As a part of installing Ubuntu on my computer, i apt-get installed cabal. After running cabal update, i ran cabal install darcs, discovered that i needed curses, installed libncurses5-dev, discovered that i needed curl, installed that, and when it still didn't work, ran cabal install darcs -f-curl.
This worked, but darcs doesn't run from the shell when just running darcs. It was installed in ~/.cabal/bin/darcs, so i made a link from there to /bin. Darcs now works from the shell, but i am left wondering if there is a better way to do it, perhaps running cabal install under sudo?
cabal by default installs without needing administrator privileges. As such, it can't install into e.g. /usr/bin. Since $HOME/bin isn't a standard, the compromise is to put executables into ~/.cabal/bin.
You can then add ~/.cabal/bin to your path to see these executables (e.g. by appending the following to your .bashrc:
export $PATH=$PATH:~/.cabal/bin
If you have admin privledges in your machine, by all means use the --global flag.
Firstly, do not use sudo. Even when run under root, cabal builds stuff in your home directory, so using sudo indiscriminately can lead to root-owned files in your ~/.cabal. Instead, set root-cmd: sudo in your /.cabal/config, then cabal with the --global option will invoke that command as appropriate.
Secondly, using local installs is generally more convenient and safer than using global ones, because you keep your package manager and cabal separate. So you could just as well add this to your ~/.bashrc:
export PATH=$HOME/.cabal/bin:$PATH
and then continue to install stuff without using sudo at all.
See also: a slightly over-pessimistic, but nevertheless informative article on cabal package management.
The --global flag is what is needed in this case. Running sudo cabal install darcs -f-curl --global, with -f-curl because cabal couldn't find curl, --global to make it install into usr/local/bin, and sudo just to be sure, works.
Related
Is it possible to use RPM or YUM or any other package manager in Linux, specifically CentOS, to install a package either already downloaded or from repo to a custom location without admin/root access?
I tried building from sources, using cmake, configure, make, make install etc, but, it ended up having so many dependencies one after other.
Or are there any better alternatives?
It is possible to use yum and rpm to install any package in the repository of the distribution. Here is the recipe:
Find the package name
Use yum search.
Download
Download the package and all of its dependencies using yumdownloader (which is available on CentOS by default). You'll need to pass it --resolve to get dependency resolution. yumdownloader downloads to the current directory unless you specify a --destdir.
mkdir -p ~/rpm
yumdownloader --destdir ~/rpm --resolve vim-common
Choose a prefix location
It might be ~, ~/centos, or ~/y. If your home is slow because it is on a network file system, you can put it in /var/tmp/....
mkdir ~/centos
Extract all .rpm packages
Extract all .rpm packages to your chosen prefix location.
cd ~/centos && rpm2cpio ~/rpm/x.rpm | cpio -id
rpm2cpio outputs the .rpm file as a .cpio archive on stdout.
cpio reads it from from stdin
-i means extract (to the current directory)
-d means create missing directory
You can optionally use -v: verbose
Configure the environment
You will need to configure the environment variable PATH and LD_LIBRARY_PATH for the installed packages to work correctly. Here is the corresponding sample from my ~/.bashrc:
export PATH="$HOME/centos/usr/sbin:$HOME/centos/usr/bin:$HOME/centos/bin:$PATH"
export MANPATH="$HOME/centos/usr/share/man:$MANPATH"
L='/lib:/lib64:/usr/lib:/usr/lib64'
export LD_LIBRARY_PATH="$HOME/centos/usr/lib:$HOME/centos/usr/lib64:$L"
Edited note (thanks to #AmitNaidu for pointing out my mistake):
According to bash documentation about startup files, when connecting to a server via ssh, only .bashrc is sourced:
Invoked by remote shell daemon
Bash attempts to determine when it is being run with its standard input connected to a network connection, as when executed by the remote shell daemon, usually rshd, or the secure shell daemon sshd. If Bash determines it is being run in this fashion, it reads and executes commands from ~/.bashrc, if that file exists and is readable.
Now if you want to install a lot of packages that way, you might want to automate the process. If so, have a look at this repository.
