| layout | title | navbar |
|---|---|---|
default |
SoC 2016 Ideas |
false |
This is the idea page for Summer of Code 2016 for Git and libgit2.
Please read this section completely before reading the idea list below.
It is required that students who want to apply to the Git project for the Summer of Code 2016 complete a tiny, code-related "microproject" as part of their application. Please refer to our guidelines and suggestions for microprojects for more information. Completing a microproject is not only an important way for us to get experience with applicants, but it will also help applicants become familiar with Git's development and submission process.
A complete GSoC application should include a presentation of yourself (include any argument that may convince mentors that you are able to complete the project) and detailed explanations about your project. Ideas below are just ... ideas! The list is not exhaustive, and more importantly each idea only includes a summary of what is to be done. An application must include detailed plans on the design, timeline ... A typical application takes several pages. You should already have read the GSoC Student Guide by now, but re-read it if needed.
Also, working in Git project is not only about writing your own patches. Constructively critiquing design and implementation of patches by other people is also an important skill you need to learn in order to effectively collaborate with others. So, if you have time and inclination, it would be beneficial to read and understand other applicants' patches (or any other patch submitted to the mailing-list), think if you agree that the problem they are trying to solve is worth solving, the approach they are taking is the best way (or if you think of a better way to solve it), etc., and respond to their patches with the result of your thinking as a review.
Please, include link(s) to the mailing-list discussion(s) related to your microproject in your application (e.g. linking to gmane). If you participate in the review of other patches, then you may also include links to discussions that would support your application.
Students must send drafts of their proposal on the mailing-list before submitting it officially to GSoC to get feedback from the community. They are strongly encourraged to publish a draft on the official GSoC website and post it to the mailing list for discussion.
Getting your proposal right can follow the same process as usual patch
submission for Git, as described in the
microprojects page and
in Documentation/SubmittingPatches in Git's source code. It is also
expected that you will send several versions of your draft, responding
to comments on the list. If you are not sure about your proposal, you
can discuss that in the same email where you introduce yourself or in
separate emails. Please use "[GSoC]" at the beginning of such emails.
In summary, all applicants must (not necessarily in this order):
-
Complete a microproject.
-
Write a detailed application explaining their project.
-
Discuss their project by posting drafts of their application on the mailing-list long before the deadline.
In your application, it is a good idea to:
-
Include link(s) to the mailing-list discussion(s) related to the project you chose in your application, for example previous discussions or patch series about the topic. There might be interesting discussions about the topics that are several year old. It is also a good idea to summarize them.
-
Include link(s) to the mailing-list discussion(s) related to the previous drafts of your application itself.
-
Include link(s) to the mailing-list discussion(s) related to your microproject.
-
Include what is suggested in the GSoC Student Guide
(gmane can be used for searching the mailing list and linking to previous discussions.)
In 2016, the Git organization has very limited mentoring capacity. These days we usually accept between 0 and 2 students per year.
Students: Please consider these ideas as starting points for generating proposals. We are also more than happy to receive proposals for other ideas related to Git or libgit2. For libgit2, see the bottom of the list and the libgit2 list of projects.
In addition to what we discussed, I think files named by include.path should be honored without checking the ownership. Cf. http://article.gmane.org/gmane.comp.version-control.git/243628
Takes eol attributes into account to correct lossage/addition of CR at the end of line, with updates to "am". Cf. http://article.gmane.org/gmane.comp.version-control.git/257445
- Language: C, shell (bash)
- Difficulty: medium
- Possible mentors: Matthieu Moy, Christian Couder
When your project is strictly "new features are merged into trunk, never the other way around", it is handy to be able to first find a merge on the trunk that merged a topic to point fingers at when a bug appears, instead of having to drill down to the individual commit on the faulty side branch. Cf. http://thread.gmane.org/gmane.comp.version-control.git/264661/focus=264720
"git bisect" is initially meant to find regressions (i.e. the new code is bad, and the old one used to work). The opposite may happen too: look for the commit which fixed a bug. This is why "git bisect terms" was added in Git 2.7. Still, when starting the bisection, the user has to know which behavior is the "new" one and which is the "old" one. This may not be obvious when comparing two commits which have no direct ancestry relation.
For example, commit A is 'fast', commit B is 'slow'. If neither commit is an ancestor of the other, then Git will test the common ancestor (say, C) of A and B. If C is 'fast' then we now know that we're looking for a transition from 'fast' to 'slow' between C and B. If C is 'slow', then we are looking for a transition from 'slow' to 'fast' between C and A. Currently, in this situation, Git will stop with an error message if 'fast' is the new term, 'slow' the old one, and if C is 'fast'. Instead of failing, it could offer the user to reverse the terms and continue the bisection.
See discussion on the subject here.
The change is controversial, hence a substantial part of the project would be to define what is the right behavior (should it be activated by default? what should be configurable and how? ...). Technically, the project is probably less than what would be expected for a GSoC hence it should be grouped with another one (typically another bisect-related idea).
"git-bisect.sh" is already using git bisect--helper for some
functionality. More features could be moved one by one into git bisect--helper by porting them from shell to C until both
"git-bisect.sh" and "builtin/bisect--helper.c" could be fully replaced
by "builtin/bisect.c".
Since 84e6fb9 (status: give more information during rebase -i, 2015-07-06), "git status" gives some useful information like previously applied and next commits when an interactive rebase is stopped. The same could be done for non-interactive rebase.
This is a small project that can be a used as a warm-up (perhaps: "extended warm-up"?) before tackling another one.
