfills the requirements of the LGPL. Further, on http://cnswww.cns.cwru.edu/php/chet/readline/rltop.html: "Readline is free software, distributed under the terms of the GNU General Public License, version 2. This means that if you want to use Readline in a program that you release or distribute to anyone, the program must be free software and have a GPL-compatible license." For your distribution to adhere to this, it appears to require GHC to have a GPL-compatible license. I don't believe it does. > ... Best regards Thorkil

5 13

Re: gmp
by Thorkil Naur 17 Jan '08

17 Jan '08

Hello, On Thursday 17 January 2008 17:57, Christian Maeder wrote: > I understand that gmp is needed for the certain libraries like the > Prelude with Double and Integer. > > But I do not understand why gmp is so deeply buried in the rts. > Are the basic types Int and Pointer not enough to write a compiler like ghc? You probably know about this, but just to make sure: Did you take a look at http://hackage.haskell.org/trac/ghc/wiki/ReplacingGMPNotes? > > Cheers Christian > ... Best regards Thorkil

1 0

Re: bindist for Intel MacOS X 10.4 (Tiger) with static libs
by Thorkil Naur 17 Jan '08

17 Jan '08

Hello, On Thursday 17 January 2008 05:24, Manuel M T Chakravarty wrote: > Thorkil Naur: > > Hello, > > > > On Tuesday 08 January 2008 15:07, Christian Maeder wrote: > >> Hi, > >> > >> I've succeeded in building a binary distribution that uses static > >> libraries for gmp and readline. libreadline.a, libncurses.a and > >> libgmp.a > >> with corresponding header files are included. (For license issues ask > >> someone else.) > > > > On http://gmplib.org/ we find: > > > > "GMP is distributed under the GNU LGPL. This license makes the > > library free to > > use, share, and improve, and allows you to pass on the result. The > > license > > gives freedoms, but also sets firm restrictions on the use with non- > > free > > programs." > > > > I have not attempted to check whether your distribution fulfills the > > requirements of the LGPL. > > It does fullfil them. The source code of all components of the system > is available enabling users to build the same software with a > different version of GMP. That's all that the LGPL requires of > software linked against a LGPL library. > > > > Further, on http://cnswww.cns.cwru.edu/php/chet/readline/rltop.html: > > > > "Readline is free software, distributed under the terms of the GNU > > General > > Public License, version 2. This means that if you want to use > > Readline in a > > program that you release or distribute to anyone, the program must > > be free > > software and have a GPL-compatible license." > > > > For your distribution to adhere to this, it appears to require GHC > > to have a > > GPL-compatible license. Sorry, my use of the term "GPL-compatible" here is wrong. What I should have written was that "it appears to require GHC to be distributed under the GPL". > > I don't believe it does. > > It does. GHC's codebase is a mix of BSD3, LGLP, and GPL. They are > perfectly compatible. See <http://www.fsf.org/licensing/licenses/index_html > >. > > Manuel > Best regards Thorkil

1 0

Re: Integrating editline with ghc
by Thorkil Naur 17 Jan '08

17 Jan '08

Hello, On Wednesday 16 January 2008 22:05, Judah Jacobson wrote: > Hi all, > > I have managed to build ghc using the initial release of the editline package: > > Hackage link: http://hackage.haskell.org/cgi-bin/hackage-scripts/package/editline-0.1 > Haddock: http://code.haskell.org/editline/dist/doc/html/editline/ > > As I've mentioned before, there are two independent modules: > - System.Console.Editline is a very basic (and experimental) interface > to the native editline APIs. > - System.Console.Editline.Readline contains the readline APIs provided > by the editline library (mostly a cut/paste of > System.Console.Readline). Excellent! > > Currently I'm using just the latter as a drop-in replacement for > System.Console.Readline in ghci. I have added a --with-editline flag > to ghc's configure script, which has no effect if it's not specified, > and otherwise does the following: > > - Throw an error (at configure time) if editline isn't present (as > $hardtop/libraries/editline) That's the way. > - Use the editline package instead of readline when building ghc stage 2 > - Use CPP to make InteractiveUI.hs (the main ghci loop) import > System.Console.Editline.Readline instead of System.Console.Readline. > > Does that sound like the right way to handle this? If so, I'll send a > darcs patch. An alternative that would make the GHC configure script more symmetric with respect to command line editor would be to have --with-line-editor=editline, --with-line-editor=readline and also, perhaps, --with-line-editor=none (or even --with-line-editor=). All of these with, hopefully, obvious meanings. On top of this, one could have --with-edit-line=automatic with some automatic selection taking place. And the default? I'm sure that my favorite --with-line-editor=none will not be considered practical, so I will leave this most difficult choice to others. > > Also, should editline be made a boot-package or an extra-package (or neither)? > > Thanks, > -Judah > ... Thanks a lot again. Best regards Thorkil