Extra note: if you are trying to install any of gcc, zlib, make, cmake, git, fish, zsh or tmux , you should really consider using conda, see my other answer.
TL;DR Use Miniconda, conda-forge is amazing.
curl "https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh" | sh
Or, alternatively:
curl https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh > Miniconda.sh
bash Miniconda.sh -b -p ~/conda
# -b is used to specify that this is done "in batch", so skip the EULA prompt
# -p lets you specify where you want conda installed
Commonly wanted packages:
gcc conda install gcc
zlib conda install zlib
make conda install make
cmake conda install cmake
git conda install git
fish conda install -c conda-forge fish
zsh conda install -c ActivisionGameScience zsh
tmux conda install -c conda-forge tmux
This tmux has a bug with the name of the ncurse library it uses. You can work around it by going to your da/lib folder and symlinking ln -sT libtinfow.so.6.1 libtinfo.so.6
For the rest, you can try https://anaconda.org/search?q=.
I've tried for a long time to get a package manager to work well on CentOS/RedHat but without success. The best I could do was to install a Gentoo Prefix at the correct location on another CentOS with root access, then scp a .tar.xz of the whole installation to the target server (only way to get a proper gcc for Gentoo Prefix). I could emerge (build & install) packages on the target server but kept hitting problems with locals and permissions.
I recently achieved a user installation of some interesting packages using conda. Here is how to install it from the command line:
curl "https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh" | sh
If like me, your home folder is hosted on a remote drive (a network file system), you might not want to install it in your home folder, so you might want to use something like mkdir /var/tmp/lo then specify an installation folder like /var/tmp/lo/da during the installation.
You'll then be able to install quite a lot of packages, though maybe not all those you wanted. Most of the time, if it is not in the default channel, it will be in conda-forge. You can check for existing packages at https://anaconda.org/search?q=
Other package managers I've tried to use after conda:
Linuxbrew
I thought that with that it would be easy to install homebrew (linuxbrew) but their sources are messy and use hard-coded absolute path to ruby interpreter, which fails because it isn't the last version and so on and so on and I gave up.
Nix
Nix still requires you to use the /nix folder. They hard-coded it too and it's hard to sed it correctly from every download it has to do during the installation (let alone updates).
Gentoo Prefix
I expect Gentoo Prefix to be easier to install directly now that we gcc can be used on the target server. -- Ok, I tried but met permissions bugs during installation (2018-09-28):
portage.exception.OperationNotPermitted: chown(b'~/gentoo/tmp/var/tmp/portage/sys-apps/gentoo-functions-0.12/image/var', 2000, 2000)
PkgSrc
I'm going to try pkgsrc now. -- Use (older) version 64-bit EL 6.x if on CentOS 6 or if encountering (G)LibC version issues with the 7.x one. -- No luck, pkgsrc hard codes /usr/pkg/sbin and /usr/pkg/bin. So it can't be used as user, unless maybe setting up a fakechroot environment. But I've never done that and I expect usability issues.
Please comment/answer if you succeed in installing any other package manager.
Download the packages, and indicate to include dependencies with the --resolve flag.
yumdownloader --resolve openslide-tools
Iterate over all downloaded rpm files.
for i in *.rpm; do rpm2cpio $i | cpio -idv; done
the output will be stored in your present working directory $PWD/usr/*
This answer by goldilocks sounds like what you are looking for.
https://unix.stackexchange.com/a/61295
It's still not a pretty process, but seems easier than building from source.
Otherwise you might want to look into non-root package managers as an alternative to yum.
Yes it is. If the software is packaged in repos. And admin installed
PackageKit-command-not-found package.
See:
https://fedoraproject.org/wiki/Features/PackageKitCommandNotFound
I'm a newbie to haskell and cabal, so I'm probably missing something simple.
I updated cabal-install:
sudo cabal install cabal-install
Password:
Resolving dependencies...
Configuring cabal-install-1.22.0.0...
Building cabal-install-1.22.0.0...
