Re: [Haskell-cafe] cabal: internal error when reading package index
On 18 September 2013 19:23, Henk-Jan van Tuyl hjgt...@chello.nl wrote: L.S., I was trying to install a package from a local drive and got the following message: cabal install Resolving dependencies... cabal: internal error when reading package index: could not read tar file entryThe package index or index cache is probably corrupt. Running cabal update might fix it. Then I tried cabal update: cabal update Downloading the latest package list from hackage.haskell.org Skipping download: Local and remote files match. Note: there is a new version of cabal-install available. To upgrade, run: cabal install cabal-install Trying the command cabal install cabal-install gave the same internal error message as for the first command. What can I do about this? Maybe try deleting the current download? Namely ~/.cabal/packages/hackage.haskell.org/00-index.* Regards, Henk-Jan van Tuyl -- Folding@home What if you could share your unused computer power to help find a cure? In just 5 minutes you can join the world's biggest networked computer and get us closer sooner. Watch the video. http://folding.stanford.edu/ http://Van.Tuyl.eu/ http://members.chello.nl/hjgtuyl/tourdemonad.html Haskell programming -- ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com http://IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal --enable-tests
I don't think so. Perhaps we should set one. What's your use case? Perhaps you could describe it in a new bug report at https://github.com/haskell/cabal/issues On Mon, Sep 9, 2013 at 7:29 PM, satvik chauhan mystic.sat...@gmail.comwrote: Hi cafe, I wanted to ask this as I couldn't find this in cabal documentation. Is there any CCP macro set when a package is configured with --enable-testing? If not is there a way to do that? -Satvik ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev: installing/running with profiling enabled
Hi Rogan, Thanks for your reply. I don'th think the -prof flag is necessary; but if it is, then it may also be misinterpreted by cabal-dev; generally, you need to pass flags that only use a single leading - to cabal-dev with the --flags=... option. (eg: 'cabal-dev install --enable-library-profiling --enable-executable-profiling --flags=-prof' This command worked for me: $ cabal-dev install --enable-library-profiling --enable-executable-profiling OK, so I removed `cabal-dev` directory, and installed again with this command. But it still failed with same error message(the flag -p requires the program to be built with -prof). Then I also tried removing cabal-dev again and installing again but this time with extra `--flags=-prof` parameter. Failed with same error. Just to be sure I created a new cabal project with only one file: 'Hello.hs' with contents 'main = putStrLn hello' and tried to install it with cabal-dev. And `./cabal-dev-/bin/Hello +RTS -p` failed with same error again. Maybe something wrong with my system? My first thought was maybe cabal-dev is using some pre-installed library, and it's not recompiling the whole world for each sandbox. But I don't know how to check this. Any ideas on that? Thanks, --- Ömer Sinan Ağacan http://osa1.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev: installing/running with profiling enabled
On Tue, Jul 02, 2013 at 09:07:18AM +0300, Ömer Sinan Ağacan wrote: OK, so I removed `cabal-dev` directory, and installed again with this command. But it still failed with same error message(the flag -p requires the program to be built with -prof). Maybe you are running into this: https://github.com/haskell/cabal/issues/1199 Check your version of the Cabal library (cabal --version should tell you). I think this was only broken for a few releases, but I don't remember which ones. pgp4n6mPD4Zps.pgp Description: PGP signature ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev: installing/running with profiling enabled
Maybe you are running into this: https://github.com/haskell/cabal/issues/1199 Check your version of the Cabal library (cabal --version should tell you). I think this was only broken for a few releases, but I don't remember which ones. Great, it worked. I first updated Cabal library (with `cabal install cabal`) and cabal-install program(`cabal install cabal-install`) and then reinstalled cabal-dev. Now it works for my hello world program. Hopefully it will also work with my main program(it's compiling now, and it will take some time). Thanks! --- Ömer Sinan Ağacan http://osa1.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev: installing/running with profiling enabled
Hi Ömer, I've replied in-line below. On Mon, Jul 1, 2013 at 2:34 PM, Ömer Sinan Ağacan omeraga...@gmail.comwrote: So I tried installing the program in a fresh cabal-dev environment with profiling enabled, so that all dependencies would be also installed with profiling enabled. But for some reason even after installing with `cabal-dev install --enable-executable-profiling -prof` I still can't run this program with `caba-dev/bin/program +RTS -p`, it fails with 'program: the flag -p requires the program to be built with -prof'. Because you're installing libraries (into the cabal-dev sandbox) you'll also need to add the --enable-library-profiling flag, in addition to the --enable-executable-profiling flag. Cabal-dev could infer the former from the later, but we haven't done that (this doesn't seem to come up very often). I don'th think the -prof flag is necessary; but if it is, then it may also be misinterpreted by cabal-dev; generally, you need to pass flags that only use a single leading - to cabal-dev with the --flags=... option. (eg: 'cabal-dev install --enable-library-profiling --enable-executable-profiling --flags=-prof' This command worked for me: $ cabal-dev install --enable-library-profiling --enable-executable-profiling Note that you must remove the existing cabal-dev sandbox directory before running that command (or the installed dependencies that were built without library profiling can cause the build to fail. --Rogan ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal config file Guide
On 13-05-25 04:52 PM, Daniel Díaz Casanueva wrote: As you already know, cabal-install is configured in the file config. It has a lot of fields, but I didn't find a single place where each field is explained with detail. There is none, but my new and timely http://www.vex.net/~trebla/haskell/cabal-cabal.xhtml talks about a few, and one of them is a real surprise. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-install 1.16.0.2 on Mac
On Thu, Apr 11, 2013 at 1:19 AM, Richard A. O'Keefe o...@cs.otago.ac.nzwrote: On 11/04/2013, at 12:56 PM, Brandon Allbery wrote: Xcode 4.2 and on do not use /Developer at all. You have an older Xcode on your system somehow, which does not understand newer object files; you should remove the entire /Developer tree. (Xcode, in order to be distributable via the App Store, is completely self-contained in /Applications/Xcode.app.) Unfortunately, I cannot. I _am_ able to install stuff, but uninstalling generally gives me problems, and removing /Developer is something I'm not allowed to do. I think you need to discuss that with whoever made that dictum; requiring that a system be broken is not generally a good idea. Many software packages will find it and use outdated programs or frameworks as a result. It really needs to not be there at all. (Newer Xcode should actually complain and tell you to run the removal script on startup, because its presence can even break Xcode under some circumstances.) -- brandon s allbery kf8nh sine nomine associates allber...@gmail.com ballb...@sinenomine.net unix, openafs, kerberos, infrastructure, xmonadhttp://sinenomine.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-install 1.16.0.2 on Mac
The basic problem is that the University has a strict policy that academic staff must not have root access on any machine that is connected to the University network. I was given an administrator account so that I could resume the printer and install (some) stuff, but /Developer is owned by root, and I will be given root access on the Greek Calends. I would have thought that many organisations would have similar policies. On 12/04/2013, at 2:44 AM, Brandon Allbery wrote: (Newer Xcode should actually complain and tell you to run the removal script on startup, because its presence can even break Xcode under some circumstances.) 4.6.1 was the latest available in March when I installed it, and it _didn't_ complain or tell me to run any removal script. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-install 1.16.0.2 on Mac
On Thu, Apr 11, 2013 at 7:41 PM, Richard A. O'Keefe o...@cs.otago.ac.nzwrote: The basic problem is that the University has a strict policy that academic staff must not have root access on any machine that is connected to the University network. I was given an administrator account so that I could resume the printer and install (some) stuff, but /Developer is owned by root, and I will be given root access on the Greek Calends. I would have thought that many organisations would have similar policies. Well, yes (I was one of those admins, although not at your university, for many years), but if they are installing machines with both Xcode 4.6 under /Applications and Xcode 4.1 or earlier under /Developer, they are installing broken machines that will fail to build many packages and where Xcode may malfunction. /Developer should not exist on a machine with Xcode 4.2 or later installed, at all. You should contact an administrator about this and have them fix both installed machines and their installation images or maintenance routines (whatever they went with for OS X). sudo /Developer/Library/uninstall-devtools --mode=all If they need an official reference on this, I can dig up the relevant Apple knowledge base article. On 12/04/2013, at 2:44 AM, Brandon Allbery wrote: (Newer Xcode should actually complain and tell you to run the removal script on startup, because its presence can even break Xcode under some circumstances.) 4.6.1 was the latest available in March when I installed it, and it _didn't_ complain or tell me to run any removal script. I have heard that it is sometimes inconsistent about this; sadly, just because it didn't notice the older version doesn't mean the older version won't cause breakage. (As you saw.) -- brandon s allbery kf8nh sine nomine associates allber...@gmail.com ballb...@sinenomine.net unix, openafs, kerberos, infrastructure, xmonadhttp://sinenomine.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-install 1.16.0.2 on Mac
On Apr 11, 2013, at 6:53 PM, Brandon Allbery wrote: /Developer should not exist on a machine with Xcode 4.2 or later installed, at all. Unfortunately this is not completely true - there are some SDKs that still install stuff in /Developer (NVIDIA comes to mind) but it's pretty obvious that it's not XCode-related. Just because you have /Developer present doesn't mean you're harboring an old XCode. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-install 1.16.0.2 on Mac
On Wed, Apr 10, 2013 at 8:36 PM, Richard A. O'Keefe o...@cs.otago.ac.nzwrote: /Developer/usr/bin/strip: object: /home/cshome/o/ok/.cabal/bin/cabal malformed object (unknown load command 15) Xcode 4.2 and on do not use /Developer at all. You have an older Xcode on your system somehow, which does not understand newer object files; you should remove the entire /Developer tree. (Xcode, in order to be distributable via the App Store, is completely self-contained in /Applications/Xcode.app.) -- brandon s allbery kf8nh sine nomine associates allber...@gmail.com ballb...@sinenomine.net unix, openafs, kerberos, infrastructure, xmonadhttp://sinenomine.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-install 1.16.0.2 on Mac
On 11/04/2013, at 12:56 PM, Brandon Allbery wrote: Xcode 4.2 and on do not use /Developer at all. You have an older Xcode on your system somehow, which does not understand newer object files; you should remove the entire /Developer tree. (Xcode, in order to be distributable via the App Store, is completely self-contained in /Applications/Xcode.app.) Unfortunately, I cannot. I _am_ able to install stuff, but uninstalling generally gives me problems, and removing /Developer is something I'm not allowed to do. However, putting /Applications/Xcode.app/Contents/Developer/usr/bin at the front of my $PATH seems to do the job. Thanks. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal install pandoc
Hello Albert, On 04/01/2013 11:41 PM, Albert Y. C. Lai wrote: On 13-04-01 06:26 AM, Roger Mason wrote: It turned out that there was a stale version of 'array' lurking in the ghc package db. In spite of reinstalling ghc it did not go away until I unregistered it. I think it was persisting because re-installing ghc simply unpacked over the old directory leaving that pre-existing file intact. See my http://www.vex.net/~trebla/haskell/sicp.xhtml for how does GHC know or not know what libs you have. In particular, it has very little to do with files, and clearing GHC is only half the story. And how to have the same kind of problems recur in the future. Thank you. I have read and filed away the article for future reference. I guess the best (least error prone) method of installing ghc and packages is to obtain a binary ghc (outside one's package manager), build haskell platform and then maintain ghc and packages outside the distro package manager. Comments welcome. Roger This electronic communication is governed by the terms and conditions at http://www.mun.ca/cc/policies/electronic_communications_disclaimer_2012.php ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc
