Rip is an attempt to create a next generation packaging system for Ruby.
For more thorough documentation please see the Rip site:
Let's get right to it then, shall we?
First we install rip.
$ sudo gem install rip
Did that work?
$ rip check All systems go.
Yep. Let's see what libraries rip knows about.
$ rip list ripenv: base nothing installed
None. Really? Let's try to require Grit.
$ ruby -r grit -e 'puts Grit' ruby: no such file to load -- grit (LoadError)
Whoops. Not found. Let's install the latest using rip.
$ rip install git://github.com/defunkt/grit.git v1.1.1b Successfully installed grit v1.1.1 $ rip list ripenv: base diff-lcs (491fbc0) mime-types (v1.16) grit (v1.1.1)
Great, now we have Grit and all its dependencies.
$ ruby -r grit -e 'puts Grit' Grit $ ruby -r grit -e 'puts MIME::Types' MIME::Types
And we don't need any magical
require statements to use them!
If we'd like, we can now move between rip environments to mix and match libraries.
$ rip env create my_fresh_env ripenv: created my_fresh_env $ ruby -r grit -e 'puts Grit' ruby: no such file to load -- grit (LoadError)
And so much more.
Rip can install from a variety of sources: directories, single files, git repositories - even Rubygems.
Adding a new package is easy and we expect Rip to support more formats in the future.
Virtual environments ("ripenvs") can be created so multiple versions of a package may be installed and used by different applications concurrently.
ripenvs are easy to create, copy, delete, and share. Recipes for creating ripenvs are trivial to generate and publish.
Dependency checking happens when Rip packages are installed, not when packages run.
A dependency graph is constructed for the entire virtual environment, making it easy to debug leaf-node transitive dependency issues.
Installation and dependency errors should be clear and give as much information as possible in order to help you fix the problem.
Rip itself should work without anything but the Ruby standard library, for maximum portability.
Rip's support for a variety of package types means there is nothing to build and distribute.
Tag your Git repository and publicize the latest version, or just pass around Gists. Rip does not care.
Rip dependencies are listed as separate lines in a plaintext file and can reference any package type. As a result, Rip packages can depend on existing Rubygems that aren't available from any other source.
This means projects unaware of Rip can be installed by Rip and managed by ripenvs. Adding the dependencies yourself is easy.
Rip makes it easy to have multiple environments with different versions of libraries.
You could even clone a ripenv then upgrade a single library to test its impact on the environment as a whole. Installation not go smoothly? Delete the new ripenev then continue using the stable one.
With Rip, version conflicts in dependencies are simpler to resolve: you know exactly what version of which libraries are requesting which versions of the same library at installation time. As a result conflicts are resolved when you're thinking about installing your code, not later on when you're thinking about running it.
Rip requires no changes to your code, only an optional
deps.rip file added
to the root of your project. As a result you do not force Rip on anyone else
and individuals are free to re-package your code using other systems.
There is no canonical server for Rip packages, which may be good or bad.
Each Rip package optionally contains at the root a
identifying it as a Rip package and listing its dependencies.
Let's take the ambition project as an example. This is its
deps.rip in full:
git://github.com/drnic/rubigen.git REL-1.3.0 git://github.com/seattlerb/ruby2ruby.git e3cf57559 # 1.1.8 git://github.com/seattlerb/parsetree.git 480ede9d9 # 2.1.1
If you were to run
rip install git://github.com/defunkt/ambition
the following steps would occur:
- The source would be fetched and unpacked as
ambitionin the cwd
- The source of each dependency in
deps.ripwould be fetched
- Each dependency would be unpacked into the current ripenv at the revision or tag specified
- Each dependency's
deps.ripwould be fetched and unpacked into the ripenv, etc
As this process unfolds, a mapping of libraries and versions is kept in memory. When a library is declared multiple times at different versions the process is halted and the error reported.
If you've cloned
ambition on your own you can still install the
rip install deps.rip
The easiest way to mass uninstall packages is to delete your ripenv
and create a new one. Otherwise,
rip uninstall package will do the trick.
Rip will complain if you attempt to uninstall a package that others depend
on. To remove the package anyway, use
-y. To remove the package and the
-d (for dependents).
Installing a package that is not Rip aware but needs an extension
Rip will attempt to run
rake rip:install in your library if a
Rakefile is found. If you need to compile your C extension or do
any other work, this is the place.
Rip is currently user-specific.
Here is a typical directory structure for Rip:
rip/ - rip-packages/ - active/ - bin/ - lib/ - base/ - bin/ - lib/ - base.ripenv - cheat/ - bin/ - lib/ - cheat.ripenv - thunderhorse/ - bin/ - lib/ - thunderhorse.ripenv
The above contains three ripenvs:
ripenv contains directories for executable binaries and Ruby source files.
They also include a generated
.ripenv file containing metadata about
the ripenev and its packages.
This individual may use
base for general tomfoolery (it's the
cheat for developing their Cheat application, and
thunderhorse for working on their new Thunderhorse project.
active is a symlink to the current, active ripenv. We also see a
rip-packages directory. This is where Rip stores the raw repositories.
Let us focus on the
rip/ - cheat/ - bin/ - camping - lib/ - markaby/ - builder.rb - cssproxy.rb - metaid.rb - rails.rb - tags.rb - template.rb - markaby.rb - camping/ - db.rb - fastcgi.rb - reloader.rb - session.rb - webrick.rb - camping-unabridged.rb - camping.rb - cheat.ripenv
When using the
cheat ripenv, a
camping binary will be in our
When running Ruby scripts, or even executing the
rip/cheat/lib will be in our $LOAD_PATH. Therefor we may, for instance,
require "markaby" in a Ruby script and it will succeed.
In any other ripenv, the
cheat ripenv's binaries and libraries are as
good as non-existant.
Want to get a copy of your local environment on your deployment
server? Generate a
.rip file with
rip freeze then upload and
Currently it's UNIX-only. This is because Rip needs to manipulate the RUBYLIB and PATH environment variables so that Ruby knows where to find installed Rip packages.
As a result, the setup script expects you to be running bash or zshell.
- Chris Wanstrath (that's me!)
- Jeff Hodges
- Tom Preston-Werner
- John Barnette
- Blake Mizerany
- Ryan Tomayko
- Pat Nakajima
- Coda Hale for the phrase "leaf-node transitive dependency issues"
- GitHub for sponsoring development