It is strongly recommended that students who want to apply to the Git project for the Summer of Code 2016 submit a small code-related patch to the Git project as part of their application. Think of these microprojects as the “Hello, world” of getting involved with the Git project; the coding aspect of the change can be almost trivial, but to make the change the student has to become familiar with many of the practical aspects of working on the Git project.
NOTE: Students who plan to work on libgit2, which is also under the Git umbrella in the Google Summer of Code, should refer to the libgit2 list of projects rather than the list below.
Git development is based on sending successive versions of patches or patch series to the mailing list until they are considered good and correct by the reviewers and Junio Hamano, the maintainer, who will merge them. This process usually takes quite some time. By sending drafts of your microproject patches to the mailing list long before the deadline, you can show us that you are willing and able to work well using the Git development process.
It is expected that what you send will need several rounds of reviews and discussions. If you are not sure at all about a patch you can put “[RFC/PATCH]” at the beginning of its subject.
Consider a sample email thread, which shows how a developer proposed a change and a patch to implement it. The problem being solved, the design of the proposed solution, and the implementation of that design were all reviewed and discussed, and after several iterations an improved version of the patch was accepted into our codebase. As a GSoC student, you will be playing the role of the developer and engaging in a similar discussion. Get familiar with the flow, need for clarity on both sides (i.e. you need to clearly defend your design, and need to ask clarifications when questions/suggestions you are offered are not clear enough), the pace at which the discussion takes place, and the general tone of the discussion, to learn what is expected of you.
To complete a microproject, you will have to go through approximately the following steps:
Download the source code: clone the repository using the Git via
Git instructions and read the
README file.
Build the source code: this is described in the file INSTALL.
Glance over our coding guidelines in the file
Documentation/CodingGuidelines. We take things like proper code
formatting very seriously.
Read about the process for submitting patches to Git: this is
described in Documentation/SubmittingPatches.
Make the actual change. (Funny, this is the only part they teach you about in college.)
Run the test suite and make sure it passes 100%: this is described
in the file t/README. (If you have added new functionality, you
should also add new tests, but most microprojects will not add new
functionality.)
Commit your change. Surprise: we use Git for that, so you will need
to gain at least
a basic familiarity with using
Git. Make sure to write a good commit message that explains the
reason for the change and any ramifications. Remember to make sure
that you agree with our “Developer’s Certificate of Origin” (whose
text is contained in Documentation/SubmittingPatches), and to
signify your agreement by adding a Signed-off-by line.
Optional, but recommended: Create a fork of Git on github.com, sign in to Travis CI with your GitHub account, accepting the GitHub access permissions confirmation, and enable builds of your Git fork in your Travis CI profile (you only have to do these steps once). Afterwards you can push your changes to your fork and on https://travis-ci.org/your-github-name/git/builds (example here) you can see if your changes pass the Git test suite on Ubuntu Linux and on Mac OS X.
Submit your change to the Git mailing list. For this step you
probably want to use the commands git format-patch and git
send-email. Make sure that your email is formatted correctly: send
a test version of the email to yourself and see if you can apply it
to your repository using git am.
Expect feedback, criticism, suggestions, etc. from the mailing list.
Respond to it! and follow up with improved versions of your change. Even for a trivial patch you shouldn’t be surprised if it takes two or more iterations before your patch is accepted. This is the best part of participating in the Git community; it is your chance to get personalized instruction from very experienced peers!
The coding part of the microproject should be very small (say, 10-30 minutes). We don’t require that your patch be accepted into master by the time of your formal application; we mostly want to see that you have a basic level of competence and especially the ability to interact with the other Git developers.
When you submit your patch, please mention that you plan to apply for the GSoC. This will ensure that we take special care not to overlook your application among the large pile of others.
Students: Please attempt only ONE microproject. We want quality, not quantity! (Also, it takes work to collect the ideas, and it would be nice to have enough microprojects for everybody.) If you’ve already done a microproject and are itching to do more, then get involved in other ways, like finding and fixing other problems in the code, or improving the documentation or code comments, or helping to review other people’s patches on the mailing list, or answering questions on the mailing list or in IRC, or writing new tests, etc., etc. In short, start doing things that other Git developers do!
The following are just ideas. Any small code-related change would be suitable. Just remember to keep the change small! It is much better for you to finish a small but complete change than to try something too ambitious and not get it done.