Hello Brent, On 03/31/2013 04:53 PM, Brent Yorgey wrote: It looks like your entire Haskell Platform installation is completely hosed. Sad to say, but I think your best bet is to simply reinstall the Haskell Platform. -Brent ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe It turned out that there was a stale version of 'array' lurking in the ghc package db. In spite of reinstalling ghc it did not go away until I unregistered it. I think it was persisting because re-installing ghc simply unpacked over the old directory leaving that pre-existing file intact. 'ghc-pkg check' shows no errors and I have successfully installed pandoc and some other packages. Roger This electronic communication is governed by the terms and conditions at http://www.mun.ca/cc/policies/electronic_communications_disclaimer_2012.php ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc
On 13-04-01 06:26 AM, Roger Mason wrote: It turned out that there was a stale version of 'array' lurking in the ghc package db. In spite of reinstalling ghc it did not go away until I unregistered it. I think it was persisting because re-installing ghc simply unpacked over the old directory leaving that pre-existing file intact. See my http://www.vex.net/~trebla/haskell/sicp.xhtml for how does GHC know or not know what libs you have. In particular, it has very little to do with files, and clearing GHC is only half the story. And how to have the same kind of problems recur in the future. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc
On Fri, Mar 29, 2013 at 08:05:47AM -0230, Roger Mason wrote: Thank you for your response. 'ghc-pkg check' shows some problems: http://pastebin.ca/2344794 On 03/28/2013 08:01 PM, Patrick Wheeler wrote: So I printed off the requirements for pandoc on a empty ghc-7.6.2 install you can find it at: http://hpaste.org/84794 I do not see any odd package versions listed in what you posted so far. No promise I will be able to help afterwards but it might help to see the full log, and then again with verbosity turned on. So seperate pastes for: * `cabal install pandoc --dry-run` * `cabal install pandoc --dry-run --verbose=2` * `cabal install pandoc --dry-run --verbose=3` You might also want to run a `ghc-pkg check` to check to see if your packages are consistent/unbroken. 'ghc-pkg check' shows some problems: http://pastebin.ca/2344794 It looks like your entire Haskell Platform installation is completely hosed. Sad to say, but I think your best bet is to simply reinstall the Haskell Platform. -Brent ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc
On Thu, 28 Mar 2013 19:08:46 +0100, Roger Mason rma...@mun.ca wrote: I installed ghc (7.6.2) on an Arch Linux machine. I'm trying to install pandoc via cabal but it fails: ... Configuring text-0.11.2.3... Warning: This package indirectly depends on multiple versions of the same package. This is highly likely to cause a compile failure. package deepseq-1.3.0.1 requires array-0.4.0.1 package text-0.11.2.3 requires array-0.4.0.1 Building text-0.11.2.3... Preprocessing library text-0.11.2.3... command line: cannot satisfy -package-id array-0.4.0.1-db49bb8b0087ae85b5875d4c0cc12874 (use -v for more information) Failed to install text-0.11.2.3 ... I had something similar with Ubuntu (before there was a binary package available for this platform); I installed several packages, that gave such message, again. That solved it. Regards, Henk-Jan van Tuyl -- http://Van.Tuyl.eu/ http://members.chello.nl/hjgtuyl/tourdemonad.html Haskell programming -- ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc
Thank you for your response. 'ghc-pkg check' shows some problems: http://pastebin.ca/2344794 On 03/28/2013 08:01 PM, Patrick Wheeler wrote: So I printed off the requirements for pandoc on a empty ghc-7.6.2 install you can find it at: http://hpaste.org/84794 I do not see any odd package versions listed in what you posted so far. No promise I will be able to help afterwards but it might help to see the full log, and then again with verbosity turned on. So seperate pastes for: * `cabal install pandoc --dry-run` * `cabal install pandoc --dry-run --verbose=2` * `cabal install pandoc --dry-run --verbose=3` You might also want to run a `ghc-pkg check` to check to see if your packages are consistent/unbroken. 'ghc-pkg check' shows some problems: http://pastebin.ca/2344794 Thanks for any help you can offer. Roger This electronic communication is governed by the terms and conditions at http://www.mun.ca/cc/policies/electronic_communications_disclaimer_2012.php ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc
Hello, On 03/29/2013 06:47 AM, Henk-Jan van Tuyl wrote: On Thu, 28 Mar 2013 19:08:46 +0100, Roger Mason rma...@mun.ca wrote: I installed ghc (7.6.2) on an Arch Linux machine. I'm trying to install pandoc via cabal but it fails: ... Configuring text-0.11.2.3... Warning: This package indirectly depends on multiple versions of the same package. This is highly likely to cause a compile failure. package deepseq-1.3.0.1 requires array-0.4.0.1 package text-0.11.2.3 requires array-0.4.0.1 Building text-0.11.2.3... Preprocessing library text-0.11.2.3... command line: cannot satisfy -package-id array-0.4.0.1-db49bb8b0087ae85b5875d4c0cc12874 (use -v for more information) Failed to install text-0.11.2.3 ... I had something similar with Ubuntu (before there was a binary package available for this platform); I installed several packages, that gave such message, again. That solved it. Regards, Henk-Jan van Tuyl It appears in my case that cabal may be looking in a strange place for installed pacckages. At least, that is how I interpret the output I just pasted here: http://pastebin.ca/2344794 Thanks, Roger This electronic communication is governed by the terms and conditions at http://www.mun.ca/cc/policies/electronic_communications_disclaimer_2012.php ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc [solved]
Hello, On 03/29/2013 08:13 AM, Roger Mason wrote: Hello, It appears in my case that cabal may be looking in a strange place for installed pacckages. At least, that is how I interpret the output I just pasted here: http://pastebin.ca/2344794 Thanks, Roger ghc-pkg check showed that there were problems with 'array'. ghc-pkg unregister and a fresh installation of ghc and cabal-install have fixed the problem. Thanks to all who responded. Roger This electronic communication is governed by the terms and conditions at http://www.mun.ca/cc/policies/electronic_communications_disclaimer_2012.php ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc
To side step the issue, Pandoc is available via the ArchHaskell repos (package name `haskell-pandoc`): https://wiki.archlinux.org/index.php/Haskell_package_guidelines -M ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc
hello, On 03/28/2013 04:11 PM, Mark Fredrickson wrote: To side step the issue, Pandoc is available via the ArchHaskell repos (package name `haskell-pandoc`): https://wiki.archlinux.org/index.php/Haskell_package_guidelines -M Yes, I know. I wanted to avoid having a mixture of packages installed by pacman and others (not available in the repo) installed using cabal. Thanks, Roger This electronic communication is governed by the terms and conditions at http://www.mun.ca/cc/policies/electronic_communications_disclaimer_2012.php ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install pandoc
So I printed off the requirements for pandoc on a empty ghc-7.6.2 install you can find it at: http://hpaste.org/84794 I do not see any odd package versions listed in what you posted so far. No promise I will be able to help afterwards but it might help to see the full log, and then again with verbosity turned on. So seperate pastes for: * `cabal install pandoc --dry-run` * `cabal install pandoc --dry-run --verbose=2` * `cabal install pandoc --dry-run --verbose=3` You might also want to run a `ghc-pkg check` to check to see if your packages are consistent/unbroken. On Thu, Mar 28, 2013 at 7:52 PM, Roger Mason rma...@mun.ca wrote: hello, On 03/28/2013 04:11 PM, Mark Fredrickson wrote: To side step the issue, Pandoc is available via the ArchHaskell repos (package name `haskell-pandoc`): https://wiki.archlinux.org/**index.php/Haskell_package_**guidelineshttps://wiki.archlinux.org/index.php/Haskell_package_guidelines -M Yes, I know. I wanted to avoid having a mixture of packages installed by pacman and others (not available in the repo) installed using cabal. Thanks, Roger This electronic communication is governed by the terms and conditions at http://www.mun.ca/cc/policies/**electronic_communications_** disclaimer_2012.phphttp://www.mun.ca/cc/policies/electronic_communications_disclaimer_2012.php __**_ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/**mailman/listinfo/haskell-cafehttp://www.haskell.org/mailman/listinfo/haskell-cafe -- Patrick Wheeler patrick.john.whee...@gmail.com patrick.j.whee...@rice.edu patrick.whee...@colorado.edu ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install oddities
On Tue, Mar 12, 2013 at 3:21 PM, Tycho Andersen ty...@tycho.ws wrote: Below is some sample output from a failing package: ps168825:~/playground$ cabal install network Resolving dependencies... Configuring network-2.4.1.2... configure: WARNING: unrecognized options: --with-compiler, --with-gcc checking build system type... x86_64-unknown-linux-gnu checking host system type... x86_64-unknown-linux-gnu checking for gcc... gcc checking whether the C compiler works... yes checking for C compiler default output file name... a.out checking for suffix of executables... checking whether we are cross compiling... configure: error: in `/tmp/network-2.4.1.2-28534/network-2.4.1.2': configure: error: cannot run C compiled programs. cabal install unpacks a package into /tmp in order to build it. My guess is your OS has /tmp mounted noexec. I don't know offhand how you override this in cabal. -- brandon s allbery kf8nh sine nomine associates allber...@gmail.com ballb...@sinenomine.net unix, openafs, kerberos, infrastructure, xmonadhttp://sinenomine.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install oddities
On Tue, Mar 12, 2013 at 03:28:08PM -0400, Brandon Allbery wrote: cabal install unpacks a package into /tmp in order to build it. My guess is your OS has /tmp mounted noexec. I don't know offhand how you override this in cabal. Yep, you're exactly right. Thank you! I also couldn't figure out a way to override it. \t ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install ghc-mod installs 3 years old version
Doesn't Cabal tend to install library packages under the .cabal folder? So blowing it away gets rid of the problematic ones. (And everything else.) On 25 Feb 2013, at 16:56, Brent Yorgey wrote: On Sun, Feb 24, 2013 at 02:33:55PM +, Niklas Hambüchen wrote: You are right, my ghc-7.4.2 was broken in ghc-pkg list; I fixed the problem by killing my .cabal folder (as so often). Surely you mean by killing your .ghc folder? I do not see what effect killing your .cabal folder could possibly have on broken packages. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install ghc-mod installs 3 years old version
On 13-03-01 05:10 AM, Malcolm Wallace wrote: Doesn't Cabal tend to install library packages under the .cabal folder? So blowing it away gets rid of the problematic ones. (And everything else.) You need to perform scientific experiments to refute that claim, then see my http://www.vex.net/~trebla/haskell/sicp.xhtml#ident then perform more scientific experiments to try to refute my claim (and see that my claim passes your scrutiny). ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install ghc-mod installs 3 years old version
On Fri, Mar 1, 2013 at 12:08 PM, Albert Y. C. Lai tre...@vex.net wrote: On 13-03-01 05:10 AM, Malcolm Wallace wrote: Doesn't Cabal tend to install library packages under the .cabal folder? So blowing it away gets rid of the problematic ones. (And everything else.) You need to perform scientific experiments to refute that claim, then see my At least some versions of cabal-install do put the actual library install trees under .cabal/lib, then register them under .ghc. -- brandon s allbery kf8nh sine nomine associates allber...@gmail.com ballb...@sinenomine.net unix, openafs, kerberos, infrastructure, xmonadhttp://sinenomine.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install ghc-mod installs 3 years old version
On Sun, Feb 24, 2013 at 02:33:55PM +, Niklas Hambüchen wrote: You are right, my ghc-7.4.2 was broken in ghc-pkg list; I fixed the problem by killing my .cabal folder (as so often). Surely you mean by killing your .ghc folder? I do not see what effect killing your .cabal folder could possibly have on broken packages. -Brent ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install ghc-mod installs 3 years old version
Yep, I usually kill ~/.ghc and ~/.cabal for this kind of reset. On Mon 25 Feb 2013 16:56:56 GMT, Brent Yorgey wrote: On Sun, Feb 24, 2013 at 02:33:55PM +, Niklas Hambüchen wrote: You are right, my ghc-7.4.2 was broken in ghc-pkg list; I fixed the problem by killing my .cabal folder (as so often). Surely you mean by killing your .ghc folder? I do not see what effect killing your .cabal folder could possibly have on broken packages. -Brent ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install ghc-mod installs 3 years old version
You are right, my ghc-7.4.2 was broken in ghc-pkg list; I fixed the problem by killing my .cabal folder (as so often). Do you know if it is possible to make ghc-pkg list print some actual text when packages are broken instead of writing them in red (which goes away on output redirection)? Thanks Niklas On 24/02/13 07:34, Ivan Lazar Miljenovic wrote: Which version of GHC (and hence base, etc.)? My guess is that for some reason it thinks that some requirement of the later versions is incompatible with your version of GHC. Maybe explicitly try cabal install 'ghc-mod = 1.11.4' and see why it doesn't like it. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install ghc-mod installs 3 years old version
