Ruby on Rails is not "someone else's framework." Over the years, hundreds of people have contributed to Ruby on Rails ranging from a single character to massive architectural changes or significant documentation -- all with the goal of making Ruby on Rails better for everyone. Even if you don't feel up to writing code or documentation yet, there are a variety of other ways that you can contribute, from reporting issues to testing patches.
Ruby on Rails uses "GitHub Issue Tracking":https://github.com/rails/rails/issues to track issues (primarily bugs and contributions of new code). If you've found a bug in Ruby on Rails, this is the place to start. You'll need to create a (free) GitHub account in order to submit an issue, to comment on them or to create pull requests.
NOTE: Bugs in the most recent released version of Ruby on Rails are likely to get the most attention. Also, the Rails core team is always interested in feedback from those who can take the time to test _edge Rails_ (the code for the version of Rails that is currently under development). Later in this guide you'll find out how to get edge Rails for testing.
If you've found a problem in Ruby on Rails which is not a security risk, do a search in GitHub under "Issues":https://github.com/rails/rails/issues in case it was already reported. If you find no issue addressing it you can "add a new one":https://github.com/rails/rails/issues/new. (See the next section for reporting security issues.)
At the minimum, your issue report needs a title and descriptive text. But that's only a minimum. You should include as much relevant information as possible. You need at least to post the code sample that has the issue. Even better is to include a unit test that shows how the expected behavior is not occurring. Your goal should be to make it easy for yourself -- and others -- to replicate the bug and figure out a fix.
Then, don't get your hopes up! Unless you have a "Code Red, Mission Critical, the World is Coming to an End" kind of bug, you're creating this issue report in the hope that others with the same problem will be able to collaborate with you on solving it. Do not expect that the issue report will automatically see any activity or that others will jump to fix it. Creating an issue like this is mostly to help yourself start on the path of fixing the problem and for others to confirm it with an "I'm having this problem too" comment.
WARNING: Please do not report security vulnerabilities with public GitHub issue reports. The "Rails security policy page":http://rubyonrails.org/security details the procedure to follow for security issues.
Please don't put "feature request" items into GitHub Issues. If there's a new feature that you want to see added to Ruby on Rails, you'll need to write the code yourself - or convince someone else to partner with you to write the code. Later in this guide you'll find detailed instructions for proposing a patch to Ruby on Rails. If you enter a wishlist item in GitHub Issues with no code, you can expect it to be marked "invalid" as soon as it's reviewed.
To move on from submitting bugs to helping resolve existing issues or contributing your own code to Ruby on Rails, you _must_ be able to run its test suite. In this section of the guide you'll learn how to set up the tests on your own computer.
Ruby on Rails uses git for source code control. The "git homepage":http://git-scm.com/ has installation instructions. There are a variety of resources on the net that will help you get familiar with git:
The test suite must pass with any submitted code. No matter whether you are writing a new patch, or evaluating someone else's, you need to be able to run the tests.
This command will install all dependencies except the MySQL and PostgreSQL Ruby drivers. We will come back to these soon. With dependencies installed, you can run the test suite with:
If you want to run the tests located in a specific directory use the +TEST_DIR+ environment variable. For example, this will run the tests of the +railties/test/generators+ directory only:
The test suite runs with warnings enabled. Ideally, Ruby on Rails should issue no warnings, but there may be a few, as well as some from third-party libraries. Please ignore (or fix!) them, if any, and submit patches that do not issue new warnings.
As of this writing (December, 2010) they are specially noisy with Ruby 1.9. If you are sure about what you are doing and would like to have a more clear output, there's a way to override the flag:
The test suite of Active Record attempts to run four times: once for SQLite3, once for each of the two MySQL gems (+mysql+ and +mysql2+), and once for PostgreSQL. We are going to see now how to set up the environment for them.
WARNING: If you're working with Active Record code, you _must_ ensure that the tests pass for at least MySQL, PostgreSQL, and SQLite3. Subtle differences between the various adapters have been behind the rejection of many patches that looked OK when tested only against MySQL.
The Active Record test suite requires a custom config file: +activerecord/test/config.yml+. An example is provided in +activerecord/test/config.example.yml+ which can be copied and used as needed for your environment.