Some pieces of code are currently duplicated between wt-status.c and
git-rebase--interactive.sh (read_rebase_todolist/abbrev_sha1_in_line
in C and expand_todo_ids/collapse_todo_ids in shell). It would be nice
to let the shell version use directly the C version through a C helper
to reduce code duplication and increase consistency (we've already
been hit by subtle difference between different versions of a
git-rebase-todo file ...).
This would need:
-
libification of read_rebase_todolist/abbrev_sha1_in_line
-
Introduction of a new internal command like
git-rebase--helper(seebisect--helperfor a prior example) that call it -
Use it from
git-rebase--interactive.sh
This would open the door to other uses of C within "rebase -i", and possibly a step-by-step port of the shell code to C in the long run.
This is a relatively small project that should be combined with another one to give a proper GSoC project.
"git config", when removing the last variable in a section, leaves an empty section header behind. Anybody who wants to improve this needs to consider ramifications of leaving or removing comments. Cf. http://article.gmane.org/gmane.comp.version-control.git/219524
Cf. http://article.gmane.org/gmane.comp.version-control.git/235874
When merging a commit which has tracked files with the same name as local untracked files, Git refuses to proceed. It could be nice to:
-
Accept the situation without conflict when the tracked file has the exact same content as the local untracked file (which would become tracked). No data is lost, nothing can be committed accidentally.
-
Possibly, for fast-forward merges, if a local files belongs to the index but not to the last commit, attempt a merge between the upstream version and the local one (resulting in the same content as if the file had just been committed, but without introducing an extra commit).
Recent versions SVN do something similar: on update, it considers added but not committed files like normal tracked files, and attempts a merge of the upstream version with the local one (which always succeeds when the files have identical content). Attempting a merge for non-fast forward cases would probably not make sense: it would mix changes coming from the merge with other changes that do not come from a commit.
One difficulty with this project is to identify uncontroversial cases (where Git should merge without complaining) and all the possible corner-cases.
Many components of Git are still in the form of shell and Perl scripts. While this is an excellent choice as long as the functionality is improved, it causes problems in production code – in particular on multiple platforms, e.g. Windows (think: POSIX-to-Windows path conversion issues).
The idea of this project is to dive into the Git source code and convert a couple of shell and/or Perl scripts into portable and performant C code, making it a so-called "built-in".
(Un)fortunately, the easy-to-port scripts like git-pull.sh are
already ported by now. It is still possible to start with something
small by porting portions of existing shell-scripts to C using a C
helper inside the existing shell-script.
Libgit2 has an implementation of pack-objects (copied from git) which does support multi-threading but not some of the other capabilities which the git implementation does like re-using deltas or copying compressed data from one packfile to another.
This would involve looking at the code in git to copy over optimisations as well as figuring out what parts of libgit2 should be changed to accomodate these new capabilities.
- Language: C
- Difficulty: medium
- Possible mentors: Carlos Martín / Ed Thomson
Libgit2 has support for the client side of the negotiation, but it's
missing server-side capabilities. We wouldn't want to simply
reimplement upload-pack or receive-pack as function calls, but
instead create the framework that takes care of the protocol details
and calls to user code for
- pushing bytes to and from the network
- deciding which references to advertise
- deciding whether an update is acceptable
- possibly more
which would allow e.g. limiting which references are shown to a particular user or make decisions about updates in callbacks instead of script hooks.
- Language: C
- Difficulty: medium
- Possible mentors: Carlos Martín / Ed Thomson
Git is an incredible powerful tool with lots of different commands that offer a variety of ways to approach source control management. Naturally every way has advantages and disadvantages which a seasoned user can carefully consider and out weight against each other. However, many new users of Git are unable to cope with this variety, at least initially. If they run into a problem they likely search the Internet and find a StackOverflow answer instructing them to run a certain Git commands to solve their problems. A rushed user (aren't we all?) might run these commands without reading the docs which might makes the problem worse.
The core of this project is to evaluate with a running prototype if it is possible to implement a "Git Beginner Mode". The mode shall be activated with the config "core.isbeginner" by Git users who prefer this safety net (default should be false).
If this mode is enabled then Git shall print a warning message before running a potentially destructive command. In addition to the warning Git shall print a command that would reverse the operation if possible. Most of the information to reverse an operation is already available via git reflog. However, the task here is to make this information more easily accessible to Git beginners.
The following commands should be guarded with this mechanism:
git rebase
git reset --hard
git clean -f
git gc --prune=now --aggressive
git push -f
This list can and should be extended by the student.
Note that this project is not technically difficult, it requires a deep understanding of Git: how each command is meant to be used, what are the potential dangers, ... Reaching a solution that effectively protects beginners without harming anyone is much harder than it seems. See for example this thread for example potential objections. If chosen, this project should be discussed in depth on the list before and after the student application.
- Language: C, shell (bash)
- Difficulty: hard
- Possible mentors: Lars Schneider
Git beginners are easily confused by the distributed nature of Git. One source of confusion are Git remotes, especially if there are multiple ones. This is a potentially big threat to cooperations as Git beginners might push changes to a public remote such as github.com instead of the private company Git server.
This project is about to implement a Git remote whitelist and blacklist using Git config.
Whitelist example:
[remote]
default = deny
message = "Are you sure you're not pushing company code?"
allowed = http://whitelisted-hosting.org
allowed = http://git-hosting.org/whitelisted-org
allowed = http://git-hosting.org/org/whitelisted-repo
Blacklist example:
[remote]
default = allow
denied = http://denied-hosting.com
If a user wants to push changes to a blacklisted remote then the push
command would print a generic error. If a remote.message is defined
then this message would be shown in addition.
- Language: C, shell (bash)
- Difficulty: medium
- Possible mentors: Lars Schneider