On 25 February 2013 01:33, Niklas Hambüchen m...@nh2.me wrote: You are right, my ghc-7.4.2 was broken in ghc-pkg list; I fixed the problem by killing my .cabal folder (as so often). Do you know if it is possible to make ghc-pkg list print some actual text when packages are broken instead of writing them in red (which goes away on output redirection)? Just run ghc-pkg check every now and then? Thanks Niklas On 24/02/13 07:34, Ivan Lazar Miljenovic wrote: Which version of GHC (and hence base, etc.)? My guess is that for some reason it thinks that some requirement of the later versions is incompatible with your version of GHC. Maybe explicitly try cabal install 'ghc-mod = 1.11.4' and see why it doesn't like it. -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com http://IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install ghc-mod installs 3 years old version
On 24 February 2013 12:38, Niklas Hambüchen m...@nh2.me wrote: Hi, I just did cabal update and cabal install ghc-mod, and for some reason it tries to install version 0.3.0 from 3 years ago: cabal install ghc-mod -v Reading available packages... Choosing modular solver. Resolving dependencies... Ready to install ghc-mod-0.3.0 Downloading ghc-mod-0.3.0... cabal --version cabal-install version 1.16.0.2 using version 1.16.0.3 of the Cabal library Does anyone have an idea why that could be? Which version of GHC (and hence base, etc.)? My guess is that for some reason it thinks that some requirement of the later versions is incompatible with your version of GHC. Maybe explicitly try cabal install 'ghc-mod = 1.11.4' and see why it doesn't like it. Thanks ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Ivan Lazar Miljenovic ivan.miljeno...@gmail.com http://IvanMiljenovic.wordpress.com ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev add-source
You need to call cabal-dev add-source on P1 again to copy over the sdist, then do a cabal-dev install. See notes under Using a sandbox-local Hackage on https://github.com/creswick/cabal-dev On Feb 8, 2013 2:22 PM, JP Moresmau jpmores...@gmail.com wrote: Hello, I'm trying to understand cabal-dev, and I seem to be missing some basic point, because I can't get dependencies between projects working properly. I have two projects, let's call them P1 and P2. P2 depends on P1, as indicated by its cabal file build-depends field. I run cabal-dev add-source ..\P1 inside P2 then cabal-dev install Everything works fine, and my project compiles, with modules in P2 calling functions from P1. Then I add a new function if an exposed module of P1. I rerun the cabal-dev add-source and cabal-dev install commands to tell P2 of the change, as per the cabal-dev documentation. I change a file in P2 to use the new function, the compilation fails. If I check into the cabal-dev folder, I see the .hi file for my changed module has not been updated (old date). What am I missing? Thanks -- JP Moresmau http://jpmoresmau.blogspot.com/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev add-source
On Fri, Feb 8, 2013 at 9:53 AM, Blake Rain blake.r...@gmail.com wrote: You need to call cabal-dev add-source on P1 again to copy over the sdist, then do a cabal-dev install. See notes under Using a sandbox-local Hackage on https://github.com/creswick/cabal-dehttps://github.com/creswick/cabal-dev With the new cabal sandboxing (due in 1.18) this won't be necessary as we create a link to the repo, instead of installing a copy. We will rebuild the linked repo as needed. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev add-source
That's exactly what I'm doing, and I was exactly following these notes, but it doesn't work. This google+ post and the answers to it ( https://plus.google.com/102016502921512042165/posts/TGaENqWfubP) lead me to at least one solution that seems to work: you need to unregister the changed package first. So cabal-dev ghc-pkg unregister --force P1 cabal-dev add-source ../P1 cabal-dev install Does the trick. There seems to be a way to pass a --reinstall flag to cabal-dev install but I haven't gotten it to work yet. Thanks JP On Fri, Feb 8, 2013 at 6:53 PM, Blake Rain blake.r...@gmail.com wrote: You need to call cabal-dev add-source on P1 again to copy over the sdist, then do a cabal-dev install. See notes under Using a sandbox-local Hackage on https://github.com/creswick/cabal-dev On Feb 8, 2013 2:22 PM, JP Moresmau jpmores...@gmail.com wrote: Hello, I'm trying to understand cabal-dev, and I seem to be missing some basic point, because I can't get dependencies between projects working properly. I have two projects, let's call them P1 and P2. P2 depends on P1, as indicated by its cabal file build-depends field. I run cabal-dev add-source ..\P1 inside P2 then cabal-dev install Everything works fine, and my project compiles, with modules in P2 calling functions from P1. Then I add a new function if an exposed module of P1. I rerun the cabal-dev add-source and cabal-dev install commands to tell P2 of the change, as per the cabal-dev documentation. I change a file in P2 to use the new function, the compilation fails. If I check into the cabal-dev folder, I see the .hi file for my changed module has not been updated (old date). What am I missing? Thanks -- JP Moresmau http://jpmoresmau.blogspot.com/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- JP Moresmau http://jpmoresmau.blogspot.com/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev add-source
Johan, thanks, that brings me to a point that I wanted to raise. I'm playing with cabal-dev because users have asked me to add support for it in EclipseFP (so projects could have their own sandbox and have dependencies between projects without polluting the main package databases). It is worth it, or should I just wait for cabal 1.18 and use the sandboxing facility? Or will the two work similarly enough that supporting both will be easy? Does the sandboxing in cabal means that tools like cabal-dev are going to get deprecated? Thanks JP On Fri, Feb 8, 2013 at 7:02 PM, Johan Tibell johan.tib...@gmail.com wrote: On Fri, Feb 8, 2013 at 9:53 AM, Blake Rain blake.r...@gmail.com wrote: You need to call cabal-dev add-source on P1 again to copy over the sdist, then do a cabal-dev install. See notes under Using a sandbox-local Hackage on https://github.com/creswick/cabal-dehttps://github.com/creswick/cabal-dev With the new cabal sandboxing (due in 1.18) this won't be necessary as we create a link to the repo, instead of installing a copy. We will rebuild the linked repo as needed. -- JP Moresmau http://jpmoresmau.blogspot.com/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev add-source
On Fri, Feb 8, 2013 at 10:07 AM, JP Moresmau jpmores...@gmail.com wrote: Johan, thanks, that brings me to a point that I wanted to raise. I'm playing with cabal-dev because users have asked me to add support for it in EclipseFP (so projects could have their own sandbox and have dependencies between projects without polluting the main package databases). It is worth it, or should I just wait for cabal 1.18 and use the sandboxing facility? Or will the two work similarly enough that supporting both will be easy? Does the sandboxing in cabal means that tools like cabal-dev are going to get deprecated? I think they will be similar enough that you could easily port the code. The new cabal sandboxing will work as follows: cabal sandbox --init cabal add-source dir and then you use cabal commands like normal (e.g. configure, build, test). No installing necessary. I cannot speak for the cabal-dev developers. We do intend to support a superset of the cabal-dev functionality eventually. What we're missing now is ghci support. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev add-source
On Friday, February 8, 2013 at 1:19 PM, Johan Tibell wrote: On Fri, Feb 8, 2013 at 10:07 AM, JP Moresmau jpmores...@gmail.com (mailto:jpmores...@gmail.com) wrote: Johan, thanks, that brings me to a point that I wanted to raise. I'm playing with cabal-dev because users have asked me to add support for it in EclipseFP (so projects could have their own sandbox and have dependencies between projects without polluting the main package databases). It is worth it, or should I just wait for cabal 1.18 and use the sandboxing facility? Or will the two work similarly enough that supporting both will be easy? Does the sandboxing in cabal means that tools like cabal-dev are going to get deprecated? I think they will be similar enough that you could easily port the code. The new cabal sandboxing will work as follows: cabal sandbox --init cabal add-source dir and then you use cabal commands like normal (e.g. configure, build, test). No installing necessary. I cannot speak for the cabal-dev developers. We do intend to support a superset of the cabal-dev functionality eventually. What we're missing now is ghci support. Which, thanks to Johan's help yesterday, can still be worked around (for now) by starting ghci with: ghci -package-conf ./cabal-sandbox/your-package-conf-folder-here/ I'm trying to get Emacs haskell-mode and inferior-haskell to play nice with this, but it's not working so far for some reason. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org (mailto:Haskell-Cafe@haskell.org) http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-dev add-source
On Fri, Feb 8, 2013 at 10:24 AM, Ozgun Ataman ozata...@gmail.com wrote: Which, thanks to Johan's help yesterday, can still be worked around (for now) by starting ghci with: ghci -package-conf ./cabal-sandbox/your-package-conf-folder-here/ You can indeed do this. For real ghci support in cabal we need to also pass -package flags, C libraries, etc to ghci. That's why it's not done yet. -- Johan ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install choosing an older version
Hi Ozgur, I'm missing some context here, but I'll release an updated version of hspec ASAP ;) Cheers, Simon ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install choosing an older version
Aha! I think I know why this happens. The latest versions of ansi-terminal and hspec do not work together. Cabal picks the latest ansi-terminal (0.6) first, then the latest hspec that doesn't conflict with this choice is 0.3.0. I can confirm this by the following: $ cabal install hspec ansi-terminal --dry-run -v Reading available packages... Choosing modular solver. Resolving dependencies... In order, the following would be installed: HUnit-1.2.5.1 (new package) ansi-terminal-0.5.5.1 (new package) hspec-expectations-0.3.0.3 (new package) random-1.0.1.1 (new package) QuickCheck-2.5.1.1 (new package) setenv-0.1.0 (new package) silently-1.2.4.1 (new package) transformers-0.3.0.0 (new package) hspec-1.4.3 (new package) When hspec comes before ansi-terminal, the latest version for hspec is selected and an older version of ansi-terminal is used. Maybe cabal-install should backtrack more and pick a *more optimal *set of latest versions, I don't know. If this is desired, a proximity of the selected versions to the latest available versions might be a good measure. Best, Ozgur ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install choosing an older version
On 25 January 2013 14:46, Ozgur Akgun ozgurak...@gmail.com wrote: The latest versions of ansi-terminal and hspec do not work together. Cabal picks the latest ansi-terminal (0.6) first, then the latest hspec that doesn't conflict with this choice is 0.3.0. If this happens because the dependency bounds of ansi-terminal are too tight then please send me a patch. Cheers, Max ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install choosing an older version
Hi Max, On 25 January 2013 15:58, Max Bolingbroke batterseapo...@hotmail.comwrote: If this happens because the dependency bounds of ansi-terminal are too tight then please send me a patch. No, actually it happens because hspec depends on ansi-terminal-0.5.*. I am cc'ing Simon Hengel, the maintainer of hspec so he is aware of this. Best, Ozgur ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal sdist warns about optimization levels
On Sunday 13 January 2013, 21:27:44, Petr P wrote:
I wonder:
(1) Is there a way how to disable the warning? As the main aim of the
library is speed, I believe -O2 is appropriate here. And since the code is
quite short, I'm quite sure the increased compile time won't be noticeable.
(2) Why does cabal complain about it at the first place? I found a
reference saying the warning is adequate:
https://github.com/haskell/cabal/issues/808
but not saying why. Maybe for complex programs -O2 prolongs compile time
too much, but libraries are usually compiled once and used many times, so
using -O2 for them seems reasonable in many cases.
Sometimes compiling with -O2 instead of just -O takes significantly longer.
Not always is the produced result faster (often, the results are identical).
So if the code produced with -O performs equally to that produced with -O2,
and the -O2 compilation takes significantly longer, choosing -O2 imposes a
cost for no benefit.
That's, I think, why the warning is considered adequate.
You can specify -O2 on a per-module basis with an
{-# OPTIONS_GHC -O2 #-}
pragma where it matters, then cabal won't complain.
Or, if you're too lazy to check the consequences of -O2 vs. -O for each module
(like I usually am, if there are more than a handful), just verify that -O2
does indeed make a significant difference for the speed of the result in some
places without increasing compile time unduly, and henceforth ignore the
warning if it does. (Re-test every couple of compiler versions.)