Pick one command that operates on branch names. Teach it the “-“ shorthand that stands for “the branch we were previously on”, like we did for “git merge -“ sometime after we introduced “git checkout -“. Cf. $gmane/230828
Pick one field of a structure that (1) is of signed integral type and (2) is used as a collection of multiple bits. Discuss if there is a good reason why it has to be a signed integral field and change it to an unsigned type otherwise. Cf. $gmane/263751
Most of git dotfiles can be located, at the user’s option, in
~/.
The suggested approach is:
See how XDG was implemented for other files (run “git log --grep
XDG” in Git’s source code) and read the XDG specification.
Implement and test the new behavior, without breaking compatibility with the old behavior.
Update the documentation
Even though the amount of code to write is small, these projects involve a lot of prior work to understand the specification and deal with all potential corner-cases.
This includes:
Many have already been added (e.g. “git am -3” in e97a5e7).
Some people always run the command with these options, and would prefer to be able to activate them by default in ~/.gitconfig.
“git diff” shows the function name corresponding to each hunk after the @@ … @@ line. For common languages (C, HTML, Ada, Matlab, …), the way to find the function name is built-in Git’s source code as regular expressions (see userdiff.c). A few languages are common enough to deserve a built-in driver, but are not yet recognized. For example, CSS, shell.
This project requires a very good knowledge of regular expressions.
git tag --contains <id>” less chatty if git tag --contains
These still use a hand-rolled option parser, which can be replaced by using the parse-options api. Each of these files can be a microproject of its own.
Automated testing is an important safety net for complex software such as Git. This micro project is about to improve the Git Travis CI integration.
Investigate if we can trigger Coverity static code analysis for the Git master and maint branch (hint: Stefan Beller already looked into this). Start here: https://scan.coverity.com/travis_ci
Automated testing is an important safety net for complex software such as Git. This micro project is about to improve the Git Travis CI integration.
Investigate if we can enable and run Clang static code analysis for the master and maint branch.
Automated testing is an important safety net for complex software such as Git. This micro project is about to improve the Git Travis CI integration.
Investigate if we can use pylint to analyze the git-p4 Python code.
Automated testing is an important safety net for complex software such as Git. This micro project is about to improve the Git Travis CI integration.
Automated testing is an important safety net for complex software such as Git. This micro project is about to improve the Git Travis CI integration.
Git’s test suit is huge and over time we have seen some flaky test. Build a web page that analyzes the Travis CI test results and prints the test that fail most often. Use this implementation as starting point: https://scribu.github.io/travis-stats/#git/git
After you have done this look at the randomly failing tests and try to figure out why they fail. See here for an example of such a test failure.
git pull --rebase” the “--[no-]autostash” optionAlthough since 53c76dc0 git-pull understands the “rebase.autoStash”
configuration option, it currently does not support the corresponding
“--[no-]autostash” command-line switch. Teach git-pull to accept this
command-line switch and pass it to git-rebase. Cf. $gmane/283557
If you don’t like for some reason the above microprojects or if you just want more choice, you may find other ideas for microprojects by searching the mailing list or the code base itself. In the code base you could search the code itself or the tests (in the “t” directory).
When you find something you are interested to work on, please ask first on the mailing list if it’s worth doing and if it’s appropriate for a microproject before starting to work on what you find. Even if it looks straitforward, there could be hidden reasons why it is too difficult or just inappropriate.
Your best bet is probably to search for strings like “FIXME”, “TODO”, “NEEDSWORK”, or maybe “NEED-WORK”, and “BUG”.
Tests are in the “t” directory and can be run by launching “make” in this directory. Doing that you will see that there are a number of tests that are marked with “# TODO known breakage”, like for example:
“not ok 28 - git checkout -f: replace submodule with a directory must fail # TODO known breakage
These tests start with “test_expect_failure” instead of “test_expect_success”. They document that something is not working as it should perhaps be working. And it might be an interesting microproject to fix that.
Note that it is especially wise to first search the mailing list and then ask on the list before working on one of these “test_expect_failure”, because if we bothered to document a failure but not fix it, that is often because the fix is non-trivial.
You could also check if some commands have no test for some of their options and it could be an interesting microproject to add a test for one of those options.
You can search the mailing list for words like “low hanging fruit”, or “low-hanging fruits”, “hint, hint”, “later”, “we should”, “I plan to”…