1 0

GHC Data.List.sort performance question
by Marcus D. Gabriel 16 Jan '08

16 Jan '08

Hello, By a rather indirect route, I discovered that I obtain an almost factor of two improvement in performance in Data.List.sort if I make one small change in the implementation of the function merge which supports mergesort and hence sortBy and sort. Admittedly, the improvement was only noticeable to me when sorting for example one million random Int. The current code in libraries/base/Data/List.hs for merge is \begin{code} merge :: (a -> a -> Ordering) -> [a] -> [a] -> [a] merge cmp xs [] = xs merge cmp [] ys = ys merge cmp (x:xs) (y:ys) = case x `cmp` y of GT -> y : merge cmp (x:xs) ys _ -> x : merge cmp xs (y:ys) \end{code} and all that I did was \begin{code} merge :: (a -> a -> Ordering) -> [a] -> [a] -> [a] merge cmp [] ys = ys merge cmp xs [] = xs merge cmp (x:xs) (y:ys) = case x `cmp` y of GT -> y : merge cmp (x:xs) ys _ -> x : merge cmp xs (y:ys) \end{code} that is, I swapped the order of the first two lines. By the way, the second version is much less likely to overflow the stack than the first version. Can some confirm this? If you confirm it, can someone explain to me why one obtains this performance improvement? I currently just do not grasp the point. Thanks, - Marcus -- Marcus D. Gabriel, Ph.D. Email: mdg(a)wanadoo.fr 213 ter, rue de Mulhouse Tel: +33.3.89.69.05.06 F68300 Saint Louis FRANCE Portable: +33.6.34.56.07.75

7 10

'backslash' build problems on windows
by Twan van Laarhoven 16 Jan '08

16 Jan '08

I am having some problems with the latest builds of GHC on windows. For some reason ghc has decided to use backslashes in filenames. This leads to problems when building the stage libraries and the stage 2 compiler. I run the scripts from a mingw bash shell. The following problems occur: - Warnings: .../ghc3228_0/ghc3228_0.lpp:1: warning: unknown escape sequence '\B' these are caused by the #line directive in the preprocessed file, something like #line 1 ".\basicTypes/BasicTypes.lhs" - When building the dependencies (in dist/build/.depend) become something like: dist/build\GHC/PrimopWrappers.o : .\GHC/PrimopWrappers.hs make then ignores these dependencies and tries to build the wrong files. - When testing, some unexpected failures occur, because the error message says ".\Something.hs" instead of "Something.hs". The solution to all these problems is to use a '/' instead of a '\'. I suspect the problem was caused by the patch: Sat Jan 12 18:28:37 W. Europe Standard Time 2008 Ian Lynagh * FilePath fixes Twan

3 3

GHC Internals Changed
by C.M.Brown 14 Jan '08

14 Jan '08

Hi, It seems that GHC's API interface has changed between 6.6 and 6.8. Most notably, JustTypecheck and GHC.dropForAlls are no longer in scope. Are the changes documented anywhere? The notes on the hackage commentary still seem to point to the previous API internals. http://hackage.haskell.org/trac/ghc/wiki/Commentary/Compiler/API Kind regards, Chris.