After some time, you tend to not even notice it anymore ;)
Cheers,
Daniel
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal sdist warns about optimization levels
On 13 January 2013 20:27, Petr P petr@gmail.com wrote: to the cabal file. Now cabal sdist complains with: 'ghc-options: -O2' is rarely needed. Check that it is giving a real benefit and not just imposing longer compile times on your users. I wonder: (1) Is there a way how to disable the warning? As the main aim of the library is speed, I believe -O2 is appropriate here. And since the code is quite short, I'm quite sure the increased compile time won't be noticeable. No, but you can just ignore it. You clearly have checked and you're satisfied it's the right thing to do, so it's fine. You don't need to hit 0 warnings, nobody is going to give you or your package black marks becuase of it! :-) (2) Why does cabal complain about it at the first place? There's lots of programs where it makes no measurable difference except to make compile times longer. To some extent it's to try to break the habbit of C programmers who always default to -O2. With gcc -O2 will almost always be significantly better than -O, but with ghc that's not the case: -O is the sensible default (almost by definition, to a first approximation, things that are always a win get put into -O, things that are sometimes a win and sometimes not go into -O2). Duncan ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal bug? repeat --reinstall
Where do we report this? On 05/01/13 02:36, Albert Y. C. Lai wrote: On 13-01-04 04:36 PM, Niklas Hambüchen wrote: I get the following: $ cabal install --only-dependencies --reinstall Resolving dependencies... All the requested packages are already installed: Use --reinstall if you want to reinstall anyway. Can somebody confirm that they see the same? I confirm. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal bug? repeat --reinstall
Ahah on Github. Filed as https://github.com/haskell/cabal/issues/1175. On Fri 11 Jan 2013 01:21:11 CET, Niklas Hambüchen wrote: Where do we report this? On 05/01/13 02:36, Albert Y. C. Lai wrote: On 13-01-04 04:36 PM, Niklas Hambüchen wrote: I get the following: $ cabal install --only-dependencies --reinstall Resolving dependencies... All the requested packages are already installed: Use --reinstall if you want to reinstall anyway. Can somebody confirm that they see the same? I confirm. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal bug? repeat --reinstall
On 13-01-04 04:36 PM, Niklas Hambüchen wrote: I get the following: $ cabal install --only-dependencies --reinstall Resolving dependencies... All the requested packages are already installed: Use --reinstall if you want to reinstall anyway. Can somebody confirm that they see the same? I confirm. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal: Retrieve directory in which binaries get installed
There's getBinDir in the generated module Paths_pkgname (dist/build/autogen/Paths_pkgname.hs). This module is generated by cabal configure and can be imported in the modules of your package. Roman * Björn Peemöller b...@informatik.uni-kiel.de [2012-12-17 11:05:52+0100] Hello cafe, I'd like to retrieve the path in which binaries get installed when I `cabal install` a package. Normally, this would be ~/.cabal/bin on my Linux, but since I share my home folder on two architectures, in ~/.cabal/config I set the user-install prefix to ~/.cabal/$arch, so binaries get installed to ~/.cabal/$arch/bin. I have the feeling that it should be rather simple to retrieve the FilePath from the Cabal API, does anyone know how to do this? Thanks in advance, Björn ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal configure cabal build cabal install
On Mon, Nov 26, 2012 at 06:21:33PM -0500, Albert Y. C. Lai wrote: cabal configure is used by a lot of programmers. Today. Why? Because they use it on their own projects. They use cabal-install as a builder, not exactly an installer. Don't most devs nowadays use sandboxing, a.k.a. cabal-dev? I know I do, I've found cabal to be impossible to work with as a development tool without sandboxing. I think there was effort to bring cabal-dev like functionality to standard cabal in the next big release. Ah yes, here: http://hackage.haskell.org/trac/hackage/wiki/SandboxedBuildsAndIsolatedEnvironments In general, I consider sandboxing really important, even without dependency hell: I will sometimes want experimental and/or old versions of libraries to depend on that I don't want installed globally. Best, Aleks signature.asc Description: Digital signature ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal configure cabal build cabal install
On Tue, 27 Nov 2012 02:20:35 -0500 Albert Y. C. Lai tre...@vex.net wrote: When cabal build succeeds, it always says: (older) registering name-version (newer) In-place registering name-version That's what it says. But use ghc-pkg and other tests to verify that no registration whatsoever has happened. It doesn't register in user package-db, it registers in it's own dist/package.conf.inplace. If it didn't you wouldn't be able to build an executable and a library in one package such that executable depends on the library. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal configure cabal build cabal install
On 12-11-27 04:40 AM, kudah wrote: On Tue, 27 Nov 2012 02:20:35 -0500 Albert Y. C. Lai tre...@vex.net wrote: When cabal build succeeds, it always says: (older) registering name-version (newer) In-place registering name-version That's what it says. But use ghc-pkg and other tests to verify that no registration whatsoever has happened. It doesn't register in user package-db, it registers in it's own dist/package.conf.inplace. If it didn't you wouldn't be able to build an executable and a library in one package such that executable depends on the library. That's fair. But it also means cabal configure cabal build is not equivalent to cabal configure cabal build cabal register --inplace which was the context when you said (newer) cabal build registers inplace automatically. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal configure cabal build cabal install
Brent Yorgey byor...@seas.upenn.edu writes: On Sun, Nov 25, 2012 at 06:09:26PM -0500, Albert Y. C. Lai wrote: If you begin with cabal configure, the correct idiom is: cabal configure [flags] cabal build [cabal haddock, if you want] cabal copy cabal register Even this does not do the same thing as 'cabal install', because it does not download and install any dependencies (whereas 'cabal install' does). ...but if he prepended a 'cabal install --only-dependencies' to the invocation sequence it would, wouldn't it? ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal configure cabal build cabal install
Nice tip, Albert! Good to know! One question I have is, is (runghc Setup.lhs) equivalent to (cabal) in runghc Setup.lhs $ [configure, build, install] ? On Mon, Nov 26, 2012 at 8:08 AM, Brent Yorgey byor...@seas.upenn.eduwrote: [cabal haddock, if you want] cabal copy cabal register Even this does not do the same thing as 'cabal install', because it does not download and install any dependencies (whereas 'cabal install' does). Brent, that's useful to know too, thanks! Fwiw, I think Albert had the backdrop of classic GNU autoconf in mind, predating all that newfangled stuff of downloading (!) dependencies (!!). -- Kim-Ee -Brent ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal configure cabal build cabal install
On 12-11-26 04:34 AM, Kim-Ee Yeoh wrote: Nice tip, Albert! Good to know! One question I have is, is (runghc Setup.lhs) equivalent to (cabal) in runghc Setup.lhs $ [configure, build, install] ? Setup defaults to --global --prefix=/usr/local cabal defaults to --user --prefix=$HOME/.cabal This confuses a lot of people. FAQ #1: Why does Setup copy abort and say no permission? Answer: because you haven't escalated privilege for writing to /usr/local FAQ #2: Why does sudo cabal install register nothing in both the global database and my database? Answer: because --user means not global, and sudo means the user is root, not you. Look under /root. Lastly, there is no Setup install. Use copy and register. On Mon, Nov 26, 2012 at 8:08 AM, Brent Yorgey byor...@seas.upenn.edu mailto:byor...@seas.upenn.edu wrote: [cabal haddock, if you want] cabal copy cabal register Even this does not do the same thing as 'cabal install', because it does not download and install any dependencies (whereas 'cabal install' does). Brent, that's useful to know too, thanks! Fwiw, I think Albert had the backdrop of classic GNU autoconf in mind, predating all that newfangled stuff of downloading (!) dependencies (!!). This is ignorant of a common workflow. cabal configure is used by a lot of programmers. Today. Why? Because they use it on their own projects. They use cabal-install as a builder, not exactly an installer. In fact, some of them do: cabal configure cabal build cabal register --inplace This has no cabal install correspondence. So you ask, but surely their own projects require some packages from hackage? Yes, surely. But those packages have already been installed in the past, once and for all. That is, when the project started, they already did cabal install yesod. This is not so old-school, is it? ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal configure cabal build cabal install
On Mon, 26 Nov 2012 18:21:33 -0500 Albert Y. C. Lai tre...@vex.net wrote: Lastly, there is no Setup install. Use copy and register. $ runghc Setup.hs --help This Setup program uses the Haskell Cabal Infrastructure. See http://www.haskell.org/cabal/ for more information. Usage: Setup.hs COMMAND [FLAGS] or: Setup.hs [GLOBAL FLAGS] Global flags: -h --helpShow this help text -V --version Print version information --numeric-version Print just the version number Commands: configure Prepare to build the package. build Make this package ready for installation. install Copy the files into the install locations. Run register. copy Copy the files into the install locations. haddock Generate Haddock HTML documentation. clean Clean up after a build. sdist Generate a source distribution file (.tar.gz). hscolour Generate HsColour colourised code, in HTML format. register Register this package with the compiler. unregisterUnregister this package with the compiler. test Run the test suite, if any (configure with UserHooks). bench Run the benchmark, if any (configure with UserHooks). help Help about commands For more information about a command use Setup.hs COMMAND --help Typical steps for installing Cabal packages: Setup.hs configure Setup.hs build Setup.hs install On Mon, 26 Nov 2012 18:21:33 -0500 Albert Y. C. Lai tre...@vex.net wrote: cabal configure cabal build cabal register --inplace (newer) cabal build registers inplace automatically. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal configure cabal build cabal install
On 12-11-27 01:02 AM, kudah wrote: On Mon, 26 Nov 2012 18:21:33 -0500 Albert Y. C. Lai tre...@vex.net wrote: Lastly, there is no Setup install. Use copy and register. $ runghc Setup.hs --help [...] install Copy the files into the install locations. Run register. copy Copy the files into the install locations. I stand corrected, thank you. But then to complete answering a previous question, Setup install does not configure, does not build, and is far from cabal install. (newer) cabal build registers inplace automatically. I see where you heard this, but it is a misrepresentation from cabal, and older cabal has always told a similar misrepresentation. When cabal build succeeds, it always says: (older) registering name-version (newer) In-place registering name-version That's what it says. But use ghc-pkg and other tests to verify that no registration whatsoever has happened. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal configure cabal build cabal install
On Sun, Nov 25, 2012 at 06:09:26PM -0500, Albert Y. C. Lai wrote: If you begin with cabal configure, the correct idiom is: cabal configure [flags] cabal build [cabal haddock, if you want] cabal copy cabal register Even this does not do the same thing as 'cabal install', because it does not download and install any dependencies (whereas 'cabal install' does). -Brent ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal isn't updating local doc index after local package upgrade
This is filed as https://github.com/haskell/cabal/issues/1051 * Richard Cobbe co...@ccs.neu.edu [2012-11-24 12:43:55-0500] Haskell Platform 2012 v2.0.0, MacOS 64-bit. (MacOS 10.8.2.) I just used cabal to upgrade the installation of a local package I'm writing, and I'm still seeing the old version of the documentation in ~/Library/Haskell/doc/index.html. How can I fix this? In more detail: this machine had greek-1.0.1 installed, but I was working on another package that needed greek-1.1.0. So I went into the directory where I keep the source for the greek package, made sure it was up-to-date, and ran cabal clean cabal configure cabal build cabal install as my normal user, not as root. As far as I can tell, the rebuild was successful; cabal's output finished with Installing library in /Users/cobbe/Library/Haskell/ghc-7.4.1/lib/greek-1.1.0/lib Registering greek-1.1.0... Updating documentation index /Users/cobbe/Library/Haskell/doc/index.html However, when I open ~/Library/Haskell/doc/index.html in my browser, I still see the documentation for greek-1.0. The docs for the new version are present, in ~/Library/Haskell/ghc-7.4.1/lib/greek-1.1.0/doc, but they don't appear in the main index. Am I missing a step in the process, or is this a bug in cabal? Is there a workaround? I do have a couple of older versions of the greek package insatlled, because there doesn't seem to be an easy way to remove obsolete packages. Could they be causing problems? If so, what's the best way to delete them? (I don't particularly mind having the older versions hanging around, as long as they're actually harmless.) Thanks much, Richard ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal isn't updating local doc index after local package upgrade
On Sat, Nov 24, 2012 at 08:37:31PM +0200, Roman Cheplyaka wrote: This is filed as https://github.com/haskell/cabal/issues/1051 Ah! Thanks for the pointer; I didn't know about that bug database. I'll watch that issue for further developments. * Richard Cobbe co...@ccs.neu.edu [2012-11-24 12:43:55-0500] Haskell Platform 2012 v2.0.0, MacOS 64-bit. (MacOS 10.8.2.) I just used cabal to upgrade the installation of a local package I'm writing, and I'm still seeing the old version of the documentation in ~/Library/Haskell/doc/index.html. How can I fix this? Richard ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
Personally, I successfully use Wine to build, ship and test for Windows. There are some pitfalls related to -optl-mwindows and encodings, but, if you launch your program with $LANG set to proper windows encoding like cp1251 and the std handles closed with 0- 1- 2-, it should crash on related errors the same way as on windows. I am not (yet) aware of any Haskell programs that don't run under Wine. On Wed, 21 Nov 2012 13:05:45 +1100 Erik de Castro Lopo mle...@mega-nerd.com wrote: So is it difficult for an open source contributor to test on windows? Hell yes! You have no idea how hard windows is in comparison to say FreeBSD. Even Apple's OS X is easier than windows, because I have friends who can give me SSH access to their machines. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
kudah wrote: Personally, I successfully use Wine to build, ship and test for Windows. There are some pitfalls related to -optl-mwindows and encodings, but, if you launch your program with $LANG set to proper windows encoding like cp1251 and the std handles closed with 0- 1- 2-, it should crash on related errors the same way as on windows. I am not (yet) aware of any Haskell programs that don't run under Wine. Thats a very interesting solution. I use Wine to run the test suite when I cross compile one of my C projects from Linux to Wine. Would you consider documenting the process of setting everything up to build Haskell programs under Wine on the Haskell Wiki? Erik -- -- Erik de Castro Lopo http://www.mega-nerd.com/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
On Sat, 24 Nov 2012 13:46:37 +1100 Erik de Castro Lopo mle...@mega-nerd.com wrote: kudah wrote: Personally, I successfully use Wine to build, ship and test for Windows. There are some pitfalls related to -optl-mwindows and encodings, but, if you launch your program with $LANG set to proper windows encoding like cp1251 and the std handles closed with 0- 1- 2-, it should crash on related errors the same way as on 1windows. I am not (yet) aware of any Haskell programs that don't run under Wine. Thats a very interesting solution. I use Wine to run the test suite when I cross compile one of my C projects from Linux to Wine. Would you consider documenting the process of setting everything up to build Haskell programs under Wine on the Haskell Wiki? Erik Aside from what I posted above it's same as on Windows, just install Haskell Platform. There's already a page on Haskell Wiki http://www.haskell.org/haskellwiki/GHC_under_Wine though it seems very outdated. I can update it with my own observations when I get an HW account, they seem to have switched to manual registration while I wasn't looking. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
I just want to say that Windows support is much better than one could get the impression from this thread. I use Haskell on Windows as well as OSX and Linux. I think it works very well now, previously one had to know a bit of trickery to get things done. I don't think I have run into any more trouble on Windows than on the unixes, certainly there has been less headaches than with OSX (mostly GHC there), and you don't get the distro-hackage tension as on Linux. It is a bit annoying that packages depending on unix don't just quit upfront instead of installing a scad of dependencies first. But a mucked up package database that this thread is about can happen on any platform. The problem seems to be that cabal-install is so wonderfully easy to use that it obscures that there are no guarantees that things will just work and that it is often quite possible to fix it yourself by a tweak. If people think that what cabal install does behind the scenes is some advanced magic, it will not occur to them that they can do cabal unpack, fix the problem, and then cabal configure, cabal build and cabal install. Niklas 2012/11/21 Erik de Castro Lopo mle...@mega-nerd.com: Albert Y. C. Lai wrote: This counter-argument is flawed. Why limit oneself to one's own household? (Garage? Basement?) Get out more! Visit a friend. If that friend is not a coder, they are unlikely to have the dev tools installed. Talk to an internet cafe owner for a special deal to run one's own programs. Ditto. Rent virtual machine time in the cloud. I've already thrown a bunch of money at the microsoft machine for very poor results. If someone else set up and ran windows VMs and gave me access that would make testing on windows far more attractive. I just found that Amazon AWS has a free teir that includes windows as an option: https://aws.amazon.com/free/ Its still a huge sink of time and effort to set one up to a state where its ready to build haskell packages. Maybe if someone set up a github project that contained a script that could be downloaded onto a bare windows machine and then bootstrap that machine into a full haskell dev machine you might see some progress on this front. Erik -- -- Erik de Castro Lopo http://www.mega-nerd.com/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
On Tue, Nov 20, 2012 at 7:34 PM, Albert Y. C. Lai tre...@vex.net wrote: On 12-11-20 08:20 PM, Johan Tibell wrote: This logic is flawed. More than 90% of computers having Windows doesn't imply that 90% of all computers in a given household runs Windows. What's the probability that your household has a Windows computer if you're a programmer that don't live with your parents? What if that programmer is an open source contributor. Surely not 90%. This counter-argument is flawed. Why limit oneself to one's own household? (Garage? Basement?) Get out more! Visit a friend. Talk to an internet cafe owner for a special deal to run one's own programs. Rent virtual machine time in the cloud. There are many creative, flexible, low-cost possibilities. The key word here is low-cost. None of them are as low as the cost of Linux, Solaris, *BSD, etc. Those are all free. There's even free VM software available for them so you don't have to dedicate a machine to it. This actually makes the argument running in the other direction more telling. It's less expensive for Windows users to get Unix/Linux than Unix/Linux users to get Windows. If you want a Haskell environment to work in, install VirtualBoxOSE (free) and a Linux distro (also free) and work on that. Of course, the real cost is that maintaining software that you aren't using on a regular basis - which includes software you do use on a platform you don't - is a PITA. Given that, why would anyone doing something for free want to spend money for (access to a) copy of Windows to build/test software they aren't going to use? Insert standard OSS rant about do it yourself here. mike ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install...