Installed cabal-install-1.22.0.0
Updating documentation index
However cabal --version says:
cabal-install version 1.18.0.5
using version 1.18.1.4 of the Cabal library
What happened to cabal-install 1.22.0.0?
There are two ways of making cabal install packages globally. Note that, as a result, cabal may require sudo.
This command will install <PACKAGE> globally:
$ cabal install <PACKAGE> --global
As a more general solution, edit the file ~/.cabal/config and set user-install to False. This will automatically set the --global flag so you can just write cabal install <PACKAGE> without any worry. Here's a snippet of my config file:
...
-- split-objs: False
-- executable-stripping: True
user-install: False
-- package-db:
-- flags:
...
You may also want to set root-cmd to sudo if it's not already, so that cabal will automatically prompt for the root password when it encounters a permission problem.
There's some more info online here.
I see that there's an updated cabal at ~/Library/Haskell/bin, so I could replace /usr/bin/cabal with a symbolic link to this copy or I could copy this binary to /usr/bin.
I'm still interested if there is a more elegant/canonical way to make sure the new cabal is what gets used by default.
TLDR: Try running hash -r
Bash has a PATH hashtable that maps commands to the location of binaries. You may still have an old version of cabal installed somewhere in your PATH (possibly in a sandbox). Since cabal is not a new command, the hashtable will keep serving up the old version. hash -r rebuilds the hashtable, so the shell will correctly find the new version (providing it appears earlier in your path than the old one).
Is there any apt-get-like program for use with Cygwin?
I already tried cyg-apt but when I try I get this error:
cyg-apt: downloading: http://cygwin.mirrors.pair.com/setup-2.bz2
cyg-apt: downloading: http://cygwin.mirrors.pair.com/setup-2.ini
cyg-apt: bad URL http://cygwin.mirrors.pair.com/setup-2.ini, exiting.
Best I have ever used:
apt-cyg package manager
You can do this using Cygwin’s setup.exe from Windows command line. Example:
cd C:\cygwin64
setup-x86_64 -q -P wget,tar,gawk,bzip2,subversion,vim
For a more convenient installer, you may want to use the
apt-cyg package manager. Its syntax is
similar to apt-get, which is a plus. For this, follow the above steps and then
use Cygwin Bash for the following steps:
wget rawgit.com/transcode-open/apt-cyg/master/apt-cyg
install apt-cyg /bin
Now that apt-cyg is installed. Here are a few examples of installing some
packages:
apt-cyg install nano
apt-cyg install git
apt-cyg install ca-certificates
Update: you can read the more complex answer, which contains more methods and information.
There exists a couple of scripts, which can be used as simple package managers. But as far as I know, none of them allows you to upgrade packages, because it’s not an easy task on Windows since there is not possible to overwrite files in use. So you have to close all Cygwin instances first and then you can use Cygwin’s native setup.exe (which itself does the upgrade via “replace after reboot” method, when files are in use).
apt-cyg
The best one for me. Simply because it’s one of the most recent. It works correctly for both platforms - x86 and x86_64. There exists a lot of forks with some additional features. For example the kou1okada fork is one of improved versions.
Cygwin’s setup.exe
It has also command line mode. Moreover it allows you to upgrade all installed packages at once.
setup.exe-x86_64.exe -q --packages=bash,vim
Example use:
setup.exe-x86_64.exe -q --packages="bash,vim"
You can create an alias for easier use, for example:
alias cyg-get="/cygdrive/d/path/to/cygwin/setup-x86_64.exe -q -P"
Then you can for example install the Vim package with:
cyg-get vim
you can always make a bash alias to setup*.exe files in $home/.bashrc
cygwin 32bit
alias cyg-get="/cygdrive/c/cygwin/setup-x86.exe -q -P"
cygwin 64bit
alias cyg-get="/cygdrive/c/cygwin64/setup-x86_64.exe -q -P"
now you can install packages with
cyg-get <package>
This got it working for me:
curl https://raw.githubusercontent.com/transcode-open/apt-cyg/master/apt-cyg > \
apt-cyg && install apt-cyg /bin
No. The only officially supported tool for downloading and updating Cygwin packages is the setup.exe file you used for the initial install, although that can be invoked with command line arguments to help the process.