2 1

can't build plugins-1.0 with ghc-6.9
by Valery V. Vorotyntsev 14 Jan '08

14 Jan '08

Hi, (The message below contains bunch of "preformatted" sections. I hope it's still readable, otherwise, please, let me know.) My quest is to install lambdabot (for the sake of offline `@hoogle' and `@pl' commands), and the one depends on `plugins'. 1) plugins.cabal should be updated ---------------------------------- A patch to .cabal file[1] is needed to get rid of config-time warning[2] and resolve build-time errors[3]. [1] vvv:~/src/plugins-1.0$ diff -u plugins.cabal{.orig,} --- plugins.cabal.orig 2008-01-06 22:23:44.657177486 +0200 +++ plugins.cabal 2008-01-06 22:25:45.160728537 +0200 @@ -33,6 +33,6 @@ src/Language/Hi/hschooks.c includes: Linker.h extensions: CPP, ForeignFunctionInterface -Build-Depends: base, Cabal, haskell-src +Build-Depends: base, Cabal, haskell-src, array, containers, directory, random, process ghc-options: -Wall -O -fasm -funbox-strict-fields -fno-warn-missing-signatures -hs-source-dir: src +hs-source-dirs: src [2] vvv:~/src/plugins-1.0$ runhaskell Setup configure --prefix=$HOME --user Warning: The field "hs-source-dir" is deprecated, please use "hs-source-dirs" Configuring plugins-1.0... [...] [3] vvv:~/src/plugins-1.0$ runhaskell Setup build Preprocessing library plugins-1.0... Building plugins-1.0... src/Language/Hi/Binary.hs:90:7: Could not find module `Data.Array.Base': it is a member of package array-0.1, which is hidden 2) it doesn't build with Cabal-1.3.2 ------------------------------------ `InstalledPackageInfo' is a type alias in GHC's Cabal[4], not a data constructor; `plugins' library[5] gets confused[6]. [4] vvv:~/src$ diff -u cabal-1.2.2.0/Distribution/InstalledPackageInfo.hs ghc/libraries/Cabal/Distribution/InstalledPackageInfo.hs --- cabal-1.2.2.0/Distribution/InstalledPackageInfo.hs 2007-11-01 17:02:35.000000000 +0200 +++ ghc/libraries/Cabal/Distribution/InstalledPackageInfo.hs 2007-12-17 17:53:21.691532864 +0200 @@ -46,7 +46,7 @@ -- This module is meant to be local-only to Distribution... module Distribution.InstalledPackageInfo ( - InstalledPackageInfo(..), + InstalledPackageInfo_(..), InstalledPackageInfo, ParseResult(..), PError(..), PWarning, emptyInstalledPackageInfo, parseInstalledPackageInfo, @@ -72,8 +72,8 @@ -- ----------------------------------------------------------------------------- -- The InstalledPackageInfo type -data InstalledPackageInfo - = InstalledPackageInfo { +data InstalledPackageInfo_ m + = InstalledPackageInfo_ { -- these parts are exactly the same as PackageDescription package :: PackageIdentifier, license :: License, @@ -87,8 +87,8 @@ category :: String, -- these parts are required by an installed package only: exposed :: Bool, - exposedModules :: [String], - hiddenModules :: [String], + exposedModules :: [m], + hiddenModules :: [m], importDirs :: [FilePath], -- contain sources in case of Hugs libraryDirs :: [FilePath], hsLibraries :: [String], @@ -107,9 +107,11 @@ } deriving (Read, Show) -emptyInstalledPackageInfo :: InstalledPackageInfo +type InstalledPackageInfo = InstalledPackageInfo_ String + +emptyInstalledPackageInfo :: InstalledPackageInfo_ m emptyInstalledPackageInfo - = InstalledPackageInfo { + = InstalledPackageInfo_ { package = PackageIdentifier "" noVersion, license = AllRightsReserved, copyright = "", [5] vvv:~/src/plugins-1.0$ sed -n 64,66p src/System/Plugins/PackageAPI.hs updImportDirs f pk@(InstalledPackageInfo { importDirs = idirs }) = pk { importDirs = f idirs } updLibraryDirs f pk@(InstalledPackageInfo { libraryDirs = ldirs }) = [6] vvv:~/src/plugins-1.0$ runhaskell Setup build Preprocessing library plugins-1.0... Building plugins-1.0... [ 1 of 24] Compiling System.Plugins.Process ( src/System/Plugins/Process.hs, dist/build/System/Plugins/Process.o ) [ 2 of 24] Compiling System.Plugins.Parser ( src/System/Plugins/Parser.hs, dist/build/System/Plugins/Parser.o ) [ 3 of 24] Compiling System.Plugins.ParsePkgConfCabal ( src/System/Plugins/ParsePkgConfCabal.hs, dist/build/System/Plugins/ParsePkgConfCabal.o ) [ 4 of 24] Compiling System.Plugins.PackageAPI ( src/System/Plugins/PackageAPI.hs, dist/build/System/Plugins/PackageAPI.o ) src/System/Plugins/PackageAPI.hs:64:20: Not in scope: data constructor `InstalledPackageInfo' src/System/Plugins/PackageAPI.hs:66:21: Not in scope: data constructor `InstalledPackageInfo' I would like to hear your suggestions. Should I try harder[7] installing Cabal-1.2.2.0? Or is this a bug of GHC's Cabal? In this case I've just reported it. :) [7] vvv:~/src/cabal-1.2.2.0$ runhaskell Setup configure --prefix=$HOME --user Configuring Cabal-1.2.2.0... Setup: At least the following dependencies are missing: base <3, filepath -any vvv:~/src/cabal-1.2.2.0$ runhaskell Setup configure --prefix=$HOME --user -f small_base Configuring Cabal-1.2.2.0... Setup: At least the following dependencies are missing: base >=3, pretty -any, directory -any, old-time -any, process -any, containers -any, filepath -any vvv:~/src/cabal-1.2.2.0$ ghc-pkg list /home/vvv/lib/ghc-6.9.20080104/package.conf: Cabal-1.3.2, array-0.1, base-3.0, bytestring-0.9, containers-0.1, directory-1.0, filepath-1.1, (ghc-6.9.20080104), haskell98-1.0.1, hpc-0.5, old-locale-1.0, old-time-1.0, packedstring-0.1, pretty-1.0, process-1.0, random-1.0, readline-1.0.1, rts-1.0, template-haskell-2.2, unix-2.2 /home/vvv/.ghc/i386-linux-6.9.20080104/package.conf: X11-1.4.1, binary-0.4.1, haskell-src-1.0.1.1, mtl-1.1.0.0, network-2.1.0.0, parsec-2.1.0.0, xmonad-0.5, xmonad-contrib-0.5 Thanks a lot. Happy hacking! -- vvv