The gem +sqlite3-ruby+ does not belong to the "db" group. Indeed, if you followed the instructions above you're ready. This is how you run the Active Record test suite only for SQLite3:
To be able to run the suite for MySQL and PostgreSQL we need their gems. Install first the servers, their client libraries, and their development files. In Ubuntu just run
We need first to delete +.bundle/config+ because Bundler remembers in that file that we didn't want to install the "db" group (alternatively you can edit the file).
In order to be able to run the test suite against MySQL you need to create a user named +rails+ with privileges on the test databases:
<shell>
mysql> GRANT ALL PRIVILEGES ON activerecord_unittest.*
to 'rails'@'localhost';
mysql> GRANT ALL PRIVILEGES ON activerecord_unittest2.*
If you’re using another database, check the files under +activerecord/test/connections+ for default connection information. You can edit these files to provide different credentials on your machine if you must, but obviously you should not push any such changes back to Rails.
You can invoke +test_jdbcmysql+, +test_jdbcsqlite3+ or +test_jdbcpostgresql+ also. See the file +activerecord/RUNNING_UNIT_TESTS+ for information on running more targeted database tests, or the file +ci/travis.rb+ for the test suite run by the continuous integration server.
If you want to add a fix to older versions of Ruby on Rails, you'll need to set up and switch to your own local tracking branch. Here is an example to switch to the 3-0-stable branch:
TIP: You may want to "put your git branch name in your shell prompt":http://qugstart.com/blog/git-and-svn/add-colored-git-branch-name-to-your-shell-prompt/ to make it easier to remember which version of the code you're working with.
As a next step beyond reporting issues, you can help the core team resolve existing issues. If you check the "Everyone's Issues":https://github.com/rails/rails/issues list in GitHub Issues, you'll find lots of issues already requiring attention. What can you do for these? Quite a bit, actually:
For starters, it helps just to verify bug reports. Can you reproduce the reported issue on your own computer? If so, you can add a comment to the issue saying that you're seeing the same thing.
If something is very vague, can you help squash it down into something specific? Maybe you can provide additional information to help reproduce a bug, or help by eliminating needless steps that aren't required to demonstrate the problem.
If you find a bug report without a test, it's very useful to contribute a failing test. This is also a great way to get started exploring the source code: looking at the existing test files will teach you how to write more tests. New tests are best contributed in the form of a patch, as explained later on in the "Contributing to the Rails Code" section.
Anything you can do to make bug reports more succinct or easier to reproduce is a help to folks trying to write code to fix those bugs - whether you end up writing the code yourself or not.
You can also help out by examining pull requests that have been submitted to Ruby on Rails via GitHub. To apply someone's changes you need first to create a dedicated branch:
Then you can use their remote branch to update your codebase. For example, let's say the GitHub user JohnSmith has forked and pushed to a topic branch "orange" located at https://github.com/JohnSmith/rails.
Once you're happy that the pull request contains a good change, comment on the GitHub issue indicating your approval. Your comment should indicate that you like the change and what you like about it. Something like:
If your comment simply says "+1", then odds are that other reviewers aren't going to take it too seriously. Show that you took the time to review the pull request.
You can help improve the Rails guides by making them more coherent, consistent or readable, adding missing information, correcting factual errors, fixing typos, or bringing it up to date with the latest edge Rails. To get involved in the translation of Rails guides, please see "Translating Rails Guides":https://wiki.github.com/lifo/docrails/translating-rails-guides.
If you're confident about your changes, you can push them directly yourself via "docrails":https://github.com/lifo/docrails. Docrails is a branch with an *open commit policy* and public write access. Commits to docrails are still reviewed, but this happens after they are pushed. Docrails is merged with master regularly, so you are effectively editing the Ruby on Rails documentation.
When working with documentation, please take into account the "API Documentation Guidelines":api_documentation_guidelines.html and the "Ruby on Rails Guides Guidelines":ruby_on_rails_guides_guidelines.html.
NOTE: As explained earlier, ordinary code patches should have proper documentation coverage. Docrails is only used for isolated documentation improvements.