From that same page:
The basic reason for not having a more full-featured package manager is that such a program would need full access to all of Cygwin's POSIX functionality. That is, however, difficult to provide in a Cygwin-free environment, such as exists on first installation. Additionally, Windows does not easily allow overwriting of in-use executables so installing a new version of the Cygwin DLL while a package manager is using the DLL is problematic.
You can use Chocolatey to install cyg-get and then install your packages with it.
For example:
choco install cyg-get
Then:
cyg-get install my-package
I want to compile my program with profiling, so I run:
$ cabal configure --enable-executable-profiling
...
$ cabal build
...
Could not find module 'Graphics.UI.GLUT':
Perhaps you havent installed the profiling libraries for package 'GLUT-2.2.2.0'?
...
$ # indeed I have not installed the prof libs for GLUT, so..
$ cabal install -p GLUT --reinstall
...
Could not find module 'Graphics.Rendering.OpenGL':
Perhaps you havent installed the profiling libraries for package 'OpenGL-2.4.0.1'?
...
So, the problem is, that unlike cabal's usual welcome behavior, cabal doesn't resolve the dependencies and install them when needing profiling libraries.
I can work around it by resolving the dependencies manually (by following errors that appear after a while of compiling):
$ cabal install -p OpenGLRaw --reinstall
$ cabal install -p StateVar --reinstall
$ cabal install -p Tensor --reinstall
$ cabal install -p ObjectName --reinstall
$ cabal install -p GLURaw --reinstall
$ cabal install -p OpenGL --reinstall
$ cabal install -p GLUT --reinstall
And then repeat for my next dependency..
Is there a better way to do this? i.e do make cabal do the work on its own as it does for normal libraries?
I've enabled library-profiling: True in my ~/.cabal/config file. From then on, any new installations will automatically enable profiling.
Unfortunately that still means I had to manually reinstall for the old packages already installed. Although, after a while of doing this manually, I now have most packages reinstalled with profiling enabled...
From a comment by Tom Lokhorst:
I do hope someone will come along with a better answer, one that would not require me to reinstall the complete Haskell Platform manually next time.
For future visitors:
The task of installing profiling versions of all installed libraries has become less of a chore, cabal (cabal-install) now keeps track of what was installed using it in the world file in the .cabal directory (on linux, that would be $HOME/.cabal, on Windows something like C:\Users\%YOU%\AppData\Roaming\cabal\, on OSX ??).
So after enabling profiling in the config file (in the same directory), and clearing GHC's package database (you can find the locations of the global and user db per ghc-pkg list nonexisting; remove the cabal-installed packages from the global database with ghc-pkg unregister packagename if you have any, rename or delete the entire user db - this is necessary because the world file only tracks explicitly installed packages, not their dependencies), installing everything with profiling support should work as follows:
$ cabal install --reinstall world --dry-run
First run with --dry-run to check for problems before actually reinstalling anything. If it would reinstall boot packages like process or directory, that's a bad sign, if you don't know how to handle it, ask on the #haskell IRC channel, one of the mailing lists, or here for guidance. If it fails to find a consistent install plan due to new versions on hackage of some packages which are incompatible with each other, that can usually be solved by editing the world file and constraining allowable versions of some packages.
Then, if you are optimistic that nothing will badly break,
$ cabal install --reinstall world
and have a nice pot of tea while GHC is busy compiling.
Daniel Fischer's answer looks good, but for some reason my ~/.cabal/world library only contained entries for libraries directly installed, and not their dependencies.
Instead, I dumped out a list of all installed libraries using
$ ghc-pkg list > list
This lists the libraries installed system-wide and locally. Therefore, I edited the list file to remove the first portion (containing libraries installed system-wide) leaving only the lines after /home/<user>/.ghc/.... Finally, I ran
$ cabal install --reinstall $(cat list)
This worked for me. You should maybe do --dry-run first. Then go make a pot of tea. Or bake a cake.
it appears there is no way right now: Ticket #282 - profiling versions of libraries not managed well "As usual the problem is lack of devevloper time to implement all these
nice features we all want."