The latest version of cabal-dev on Hackage does not seem to have had its dependencies updated for GHC 7.6. Try installing off github (https://github.com/creswick/cabal-dev). Ian Sturdy From: haskell-cafe-boun...@haskell.org [haskell-cafe-boun...@haskell.org] on behalf of Eric Velten de Melo [ericvm...@gmail.com] Sent: Tuesday, November 20, 2012 7:54 PM To: Johan Tibell Cc: Gregory Guthrie; haskell-cafe@haskell.org Subject: Re: [Haskell-cafe] cabal install... I have a dream of one day being able to install leksah without having to downgrade ghc. Right now I can't even install cabal-dev with cabal. It will break ghc if I do. 2012/11/20 Johan Tibell johan.tib...@gmail.com: On Tue, Nov 20, 2012 at 1:10 PM, Gregory Guthrie guth...@mum.edu wrote: Hmm, Now when I tried to run Leksah, I get not only some broken packages (which I can avoid for my current project), but: command line: cannot satisfy -package-id base-4.5.1.0-7c83b96f47f23db63c42a56351dcb917: base-4.5.1.0-7c83b96f47f23db63c42a56351dcb917 is unusable due to missing or recursive dependencies: integer-gmp-0.4.0.0-c15e185526893c3119f809251aac8c5b (use -v for more information) So I tried to install base, then re-install it, but both fail; Any hints? From this email and some of the previous emails it seems that your package DB is in a pretty bad state, most likely from using --force-reinstalls. When Cabal warns you that this will break stuff it actually means it. :) My suggestion is that you rm -rf ~/.ghc/x86_64-linux-7.6.1 # or equivalent on your system. Then reinstall all the packages you want by listing them all at once cabal install pkg1 pkg2 pk3 By listing them all together cabal-install tries to come up with an install plan that is globally consistent for all of them. -- Johan ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install... Trying to recover
Thanks for the suggestion, I’ll do that. Here goes: I deleted the ../user/appdata/roaming/ghc and ../cabal files, an uninstalled Haskell-platform. (No trace of anything ghc on the disk.) Then reinstalled Haskell, and ran “cabal update”, it said there was a new cabal-install, but trying to install it fails (below), so I went ahead with the current version. The error seems odd to me (cabal-install-1.16.0.2 depends on Cabal-1.16.0.3 which failed to install.), that an older version depends on a newer one? So now I have; (from Windows - Haskell-platform 2012.4.0.0) GHCi = The Glorious Glasgow Haskell Compilation System, version 7.4.2 Cabal = cabal-install version 0.14.0, using version 1.14.0 of the Cabal library I then tried to reload all my previous packages, (all at once?!), but it fails, out of memory (w/8GB of memory!) So I split it into sections, and tried the first one; it lists a lot of new installs, and then fails (full list at http://pastebin.com/5ywdUjgX) The first chunk of installs gives this: ... cabal: The following packages are likely to be broken by the reinstalls: QuickCheck-2.4.2 haskell-platform-2012.4.0.0 Use --force-reinstalls if you want to install anyway. I don't understand how it can want to break the Haskell-platform, sounds dangerous! And the second this: G:\Cabalcabal install Boolean Craft3e Craft3e GLFW GLURaw GLUT HTTP IORefCAS Me moTrie MonadCatchIO-mtl NumInstances ObjectName OpenGL OpenGLRaw QuickCheck SDL SHA StateVar Tensor abstract-deque abstract-par active aeson alex ansi-terminal array asn1-data attoparsec attoparsec-conduit base-unicode-symbols base64-bytest ring bits-atomic blaze-builder blaze-builder-conduit blaze-html blaze-markup bla ze-svg bmp buildwrapper byteorder cabal-dev case-insensitive cereal certificate clientsession cmdargs colour comonad conduit contravariant cookie cpphs cprng-ae s cpu criterion crypto-api crypto-conduit crypto-pubkey-types cryptocipher crypt ohash css-text data-default date-cache diagrams-core diagrams-lib diagrams-svg d list email-validate entropy erf failure fast-logger file-embed filepath filesyst em-conduit ghc-paths gloss gtk2hs-buildtools Resolving dependencies... In order, the following would be installed: Boolean-0.1.1 (new package) ... cabal: The following packages are likely to be broken by the reinstalls: regex-posix-0.95.1 regex-compat-0.95.1 regex-posix-0.94.4 regex-compat-0.93.1 parsec-3.1.1 fgl-5.4.2.4 fgl-5.4.2.3 QuickCheck-2.4.0.1 network-2.3.1.0 haskell-platform-2012.4.0.0 cgi-3001.1.7.4 HTTP-4000.2.5 regex-posix-0.95.2 regex-compat-0.95.1 regex-posix-0.95.1 regex-compat-0.95.1 Use --force-reinstalls if you want to install anyway. So I have a typical situation where it won't install, and gives an option to –force, but that seems to lead to more problems? Do I just have some packages which are intrinsically incompatible, and I have to choose between them? Not sure how to proceed. Any help or hints appreciated! :-) –––- Cabal install cabal-install Configuring Cabal-1.16.0.3... Warning: This package indirectly depends on multiple versions of the same package. This is highly likely to cause a compile failure. package process-1.1.0.1 requires base-4.5.0.0 package pretty-1.1.1.0 requires base-4.5.0.0 package old-time-1.1.0.0 requires base-4.5.0.0 package old-locale-1.0.0.4 requires base-4.5.0.0 package filepath-1.3.0.0 requires base-4.5.0.0 package directory-1.1.0.2 requires base-4.5.0.0 package deepseq-1.3.0.0 requires base-4.5.0.0 package containers-0.4.2.1 requires base-4.5.0.0 package bytestring-0.9.2.1 requires base-4.5.0.0 package array-0.4.0.0 requires base-4.5.0.0 package Win32-2.2.2.0 requires base-4.5.0.0 package filepath-1.3.0.0 requires base-4.5.1.0 package Cabal-1.16.0.3 requires base-4.5.1.0 package Cabal-1.16.0.3 requires filepath-1.3.0.0 package process-1.1.0.1 requires filepath-1.3.0.0 package directory-1.1.0.2 requires filepath-1.3.0.0 package integer-gmp-0.4.0.0 requires ghc-prim-0.2.0.0 package bytestring-0.9.2.1 requires ghc-prim-0.2.0.0 package base-4.5.0.0 requires ghc-prim-0.2.0.0 package integer-gmp-0.4.0.0 requires ghc-prim-0.2.0.0 package base-4.5.1.0 requires ghc-prim-0.2.0.0 package base-4.5.1.0 requires integer-gmp-0.4.0.0 package base-4.5.0.0 requires integer-gmp-0.4.0.0 Building Cabal-1.16.0.3... Preprocessing library Cabal-1.16.0.3... command line: cannot satisfy -package-id array-0.4.0.0-3cf1bc3f5cd0078adea24752c18081b9 (use -v for more information) cabal: Error: some packages failed to install: Cabal-1.16.0.3 failed during the building phase. The exception was: ExitFailure 1 cabal-install-1.16.0.2 depends on Cabal-1.16.0.3 which failed to install. (more -v details at: http://pastebin.com/Y2BuMjBP )
Re: [Haskell-cafe] cabal install... Trying to recover
You should have a ghc directory under appdata, with i386-mingw32-7.4.2\package.conf.d under it. There GHC tracks what packages it knows about. Niklas From: Gregory Guthrie Sent: 2012-11-21 15:11 To: Johan Tibell Cc: haskell-cafe@haskell.org Subject: Re: [Haskell-cafe] cabal install... Trying to recover Thanks for the suggestion, I’ll do that. Here goes: I deleted the ../user/appdata/roaming/ghc and ../cabal files, an uninstalled Haskell-platform. (No trace of anything ghc on the disk.) Then reinstalled Haskell, and ran “cabal update”, it said there was a new cabal-install, but trying to install it fails (below), so I went ahead with the current version. The error seems odd to me (cabal-install-1.16.0.2 depends on Cabal-1.16.0.3 which failed to install.), that an older version depends on a newer one? So now I have; (from Windows - Haskell-platform 2012.4.0.0) GHCi = The Glorious Glasgow Haskell Compilation System, version 7.4.2 Cabal = cabal-install version 0.14.0, using version 1.14.0 of the Cabal library I then tried to reload all my previous packages, (all at once?!), but it fails, out of memory (w/8GB of memory!) So I split it into sections, and tried the first one; it lists a lot of new installs, and then fails (full list at http://pastebin.com/5ywdUjgX) The first chunk of installs gives this: ... cabal: The following packages are likely to be broken by the reinstalls: QuickCheck-2.4.2 haskell-platform-2012.4.0.0 Use --force-reinstalls if you want to install anyway. I don't understand how it can want to break the Haskell-platform, sounds dangerous! And the second this: G:\Cabalcabal install Boolean Craft3e Craft3e GLFW GLURaw GLUT HTTP IORefCAS Me moTrie MonadCatchIO-mtl NumInstances ObjectName OpenGL OpenGLRaw QuickCheck SDL SHA StateVar Tensor abstract-deque abstract-par active aeson alex ansi-terminal array asn1-data attoparsec attoparsec-conduit base-unicode-symbols base64-bytest ring bits-atomic blaze-builder blaze-builder-conduit blaze-html blaze-markup bla ze-svg bmp buildwrapper byteorder cabal-dev case-insensitive cereal certificate clientsession cmdargs colour comonad conduit contravariant cookie cpphs cprng-ae s cpu criterion crypto-api crypto-conduit crypto-pubkey-types cryptocipher crypt ohash css-text data-default date-cache diagrams-core diagrams-lib diagrams-svg d list email-validate entropy erf failure fast-logger file-embed filepath filesyst em-conduit ghc-paths gloss gtk2hs-buildtools Resolving dependencies... In order, the following would be installed: Boolean-0.1.1 (new package) ... cabal: The following packages are likely to be broken by the reinstalls: regex-posix-0.95.1 regex-compat-0.95.1 regex-posix-0.94.4 regex-compat-0.93.1 parsec-3.1.1 fgl-5.4.2.4 fgl-5.4.2.3 QuickCheck-2.4.0.1 network-2.3.1.0 haskell-platform-2012.4.0.0 cgi-3001.1.7.4 HTTP-4000.2.5 regex-posix-0.95.2 regex-compat-0.95.1 regex-posix-0.95.1 regex-compat-0.95.1 Use --force-reinstalls if you want to install anyway. So I have a typical situation where it won't install, and gives an option to –force, but that seems to lead to more problems? Do I just have some packages which are intrinsically incompatible, and I have to choose between them? Not sure how to proceed. Any help or hints appreciated! :-) –––- Cabal install cabal-install Configuring Cabal-1.16.0.3... Warning: This package indirectly depends on multiple versions of the same package. This is highly likely to cause a compile failure. package process-1.1.0.1 requires base-4.5.0.0 package pretty-1.1.1.0 requires base-4.5.0.0 package old-time-1.1.0.0 requires base-4.5.0.0 package old-locale-1.0.0.4 requires base-4.5.0.0 package filepath-1.3.0.0 requires base-4.5.0.0 package directory-1.1.0.2 requires base-4.5.0.0 package deepseq-1.3.0.0 requires base-4.5.0.0 package containers-0.4.2.1 requires base-4.5.0.0 package bytestring-0.9.2.1 requires base-4.5.0.0 package array-0.4.0.0 requires base-4.5.0.0 package Win32-2.2.2.0 requires base-4.5.0.0 package filepath-1.3.0.0 requires base-4.5.1.0 package Cabal-1.16.0.3 requires base-4.5.1.0 package Cabal-1.16.0.3 requires filepath-1.3.0.0 package process-1.1.0.1 requires filepath-1.3.0.0 package directory-1.1.0.2 requires filepath-1.3.0.0 package integer-gmp-0.4.0.0 requires ghc-prim-0.2.0.0 package bytestring-0.9.2.1 requires ghc-prim-0.2.0.0 package base-4.5.0.0 requires ghc-prim-0.2.0.0 package integer-gmp-0.4.0.0 requires ghc-prim-0.2.0.0 package base-4.5.1.0 requires ghc-prim-0.2.0.0 package base-4.5.1.0 requires integer-gmp-0.4.0.0 package base-4.5.0.0 requires integer-gmp-0.4.0.0 Building Cabal-1.16.0.3... Preprocessing library Cabal-1.16.0.3... command line: cannot satisfy -package-id array-0.4.0.0-3cf1bc3f5cd0078adea24752c18081b9 (use -v
Re: [Haskell-cafe] cabal install... Trying to recover
On Wed, Nov 21, 2012 at 9:08 AM, Gregory Guthrie guth...@mum.edu wrote: The error seems odd to me (cabal-install-1.16.0.2 depends on Cabal-1.16.0.3 which failed to install.), that an older version depends on a newer one? There was a minor bug in the Cabal library necessitating a point release. cabal-install was unaffected; why make a new release just to have pretty versioning? I then tried to reload all my previous packages, (all at once?!), but it fails, out of memory (w/8GB of memory!) i386 or x86-64? 8GB isn't really 8GB on the former. So I split it into sections, and tried the first one; it lists a lot of new installs, and then fails (full list at http://pastebin.com/5ywdUjgX) You're explicitly asking it for a new version of HTTP, which is asking for trouble. More worrisome is that it's still asking for new versions of base, which means it's still confused about what version of ghc is installed. -- brandon s allbery kf8nh sine nomine associates allber...@gmail.com ballb...@sinenomine.net unix/linux, openafs, kerberos, infrastructure http://sinenomine.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install... Trying to recover
OK; I took HTTP out, but still get the same error; cabal: The following packages are likely to be broken by the reinstalls: QuickCheck-2.4.2 haskell-platform-2012.4.0.0 Use --force-reinstalls if you want to install anyway. One thing I notice; Ghc reports: G:\Cabalghc --version The Glorious Glasgow Haskell Compilation System, version 7.4.2 But I did notice that I had an environment variable (from some previous install, I think openCV?) of: GHC_VERSION=7.4.1 So I updated that to 7.4.2 And retried, same results. But then, ghc-pkg check reports: The following packages are broken, either because they have a problem listed above, or because they depend on a broken package. HTTP-4000.2.3 haskell-platform-2012.2.0.0 I have no idea where the 2012.2 comes from. Any suggestions? --- From: Brandon Allbery [mailto:allber...@gmail.com] Subject: Re: [Haskell-cafe] cabal install... Trying to recover So I split it into sections, and tried the first one; it lists a lot of new installs, and then fails (full list at http://pastebin.com/5ywdUjgX) You're explicitly asking it for a new version of HTTP, which is asking for trouble. More worrisome is that it's still asking for new versions of base, which means it's still confused about what version of ghc is installed. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install... Trying to recover