NOTE: To help our CI servers you can add [ci skip] to your documentation commit message to skip build on that commit. Please remember to use it for commits containing only documentation changes.
WARNING: Docrails has a very strict policy: no code can be touched whatsoever, no matter how trivial or small the change. Only RDoc and guides can be edited via docrails. Also, CHANGELOGs should never be edited in docrails.
It doesn’t matter much what name you use, because this branch will only exist on your local computer and your personal repository on Github. It won't be part of the Rails git repository.
Now get busy and add or edit code. You’re on your branch now, so you can write whatever you want (you can check to make sure you’re on the right branch with +git branch -a+). But if you’re planning to submit your change back for inclusion in Rails, keep a few things in mind:
You should not be the only person who looks at the code before you submit it. You know at least one other Rails developer, right? Show them what you’re doing and ask for feedback. Doing this in private before you push a patch out publicly is the “smoke test” for a patch: if you can’t convince one other developer of the beauty of your code, you’re unlikely to convince the core team either.
You might want also to check out the "RailsBridge BugMash":http://wiki.railsbridge.org/projects/railsbridge/wiki/BugMash as a way to get involved in a group effort to improve Rails. This can help you get started and help you check your code when you're writing your first patches.
At this point, your editor should be fired up and you can write a message for this commit. Well formatted and descriptive commit messages are extremely helpful for the others, especially when figuring out why given change was made, so please take the time to write it.
Good commit message should be formatted according to the following example:
<plain>
Short summary (ideally 50 characters or less)
More detailed description, if necessary. It should be wrapped to 72
characters. Try to be as descriptive as you can, even if you think that
the commit content is obvious, it may not be obvious to others. You
should add such description also if it's already present in bug tracker,
it should not be necessary to visit a webpage to check the history.
Description can have multiple paragraps and you can use code examples
inside, just indent it with 4 spaces:
class PostsController
def index
respond_with Post.limit(10)
end
end
You can also add bullet points:
- you can use dashes or asterisks
- also, try to indent next line of a point for readability, if it's too
Navigate to the Rails repository you just pushed to (e.g. https://github.com/your-user-name/rails) and press "Pull Request" in the upper right hand corner.
Fill in some details about your potential patch including a meaningful title. When finished, press "Send pull request". The Rails core team will be notified about your submission.
Now you need to get other people to look at your patch, just as you've looked at other people's patches. You can use the "rubyonrails-core mailing list":http://groups.google.com/group/rubyonrails-core/ or the #rails-contrib channel on IRC freenode for this. You might also try just talking to Rails developers that you know.
It’s entirely possible that the feedback you get will suggest changes. Don’t get discouraged: the whole point of contributing to an active open source project is to tap into community knowledge. If people are encouraging you to tweak your code, then it’s worth making the tweaks and resubmitting. If the feedback is that your code doesn’t belong in the core, you might still think about releasing it as a gem.
Changes that are merged into master are intended for the next major release of Rails. Sometimes, it might be beneficial for your changes to propagate back to the maintenance releases for older stable branches. Generally, security fixes and bug fixes are good candidates for a backport, while new features and patches that introduce a change in behavior will not be accepted. When in doubt, it is best to consult a rails team member before backporting your changes to avoid wasted effort.
For simple fixes, the easiest way to backport your change is to "extract a diff from your changes in master and apply them to the target branch":http://ariejan.net/2009/10/26/how-to-create-and-apply-a-patch-with-git.
First make sure your changes are the only difference between your current branch and master:
Switch over to the target branch and apply your changes:
<shell>
$ git checkout -b my_backport_branch 3-2-stable
$ git apply ~/my_changes.patch
</shell>
This works well for simple changes. However, if your changes are complicated or if the code in master has deviated significantly from your target branch, it might require more work on your part. The difficulty of a backport varies greatly from case to case, and sometimes it is simply not worth the effort.
Once you have resolved all conflicts and made sure all the tests are passing, push your changes and open a separate pull request for your backport. It is also worth noting that older branches might have a different set of build targets than master. When possible, it is best to first test your backport locally against the ruby versions listed in +.travis.yml+ before submitting your pull request.