For visitors 2016+: Just install ghc-prof
Debian Linux Systems:
sudo apt-get install ghc-prof
Arch Linux Systems:
sudo pacman -S ghc-prof
I want to compile my program with profiling, so I run:
$ cabal configure --enable-executable-profiling
...
$ cabal build
...
Could not find module 'Graphics.UI.GLUT':
Perhaps you havent installed the profiling libraries for package 'GLUT-2.2.2.0'?
...
$ # indeed I have not installed the prof libs for GLUT, so..
$ cabal install -p GLUT --reinstall
...
Could not find module 'Graphics.Rendering.OpenGL':
Perhaps you havent installed the profiling libraries for package 'OpenGL-2.4.0.1'?
...
So, the problem is, that unlike cabal's usual welcome behavior, cabal doesn't resolve the dependencies and install them when needing profiling libraries.
I can work around it by resolving the dependencies manually (by following errors that appear after a while of compiling):
$ cabal install -p OpenGLRaw --reinstall
$ cabal install -p StateVar --reinstall
$ cabal install -p Tensor --reinstall
$ cabal install -p ObjectName --reinstall
$ cabal install -p GLURaw --reinstall
$ cabal install -p OpenGL --reinstall
$ cabal install -p GLUT --reinstall
And then repeat for my next dependency..
Is there a better way to do this? i.e do make cabal do the work on its own as it does for normal libraries?
I've enabled library-profiling: True in my ~/.cabal/config file. From then on, any new installations will automatically enable profiling.
Unfortunately that still means I had to manually reinstall for the old packages already installed. Although, after a while of doing this manually, I now have most packages reinstalled with profiling enabled...
From a comment by Tom Lokhorst:
I do hope someone will come along with a better answer, one that would not require me to reinstall the complete Haskell Platform manually next time.
For future visitors:
The task of installing profiling versions of all installed libraries has become less of a chore, cabal (cabal-install) now keeps track of what was installed using it in the world file in the .cabal directory (on linux, that would be $HOME/.cabal, on Windows something like C:\Users\%YOU%\AppData\Roaming\cabal\, on OSX ??).
So after enabling profiling in the config file (in the same directory), and clearing GHC's package database (you can find the locations of the global and user db per ghc-pkg list nonexisting; remove the cabal-installed packages from the global database with ghc-pkg unregister packagename if you have any, rename or delete the entire user db - this is necessary because the world file only tracks explicitly installed packages, not their dependencies), installing everything with profiling support should work as follows:
$ cabal install --reinstall world --dry-run
First run with --dry-run to check for problems before actually reinstalling anything. If it would reinstall boot packages like process or directory, that's a bad sign, if you don't know how to handle it, ask on the #haskell IRC channel, one of the mailing lists, or here for guidance. If it fails to find a consistent install plan due to new versions on hackage of some packages which are incompatible with each other, that can usually be solved by editing the world file and constraining allowable versions of some packages.
Then, if you are optimistic that nothing will badly break,
$ cabal install --reinstall world
and have a nice pot of tea while GHC is busy compiling.
Daniel Fischer's answer looks good, but for some reason my ~/.cabal/world library only contained entries for libraries directly installed, and not their dependencies.
Instead, I dumped out a list of all installed libraries using
$ ghc-pkg list > list
This lists the libraries installed system-wide and locally. Therefore, I edited the list file to remove the first portion (containing libraries installed system-wide) leaving only the lines after /home/<user>/.ghc/.... Finally, I ran
$ cabal install --reinstall $(cat list)
This worked for me. You should maybe do --dry-run first. Then go make a pot of tea. Or bake a cake.
it appears there is no way right now: Ticket #282 - profiling versions of libraries not managed well "As usual the problem is lack of devevloper time to implement all these
nice features we all want."
For visitors 2016+: Just install ghc-prof
Debian Linux Systems:
sudo apt-get install ghc-prof
Arch Linux Systems:
sudo pacman -S ghc-prof