On Wed, Nov 21, 2012 at 12:06 PM, Gregory Guthrie guth...@mum.edu wrote: OK; I took HTTP out, but still get the same error; cabal: The following packages are likely to be broken by the reinstalls: QuickCheck-2.4.2 haskell-platform-2012.4.0.0 Use --force-reinstalls if you want to install anyway. Right, that was not intended to be a complete fix or anything, just a note. This is the important part, and what I noted immediately afterward --- did you happen to notice there was anything in the message after that first part? (Although I'm not asking this first so it also may not actually exist, I guess) One thing I notice; Ghc reports: G:\Cabalghc --version The Glorious Glasgow Haskell Compilation System, version 7.4.2 ** ** But I did notice that I had an environment variable (from some previous install, I think openCV?) of: GHC_VERSION=7.4.1 So I updated that to 7.4.2 And retried, same results. There is more going on than just that environment variable; this is what the base stuff that you have been ignoring is trying to tell you. You still have both compilers installed, and their packages are somehow jumbled together. This is breaking your installation. Unfortunately, as I am neither particularly familiar with Windows nor able to access your system (which is probably for the best for both of us), I can't really help you with figuring out why you have two GHC versions' packages mixed together. But as long as you do, cabal will be trying to upgrade the base package, which is the actual source of the breakage. -- brandon s allbery kf8nh sine nomine associates allber...@gmail.com ballb...@sinenomine.net unix/linux, openafs, kerberos, infrastructure http://sinenomine.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install... Trying to recover
Thanks. I’ll try to do another cleanup, but not sure what more I can uninstall or clean out! I did a system search for *ghc* and came up empty before reinstall; will try again. I have now managed to get from some broken packages to a broken system! ☺ --- Subject: Re: [Haskell-cafe] cabal install... Trying to recover This is the important part, and what I noted immediately afterward --- did you happen to notice there was anything in the message after that first part? (Although I'm not asking this first so it also may not actually exist, I guess) One thing I notice; Ghc reports: G:\Cabalghc --version The Glorious Glasgow Haskell Compilation System, version 7.4.2 There is more going on than just that environment variable; this is what the base stuff that you have been ignoring is trying to tell you. You still have both compilers installed, and their packages are somehow jumbled together. This is breaking your installation. Unfortunately, as I am neither particularly familiar with Windows nor able to access your system (which is probably for the best for both of us), I can't really help you with figuring out why you have two GHC versions' packages mixed together. But as long as you do, cabal will be trying to upgrade the base package, which is the actual source of the breakage. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
Let's put some numbers on this. (1) In this country, you can buy a second-hand dual core desktop for NZD 200 (roughly USD 165, EUR 130). You can buy a new laptop for NZD 400 (roughly USD 330, EUR 260). Not fancy machines, but more than adequate to compile and build stuff. Shipping adds a fair bit to prices here. So it _must_ be possible to buy a Windows box of some kind adequate for compiling, building, and testing open source software, for even less than that in North America or Europe. It's really *NOT* the price of the box-with-Windows-installed. (2) This department has a mix of Mac OS X, Linux (running on Apple dual-boot boxes), and Windows (running on Apple dual-boot boxes). The University has quite a few Windows labs. There would be _no_ students at this University who did not have ready access to a Windows machine whenever they wanted one. The servers in the department all run some flavour of UNIX, true. (3) Given an intel Solaris, intel Linux, or intel Mac OS X box, VirtualBox is free. You can run Windows in VirtualBox. Microsoft offer a full Windows 7 Professional licence to University students for USD 30. So I really don't buy the idea of a student finding it hard to get Windows. My University is part of the MSDN Academic Alliance, so staff get stuff for no money of their own. Windows 7 Home Premium is USD 200, Professional USD 300. Probably better to buy a cheap box that already has Windows. What about software? Well, Microsoft Visual Studio Professional 2012 is several times more expensive than the box it runs on, and Office is not cheap either. There are, as always, special deals, e.g., https://www.dreamspark.com/Product/Product.aspx?productid=34 seems to make VC++ 2008 available free to students, and the MSDN Academic Alliance makes this stuff easy for staff to get. For everyone else, Eclipse and NetBeans are free, and so are Cygwin and Mingw. It took me about a day to download and install a large amount of free software, giving me quite a decent environment. (Of course, if someone were paying me to do this, the University would charge NZD 150/hour, so free = NZD 1200 ...) I even had Microsoft SUA (Services for Unix Applications -- think of it as Cygwin from Microsoft but with a less horribly ugly terminal font). I had ghc and OCaml and SWI Prolog and Squeak and Dolphin Smalltalk and lots of good stuff. So it's not really the availability of software either. So am I a happy Windows hacker? Actually, no. I had a working tolerable setup under Windows Vista. Despite its bad press, I have to say I never had any trouble with Vista. Then my (the department's) Mac laptop needed something done to it -- I forget what -- and they said while we're at it, it would simplify our lives if we upgraded the Windows side to Windows 7 like everyone else has now. I said, OK, but I _really_ don't want to lose any of my programs. And they lost everything beginning with the letters M-Z, and what they didn't lose stopped working. Apparently when Windows went 64 bit they didn't leave \Program Files\ alone and add a \Program Files 64\ directory. Oh no! Now \Program Files\ was exclusively for 64-bit programs, and 32-bit ones were supposed to be in \Program Files (x86)\. You can guess what that did to the surviving remnants of my environment. How long did it take to rebuild my environment? I don't know. Except for installing Cygwin I haven't done it. The changes to the user interface -- apparently just for the sake of change, because absolutely nothing I do has become easier for me -- did nothing for my facility with the system, and having to spend half an hour installing updates every time I boot into Windows doesn't increase my enjoyment. I don't want to even _think_ about Windows 8. On 21/11/2012, at 3:21 PM, Clark Gaebel wrote: +1 to this. The friction of finding, setting up, and using Windows isn't even comparable to just sshing into another unix box and testing something quickly. As a university student, I also find it relatively rare that I get to test on a Windows machine. My personal computer runs linux, my technical friends run linux or osx, and my non-technical ones run osx. Also, all the school servers that I have access to run either FreeBSD or Linux. If I want to run something on linux system, I have about 40 different computers that I can ssh into and run code on. If I want to run something on osx, I just have to call a friend and ask if they can turn on their computer and allow me to ssh in (to my own account, of course). If I want to run something on Windows, I have to track down a friend (in person!), ask to borrow their computer for a few hours, get administrator access to install the Haskell Platform, get frustrated that HP hasn't been upgraded to 7.6, and give up. It's just not practical, especially for the large amount of small (500 LOC)
Re: [Haskell-cafe] Cabal failures...
Hi Johan.
I haven't looked in detail at the overall problem, but:
Flags chosen: base3=True, base4=True
Why is Cabal setting both base3 and base4 to True?
This looks completely fine to me.
The Cabal .cabal file is stating:
if flag(base4) { build-depends: base = 4 } else { build-depends: base 4 }
if flag(base3) { build-depends: base = 3 } else { build-depends: base 3 }
So it's relatively clear to me that both have to be true.
Cheers,
Andres
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
Thanks to all for the comparisons between apt cabal. Your reply basically explains why it is broken, and gives a rationale (cost and trouble to do it), but no prognosis for repair. My interest is in using Haskell for teaching, and so far the package system failures often present problems that I can't solve for all but the simplest examples, so I couldn't much pass this onto students! My libraries may have gotten corrupted, so when I get time I will try to reset and clean out everything and start over, but that will of course break a lot of things and take a lot of time for some re-installs, most particularly things depending on underlying C libraries. Any hints for a simple way to do this are welcome. I look forward to the outcomes from the current cabal discussions. It was also interesting to note a comment that most developers don't have access to a Windows machine for testing. With Windows at 90% of the computing market (Linux = 1.6%), this seems like a problem which might limit growth of Haskell usage. Just an observation. :-) Thanks for your article on the topic of Cabal - interesting! -- Date: Tue, 20 Nov 2012 01:03:20 -0500 From: Albert Y. C. Lai tre...@vex.net Subject: Re: [Haskell-cafe] Cabal failures... At least I paid my 3 hours to explain some cabal stuff at http://www.vex.net/~trebla/haskell/sicp.xhtml -- ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
Thanks to all for the comparisons between apt cabal. Your reply basically explains why it is broken, and gives a rationale (cost and trouble to do it), but no prognosis for repair. It's an open problem. I make do with disposable sand-boxes, using cabal-dev to build them. In this way, I can keep many (of my) libraries building, even if different libraries need incompatible versions of Hackage libraries. If I run into a problem in a library's build environment, I either fix it surgically or nuke it, without affecting any other build environments. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
On 12-11-20 08:48 AM, Gregory Guthrie wrote: It was also interesting to note a comment that most developers don't have access to a Windows machine for testing. With Windows at 90% of the computing market (Linux = 1.6%), this seems like a problem which might limit growth of Haskell usage. Just an observation. :-) There is a paradox in that sentence. The first sentence says, most developers don't have access to Windows machines for testing. But they have access to Linux machines. Then Windows machines must be a scarcity compared to Linux machines, no? So scarce, you even have difficulty borrowing or renting. Then the next sentence says, the scarcity is the other way round, Linux machines are scarce, Windows machines are abundant. OK, so why is it so hard to access something abundant, and so easy to access something scarce? ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install...
On Tue, Nov 20, 2012 at 4:10 PM, Gregory Guthrie guth...@mum.edu wrote: Hmm, Now when I tried to run Leksah, I get not only some broken packages (which I can avoid for my current project), but: ** ** command line: cannot satisfy -package-id base-4.5.1.0-7c83b96f47f23db63c42a56351dcb917: base-4.5.1.0-7c83b96f47f23db63c42a56351dcb917 is unusable due to missing or recursive dependencies: integer-gmp-0.4.0.0-c15e185526893c3119f809251aac8c5b (use -v for more information) ** ** So I tried to install base, then re-install it, but both fail; You can't install base or integer-gmp from cabal-install. They are wired into the compiler, and the only way to reinstall them is to reinstall ghc. In fact, finding a way to install ether from cabal-install will cause the kind of breakage you're seeing. (It's not supposed to be possible, at least for base. If at some point you installed integer-gmp from hackage, you need to remove it; if you installed it into the global package database, you really do have no choice but remove and reinstall ghc now.) If you installed ghc as part of the haskell platform, then you need to remove and reinstall that. -- brandon s allbery kf8nh sine nomine associates allber...@gmail.com ballb...@sinenomine.net unix/linux, openafs, kerberos, infrastructure http://sinenomine.net ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
No; the first sentence says that someone else had reported that testing on Windows was hard to do because of (a perceived) lack of access to Windows by Haskell developers... The implication is that Haskell developers (only/mainly) use *nix. I commented that if true this lack of Windows testing could limit the availability of Haskell to the largest market share of users. --- Subject: Re: [Haskell-cafe] Cabal failures... To: haskell-cafe@haskell.org On 12-11-20 08:48 AM, Gregory Guthrie wrote: It was also interesting to note a comment that most developers don't have access to a Windows machine for testing. With Windows at 90% of the computing market (Linux = 1.6%), this seems like a problem which might limit growth of Haskell usage. Just an observation. :-) There is a paradox in that sentence. The first sentence says, most developers don't have access to Windows machines for testing. But they have access to Linux machines. Then Windows machines must be a scarcity compared to Linux machines, no? So scarce, you even have difficulty borrowing or renting. Then the next sentence says, the scarcity is the other way round, Linux machines are scarce, Windows machines are abundant. OK, so why is it so hard to access something abundant, and so easy to access something scarce? -- ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install...