2 3

Problem building on NetBSD Alpha - Dynamic link dependency?
by jim burton 14 Jan '08

14 Jan '08

I'm trying to build unregistered 6.8.2 on NetBSD Alpha 2.1.0_STABLE -- I have the following problem which seems to be same as this one - http://hackage.haskell.org/trac/ghc/ticket/1860 $ ./configure --enable-hc-boot --enable-hc-boot-unregistered checking build system type... alpha-unknown-netbsd2.1.0. checking host system type... alpha-unknown-netbsd2.1.0. checking target system type... alpha-unknown-netbsd2.1.0. Canonicalised to: alpha-unknown-netbsd checking for ghc... no checking for ghc-pkg matching ... no checking for ghc-pkg... no checking whether ghc has readline package... no checking for nhc... no checking for nhc98... no checking for hbc... no checking for ld... /usr/bin/ld checking for path to top of build tree... ./configure[3201]: -v0: not found ./configure[3203]: utils/pwd/pwd: not found configure: error: cannot determine current directory there is a temporary solution mentioned in Trac; "to link the binaries generated by the bootstrap compiler statically". How do I make my own statically linked versions of pwd, ghc-pkg and anything else involved? (I don't know anything about C.) Thanks, Jim

4 6

Jumping to code every and anywhere: What about installing source and a tagfile?
by Marc Weber 14 Jan '08

14 Jan '08

See http://hackage.haskell.org/trac/hackage/ticket/207 Discussion takes place on haskell-cafe Marc Weber

1 0