Introduction

It is strongly recommended that students who want to apply to the Git project for the Summer of Code 2018 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.

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 familar 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:

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. You can use “[GSoC][PATCH …]” in the subject of the emails you send for that purpose. 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!

Ideas for microprojects

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.

Add more builtin patterns for userdiff

“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, shell.

This project requires a very good knowledge of regular expressions.

Make “git tag --contains <id>” less chatty if <id> is invalid

git tag --contains <id> prints the whole help text if <id> is invalid. It should only show the error message instead. [thread]

Git CI Improvements 1

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

Git CI Improvements 2

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.

Git CI Improvements 3

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.

Git CI Improvements 4

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 CI Improvements 5

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 tests 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.

See the commit c6f44e1da5 for example.

Use unsigned integral type for collection of bits.

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. [thread]

Move ~/.git-credential-cache to ~/.cache/git

Most of git dotfiles can be located, at the user’s option, in ~/.<file> or in ~/.config/git/<file>, following the XDG standard. ~/.git-credential-cache is still hardcoded as ~/.git-credential-cache, but should allow using the XDG directory layout too as ~/.cache/git/credential, possibly modified by $XDG_CONFIG_HOME and $XDG_CACHE_HOME).

The suggested approach is:

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.

Add configuration options for some commonly used command-line options

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.

Use dir-iterator to avoid explicit recursive directory traversal

Some places in git use raw API opendir/readdir/closedir to traverse a directory recursively, which usually involves function recursion. Now that we have struct dir_iterator (see dir-iterator.h), convert these to use the dir-iterator to simplify the code. Do only one conversion per microproject.

How to find other ideas for microprojects

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 (https://public-inbox.org/git/) 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 innappropriate.

Searching for #leftoverbits in the mailing list

People have recently started to add “#leftoverbits” to their emails when they think further small work on the topic could be useful.

You can easily search that using:

https://public-inbox.org/git/?q=leftoverbits

But don’t forget to search to check if what you find has already been addressed.

Searching the code base itself

Your best bet is probably to search for strings like “FIXME”, “TODO”, “NEEDSWORK”, or maybe “NEED-WORK”, and “BUG”.

You can also search for common patterns in the code and try to find or create a function to refactor them.

Searching the tests

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.

Searching the mailing list

You can search the mailing list for words like “low hanging fruit”, or “low-hanging fruits”, “hint, hint”, “later”, “we should”, “I plan to”…