On Tue, Nov 20, 2012 at 1:10 PM, Gregory Guthrie guth...@mum.edu wrote: Hmm, Now when I tried to run Leksah, I get not only some broken packages (which I can avoid for my current project), but: ** ** command line: cannot satisfy -package-id base-4.5.1.0-7c83b96f47f23db63c42a56351dcb917: base-4.5.1.0-7c83b96f47f23db63c42a56351dcb917 is unusable due to missing or recursive dependencies: integer-gmp-0.4.0.0-c15e185526893c3119f809251aac8c5b (use -v for more information) ** ** So I tried to install base, then re-install it, but both fail; Any hints? From this email and some of the previous emails it seems that your package DB is in a pretty bad state, most likely from using --force-reinstalls. When Cabal warns you that this will break stuff it actually means it. :) My suggestion is that you rm -rf ~/.ghc/x86_64-linux-7.6.1 # or equivalent on your system. Then reinstall all the packages you want by listing them all at once cabal install pkg1 pkg2 pk3 By listing them all together cabal-install tries to come up with an install plan that is globally consistent for all of them. -- Johan ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal install...
I have a dream of one day being able to install leksah without having to downgrade ghc. Right now I can't even install cabal-dev with cabal. It will break ghc if I do. 2012/11/20 Johan Tibell johan.tib...@gmail.com: On Tue, Nov 20, 2012 at 1:10 PM, Gregory Guthrie guth...@mum.edu wrote: Hmm, Now when I tried to run Leksah, I get not only some broken packages (which I can avoid for my current project), but: command line: cannot satisfy -package-id base-4.5.1.0-7c83b96f47f23db63c42a56351dcb917: base-4.5.1.0-7c83b96f47f23db63c42a56351dcb917 is unusable due to missing or recursive dependencies: integer-gmp-0.4.0.0-c15e185526893c3119f809251aac8c5b (use -v for more information) So I tried to install base, then re-install it, but both fail; Any hints? From this email and some of the previous emails it seems that your package DB is in a pretty bad state, most likely from using --force-reinstalls. When Cabal warns you that this will break stuff it actually means it. :) My suggestion is that you rm -rf ~/.ghc/x86_64-linux-7.6.1 # or equivalent on your system. Then reinstall all the packages you want by listing them all at once cabal install pkg1 pkg2 pk3 By listing them all together cabal-install tries to come up with an install plan that is globally consistent for all of them. -- Johan ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
On 12-11-20 05:37 PM, Gregory Guthrie wrote: No; the first sentence says that someone else had reported that testing on Windows was hard to do because of (a perceived) lack of access to Windows by Haskell developers... The implication is that Haskell developers (only/mainly) use *nix. I commented that if true this lack of Windows testing could limit the availability of Haskell to the largest market share of users. Clearly, since 90% of computers have Windows, it should be trivial to find one to test on, if a programmer wants to. Surely every programmer is surrounded by Windows-using family and friends? (Perhaps to the programmer's dismay, too, because the perpetual I've got a virus again, can you help? is so annoying?) We are not talking about BeOS. Therefore, if programmers do not test on Windows, it is because they do not want to. And why would they want to? Take webapp programmers for example. 99.999...% [1] of computers have sufficiently new web browsers. This market share is even higher than Windows. At the server side, the programmers have freedom in choosing the OS, and apparently, they choose anything but Windows, and this has never limited them in accessing 99.999...% of computer users. And this, 99.999...% web browser market share, is exactly driving Haskell growth. Not the petty 90% Windows slice. [1] http://en.wikipedia.org/wiki/0.999... ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
On Tue, Nov 20, 2012 at 5:14 PM, Albert Y. C. Lai tre...@vex.net wrote: On 12-11-20 05:37 PM, Gregory Guthrie wrote: No; the first sentence says that someone else had reported that testing on Windows was hard to do because of (a perceived) lack of access to Windows by Haskell developers... The implication is that Haskell developers (only/mainly) use *nix. I commented that if true this lack of Windows testing could limit the availability of Haskell to the largest market share of users. Clearly, since 90% of computers have Windows, it should be trivial to find one to test on, if a programmer wants to. Surely every programmer is surrounded by Windows-using family and friends? (Perhaps to the programmer's dismay, too, because the perpetual I've got a virus again, can you help? is so annoying?) We are not talking about BeOS. Therefore, if programmers do not test on Windows, it is because they do not want to. This logic is flawed. More than 90% of computers having Windows doesn't imply that 90% of all computers in a given household runs Windows. What's the probability that your household has a Windows computer if you're a programmer that don't live with your parents? What if that programmer is an open source contributor. Surely not 90%. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
Why not? Either way, I am chiming in as a programmer of many years. Unless using osx I stick with windows to avoid half-day forays into nettling technical issues that are not related to the work I am paid to perform. I would love for Haskell to work better there. On Nov 20, 2012 5:21 PM, Johan Tibell johan.tib...@gmail.com wrote: On Tue, Nov 20, 2012 at 5:14 PM, Albert Y. C. Lai tre...@vex.net wrote: On 12-11-20 05:37 PM, Gregory Guthrie wrote: No; the first sentence says that someone else had reported that testing on Windows was hard to do because of (a perceived) lack of access to Windows by Haskell developers... The implication is that Haskell developers (only/mainly) use *nix. I commented that if true this lack of Windows testing could limit the availability of Haskell to the largest market share of users. Clearly, since 90% of computers have Windows, it should be trivial to find one to test on, if a programmer wants to. Surely every programmer is surrounded by Windows-using family and friends? (Perhaps to the programmer's dismay, too, because the perpetual I've got a virus again, can you help? is so annoying?) We are not talking about BeOS. Therefore, if programmers do not test on Windows, it is because they do not want to. This logic is flawed. More than 90% of computers having Windows doesn't imply that 90% of all computers in a given household runs Windows. What's the probability that your household has a Windows computer if you're a programmer that don't live with your parents? What if that programmer is an open source contributor. Surely not 90%. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
On 12-11-20 08:20 PM, Johan Tibell wrote: This logic is flawed. More than 90% of computers having Windows doesn't imply that 90% of all computers in a given household runs Windows. What's the probability that your household has a Windows computer if you're a programmer that don't live with your parents? What if that programmer is an open source contributor. Surely not 90%. This counter-argument is flawed. Why limit oneself to one's own household? (Garage? Basement?) Get out more! Visit a friend. Talk to an internet cafe owner for a special deal to run one's own programs. Rent virtual machine time in the cloud. There are many creative, flexible, low-cost possibilities. If one wants to. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
On Tue, Nov 20, 2012 at 5:34 PM, Albert Y. C. Lai tre...@vex.net wrote: This counter-argument is flawed. Why limit oneself to one's own household? (Garage? Basement?) Get out more! Visit a friend. Talk to an internet cafe owner for a special deal to run one's own programs. Rent virtual machine time in the cloud. There are many creative, flexible, low-cost possibilities. If one wants to. Clearly this different approaches have different costs. Fixing a bug from my couch or asking some stranger at a cafe if I can install msys is quite different things. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
Albert Y. C. Lai wrote: Clearly, since 90% of computers have Windows, it should be trivial to find one to test on, if a programmer wants to. Surely every programmer is surrounded by Windows-using family and friends? (Perhaps to the programmer's dismay, too, because the perpetual I've got a virus again, can you help? is so annoying?) We are not talking about BeOS. Therefore, if programmers do not test on Windows, it is because they do not want to. I have been an open source contributor for over 15 years. All the general purpose machines in my house run Linux. My father's and my mother-in-law's computers also run Linux (easier for me to provide support). For testing software, I have a PowerPC machine and virtual machines running various versions of Linux, FreeBSD and OpenBSD. What I don't have is a windows machine. I have, at numerous times, spent considerable amounts of time (and even real money for licenses) setting up (or rather trying to) windows in a VM and it is *always* considerably more work to set up, maintain and fix when something goes wrong. Setting up development tools is also a huge pain in the ass. And sooner or later they fail in some way I can't fix and I have to start again. Often its not worth the effort. At my day job we have on-demand windows VMs, but I am not officially allowed (nor do I intend to start) to use those resources for my open source work. So is it difficult for an open source contributor to test on windows? Hell yes! You have no idea how hard windows is in comparison to say FreeBSD. Even Apple's OS X is easier than windows, because I have friends who can give me SSH access to their machines. Erik -- -- Erik de Castro Lopo http://www.mega-nerd.com/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
Albert Y. C. Lai wrote: This counter-argument is flawed. Why limit oneself to one's own household? (Garage? Basement?) Get out more! Visit a friend. If that friend is not a coder, they are unlikely to have the dev tools installed. Talk to an internet cafe owner for a special deal to run one's own programs. Ditto. Rent virtual machine time in the cloud. I've already thrown a bunch of money at the microsoft machine for very poor results. If someone else set up and ran windows VMs and gave me access that would make testing on windows far more attractive. I just found that Amazon AWS has a free teir that includes windows as an option: https://aws.amazon.com/free/ Its still a huge sink of time and effort to set one up to a state where its ready to build haskell packages. Maybe if someone set up a github project that contained a script that could be downloaded onto a bare windows machine and then bootstrap that machine into a full haskell dev machine you might see some progress on this front. Erik -- -- Erik de Castro Lopo http://www.mega-nerd.com/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
+1 to this. The friction of finding, setting up, and using Windows isn't even comparable to just sshing into another unix box and testing something quickly. As a university student, I also find it relatively rare that I get to test on a Windows machine. My personal computer runs linux, my technical friends run linux or osx, and my non-technical ones run osx. Also, all the school servers that I have access to run either FreeBSD or Linux. If I want to run something on linux system, I have about 40 different computers that I can ssh into and run code on. If I want to run something on osx, I just have to call a friend and ask if they can turn on their computer and allow me to ssh in (to my own account, of course). If I want to run something on Windows, I have to track down a friend (in person!), ask to borrow their computer for a few hours, get administrator access to install the Haskell Platform, get frustrated that HP hasn't been upgraded to 7.6, and give up. It's just not practical, especially for the large amount of small (500 LOC) packages on Hackage. - Clark On Tue, Nov 20, 2012 at 9:05 PM, Erik de Castro Lopo mle...@mega-nerd.comwrote: Albert Y. C. Lai wrote: Clearly, since 90% of computers have Windows, it should be trivial to find one to test on, if a programmer wants to. Surely every programmer is surrounded by Windows-using family and friends? (Perhaps to the programmer's dismay, too, because the perpetual I've got a virus again, can you help? is so annoying?) We are not talking about BeOS. Therefore, if programmers do not test on Windows, it is because they do not want to. I have been an open source contributor for over 15 years. All the general purpose machines in my house run Linux. My father's and my mother-in-law's computers also run Linux (easier for me to provide support). For testing software, I have a PowerPC machine and virtual machines running various versions of Linux, FreeBSD and OpenBSD. What I don't have is a windows machine. I have, at numerous times, spent considerable amounts of time (and even real money for licenses) setting up (or rather trying to) windows in a VM and it is *always* considerably more work to set up, maintain and fix when something goes wrong. Setting up development tools is also a huge pain in the ass. And sooner or later they fail in some way I can't fix and I have to start again. Often its not worth the effort. At my day job we have on-demand windows VMs, but I am not officially allowed (nor do I intend to start) to use those resources for my open source work. So is it difficult for an open source contributor to test on windows? Hell yes! You have no idea how hard windows is in comparison to say FreeBSD. Even Apple's OS X is easier than windows, because I have friends who can give me SSH access to their machines. Erik -- -- Erik de Castro Lopo http://www.mega-nerd.com/ ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
Hi Greg, On Mon, Nov 19, 2012 at 1:25 PM, Gregory Guthrie guth...@mum.edu wrote: I follow the Cabal-messes threads with some interest, since that is the hardest area for me since starting to use Haskell. Probably 40-60% of all package install fail for some mysterious reason, with threats that trying to fix them will break more things, which generally is true. :-) We're working on it. Be brave, things are going to get better! I am not exert in the area, but I wonder how /why/ this is different than other package managers, like apt in Linux, I have never had any problems with it, and I would think that their dependencies are of at least similar complexities. The Linux package managers solve a different problems. They let you install a set of packages that have been manually curated and are know to work together (i.e. all version dependencies are fixed) while cabal does version resolution on packages that might not ever have been tried together. If you install Haskell packages via your distro's package manager I assume they will always install cleanly. The problem is that people want the latest bleeding edge of packages, which haven't made it into the distros yet, and hence they get to experience some of the pains associated with being on the bleeding edge. Being on Windows also makes things harder, as most developers don't have a Windows box to test their stuff on. In any case; Trying to do a cabal update I was told to try to update cabal-install, which I think means actually updating cabal (since I actually run installs via cabal install...), but that fails with this message below, and I don't know how to proceed. cabal-install is the package that includes the cabal executable. Cabal (with a capital C) is the library that cabal-install uses. The naming is unfortunate but hard to change at this point. To update cabal-install you do: $ cabal update cabal install cabal-install Make sure that the place that the cabal binary gets installed into (which is printed at the end of the install) is on your PATH. Linking C:\Users\guthrie\AppData\Local\Temp\Cabal-1.16.0.3-13880\Cabal-1.16.0.3\dist\setup\setup.exe ... Configuring Cabal-1.16.0.3... Warning: This package indirectly depends on multiple versions of the same package. This is highly likely to cause a compile failure. This is a sure sign that things are not going to work well. Could you include the output of cabal install -v cabal-install please. The output here is not enough to tell me what's going on. Please also include the output of cabal --version ghc --version Are you using the Haskell Platform, if so, which version? -- Johan ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
cabal install -v cabal-install Not sure if you're running into this one, but a configuration that wasn't working for me: 1) Install Haskell Platform 2) Install GHC 7.6.1 3) cabal install cabal-install As I recall, the error had something to do with a Cabal-generated 'Paths' file assuming the Prelude exported 'catch'. It was affecting a bunch of other packages too, which forced me to upgrade cabal-install. To get things working, I had to boot GHC 7.6 from my system PATH, upgrade cabal-install using GHC 7.4, and then put 7.6 back in the system path. After doing that, everything has worked well with GHC 7.6. -Greg On Mon, Nov 19, 2012 at 1:38 PM, Johan Tibell johan.tib...@gmail.com wrote: Hi Greg, On Mon, Nov 19, 2012 at 1:25 PM, Gregory Guthrie guth...@mum.edu wrote: I follow the Cabal-messes threads with some interest, since that is the hardest area for me since starting to use Haskell. Probably 40-60% of all package install fail for some mysterious reason, with threats that trying to fix them will break more things, which generally is true. :-) We're working on it. Be brave, things are going to get better! I am not exert in the area, but I wonder how /why/ this is different than other package managers, like apt in Linux, I have never had any problems with it, and I would think that their dependencies are of at least similar complexities. The Linux package managers solve a different problems. They let you install a set of packages that have been manually curated and are know to work together (i.e. all version dependencies are fixed) while cabal does version resolution on packages that might not ever have been tried together. If you install Haskell packages via your distro's package manager I assume they will always install cleanly. The problem is that people want the latest bleeding edge of packages, which haven't made it into the distros yet, and hence they get to experience some of the pains associated with being on the bleeding edge. Being on Windows also makes things harder, as most developers don't have a Windows box to test their stuff on. In any case; Trying to do a cabal update I was told to try to update cabal-install, which I think means actually updating cabal (since I actually run installs via cabal install...), but that fails with this message below, and I don't know how to proceed. cabal-install is the package that includes the cabal executable. Cabal (with a capital C) is the library that cabal-install uses. The naming is unfortunate but hard to change at this point. To update cabal-install you do: $ cabal update cabal install cabal-install Make sure that the place that the cabal binary gets installed into (which is printed at the end of the install) is on your PATH. Linking C:\Users\guthrie\AppData\Local\Temp\Cabal-1.16.0.3-13880\Cabal-1.16.0.3\dist\setup\setup.exe ... Configuring Cabal-1.16.0.3... Warning: This package indirectly depends on multiple versions of the same package. This is highly likely to cause a compile failure. This is a sure sign that things are not going to work well. Could you include the output of cabal install -v cabal-install please. The output here is not enough to tell me what's going on. Please also include the output of cabal --version ghc --version Are you using the Haskell Platform, if so, which version? -- Johan ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
On Mon, Nov 19, 2012 at 2:55 PM, Greg Fitzgerald gari...@gmail.com wrote: cabal install -v cabal-install Not sure if you're running into this one, but a configuration that wasn't working for me: 1) Install Haskell Platform 2) Install GHC 7.6.1 3) cabal install cabal-install As I recall, the error had something to do with a Cabal-generated 'Paths' file assuming the Prelude exported 'catch'. It was affecting a bunch of other packages too, which forced me to upgrade cabal-install. To get things working, I had to boot GHC 7.6 from my system PATH, upgrade cabal-install using GHC 7.4, and then put 7.6 back in the system path. After doing that, everything has worked well with GHC 7.6. The issue is that cabal-install-1.16.0.1 is broken on Windows. We have a new, fixed cabal-install-1.16.0.2 out, but if you were unlucky enough to install the broken one you need to delete that binary and install cabal-install again (either by using the bootstrap.sh script in the cabal-install repo or by some other means). ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
Thanks for looking at this and the help; Trying with topdown changes things, but as often is the case warns that it will break another ~60 packages if I force it, not sure if this will help me or cause the ruin of the rest of the local Haskell library universe. Should I force it?! :-) C:\Users\guthriecabal install -v --solver=topdown cabal-install Reading available packages... Resolving dependencies... In order, the following would be installed: Win32-2.2.2.0 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 array-0.4.0.0 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 deepseq-1.3.0.0 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 containers-0.4.2.1 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 old-locale-1.0.0.4 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 old-time-1.1.0.0 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 directory-1.1.0.2 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 pretty-1.1.1.0 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 process-1.1.0.1 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 Cabal-1.16.0.3 (new version) text-0.11.2.3 (reinstall) parsec-3.1.3 (reinstall) network-2.4.0.1 (new version) HTTP-4000.2.5 (reinstall) changes: network-2.3.1.0 - 2.4.0.1 time-1.4 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 random-1.0.1.1 (reinstall) changes: base-4.5.0.0 - 4.5.1.0 cabal-install-1.16.0.2 -bytestring-in-base (new package) cabal: The following packages are likely to be broken by the reinstalls: QuickCheck-2.4.2 haskell98-2.0.0.1 ghc-7.4.1 Cabal-1.14.0 bin-package-db-0.0.0.0 hpc-0.5.1.1 haskell-platform-2012.4.0.0 QuickCheck-2.5.1.1 haskell98-2.0.0.1 ghc-7.4.2 Cabal-1.14.0 bin-package-db-0.0.0.0 hpc-0.5.1.1 text-0.11.2.0 parsec-3.1.2 stm-2.3 regex-posix-0.95.1 regex-compat-0.95.1 regex-base-0.93.2 parallel-3.2.0.2 haskell2010-1.1.0.1 haskell-src-1.0.1.5 fgl-5.4.2.4 template-haskell-2.7.0.0 hoopl-3.8.7.3 binary-0.5.1.0 GLUT-2.1.2.1 network-2.3.1.0 cgi-3001.1.7.4 blaze-builder-0.3.1.0 stm-2.4 async-2.0.1.3 regex-posix-0.95.2 regex-compat-0.95.1 regex-base-0.93.2 parallel-3.2.0.3 haskell2010-1.1.0.1 haskell-src-1.0.1.5 fgl-5.4.2.4 vector-0.10.0.1 vector-algorithms-0.5.4.2 math-functions-0.1.1.2 template-haskell-2.7.0.0 hoopl-3.8.7.3 binary-0.5.1.0 GLUT-2.1.2.1 HUnit-1.2.5.1 Use --force-reinstalls if you want to install anyway. --- From: Johan Tibell [mailto:johan.tib...@gmail.com] Cc: haskell-cafe@haskell.org; Andres Löh Subject: Re: [Haskell-cafe] Cabal failures... I'm not quite sure what's going on. I've CCed Andres, who wrote the new constraint solver. One especially confusing part is this: C:\Users\guthrie\AppData\Local\Temp\Cabal-1.16.0.3-12392\Cabal-1.16.0.3\dist\set up\setup.exe configure --verbose=2 --ghc --prefix=C:\Users\guthrie\AppData\Roaming\cabal --user --flags=base4 --flags=base3 --constraint=process ==1.1.0.1 --constraint=pretty ==1.1.1.0 --constraint=old-time ==1.1.0.0 --constraint=filepath ==1.3.0.0 --constraint=directory ==1.1.0.2 --constraint=containers ==0.4.2.1 --constraint=base ==4.5.1.0 --constraint=array ==0.4.0.0 --disable-tests --disable-benchmarks Configuring Cabal-1.16.0.3... Flags chosen: base3=True, base4=True Dependency array ==0.4.0.0: using array-0.4.0.0 Dependency base ==4.5.1.0: using base-4.5.1.0 Dependency containers ==0.4.2.1: using containers-0.4.2.1 Dependency directory ==1.1.0.2: using directory-1.1.0.2 Dependency filepath ==1.3.0.0: using filepath-1.3.0.0 Dependency old-time ==1.1.0.0: using old-time-1.1.0.0 Dependency pretty ==1.1.1.0: using pretty-1.1.1.0 Dependency process ==1.1.0.1: using process-1.1.0.1 Warning: This package indirectly depends on multiple versions of the same package. This is highly likely to cause a compile failure. Why is Cabal setting both base3 and base4 to True? P.S. You can try the same command with --solver=topdown and see if that works. On Mon, Nov 19, 2012 at 8:22 PM, Gregory Guthrie guth...@mum.edu wrote: Johan, thanks for the note and information. My setup is: (Windows 7) cabal-install version 0.14.0 using version 1.14.0 of the Cabal library The Glorious Glasgow Haskell Compilation System, version 7.4.2 Haskell Platform 2012.4.0.0 . cabal-install-1.16.0.2 depends on Cabal-1.16.0.3 which failed to install. --- ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal failures...
On 12-11-19 04:25 PM, Gregory Guthrie wrote: I am not exert in the area, but I wonder how /why/ this is different than other package managers, like apt in Linux, I have never had any problems with it, and I would think that their dependencies are of at least similar complexities. I feel very strongly about the dissonance in comparing problems without comparing costs. Debian has a horde of volunteers for just the menial and manual work of perpetually finding one coherent set of versions so end users don't have to. And in practice, There is never one coherent set of versions. There is only a not-too-incoherent set of versions, and the volunteers first have to decide on it, and then manually pick patches from other versions (backporting patches, cherry-picking patches) to turn the not-too-incoherent set into a coherent set that does not exist in any pristine version. On top of that, Mark Shuttleworth actually pays money for Ubuntu to start from Debian and further test the set, pick some more patches, unpick some other patches... How many hours and/or dollars are you willing to pay for the menial, manual, perpetual chore of identifying coherent sets of versions so other people don't have to? And if a coherent set does not exist, how many are you willing to pay for backporting patches? At least I paid my 3 hours to explain some cabal stuff at http://www.vex.net/~trebla/haskell/sicp.xhtml Even the Haskell Platform, one very small set, costs volunteer hours. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-install-1.16.0.2
On Mon, Oct 15, 2012 at 7:16 PM, Johan Tibell johan.tib...@gmail.com wrote: Hi all, I've created bug fix release candidates for Cabal and cabal-install to address the bugs found after the release. Here's the list of fixed bugs: Fixed since cabal-install-1.16.0: * Fix installing from custom folder on Linux (#1058) * Change bootstrap.sh to require Cabal = 1.16 1.18 * Bump cabal-install version number to 1.16.0.1 * Bump network dependency in bootstrap.sh to 2.3.1.1 * Fix compilation error * Disable setting the jobs: $nprocs line in default ~/.cabal config * Fix building cabal-install with ghc-6.12 and older Fixed since Cabal-1.16.0.1: * Bump Cabal version number to 1.16.0.2 * Fixed warnings on the generated Paths module. The warnings are generated by the flag '-fwarn-missing-import-lists'. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] cabal-install-1.16.0.2
Hi Johan, On Mon, Oct 15, 2012 at 10:18 PM, Johan Tibell johan.tib...@gmail.com wrote: Fixed since Cabal-1.16.0.1: * Fixed warnings on the generated Paths module. The warnings are generated by the flag '-fwarn-missing-import-lists'. I tested this issue with Cabal-1.16.0.2 and the issue was fixed. Thanks, -- Andrés ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal dependencies
On Sat, 2012-10-06 at 17:02 +0200, José Lopes wrote: Hello, Hello I'm trying to understand Cabal dependencies. Why does the following situation happen? xmobar-0.15 depends on mtl-2.0.* and needs parsec All packages that will be broken, depends on parsec. But parsec is compiled with mtl-2.1.1 To install xmobar, cabal needs to reinstall parsec with mtl-2.0.1.0 Thanks, Yuras # cabal install xmobar --dry-run Resolving dependencies... In order, the following would be installed: parsec-3.1.3 (reinstall) changes: mtl-2.1.1 - 2.0.1.0 xmobar-0.15 (new package) Warning: The following packages are likely to be broken by the reinstalls: Best regards, José ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Cabal dependencies
OK. But, wouldn't it be possible for xmobar to use mtl-2.0.1.0 and for parsec to use mtl-2.1.1, while xmobar would use this parsec version? In this case, I am assuming that mtl-2.0.1.0 and mtl-2.1.1 are considered two different libraries. Thanks, José On 06-10-2012 17:17, Yuras Shumovich wrote: On Sat, 2012-10-06 at 17:02 +0200, José Lopes wrote: Hello, Hello I'm trying to understand Cabal dependencies. Why does the following situation happen? xmobar-0.15 depends on mtl-2.0.* and needs parsec All packages that will be broken, depends on parsec. But parsec is compiled with mtl-2.1.1 To install xmobar, cabal needs to reinstall parsec with mtl-2.0.1.0 Thanks, Yuras # cabal install xmobar --dry-run Resolving dependencies... In order, the following would be installed: parsec-3.1.3 (reinstall) changes: mtl-2.1.1 - 2.0.1.0 xmobar-0.15 (new package) Warning: The following packages are likely to be broken by the reinstalls: Best regards, José -- José António Branquinho de Oliveira Lopes Instituto Superior Técnico Technical University of Lisbon ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
