|Version 16 (modified by gabrielhurley, 4 years ago) (diff)|
Django is developed 100% by the community, and the more people that are actively involved in the code the better Django will be. We recognize that contributing to Django can be daunting at first and sometimes confusing even to veterans. While we have our official "Contributing to Django" documentation which spells out the technical details of triaging tickets and submitting patches, it leaves a lot of room for interpretation. This guide aims to offer more general advice on issues such as how to interpret the various stages and flags in Trac, and how new contributors can get started.
"The Spirit of Contributing"
Django uses Trac for managing our progress, and Trac is a community-tended garden of the bugs people have found and the features people would like to see added. As in any garden, sometimes there are weeds to be pulled and sometimes there are flowers and vegetables that need picking. We need your help to sort out one from the other, and in the end we all benefit together.
Like all gardens, we can aspire to perfection but in reality there's no such thing. Even in the most pristine garden there are still snails and insects. In a community garden there are also helpful people who--with the best of intentions--fertilize the weeds and poison the roses. It's the job of the community as a whole to self-manage, keep the problems to a minimum, and educate those coming into the community so that they can become valuable contributing members.
Similarly, while we aim for Trac to be a perfect representation of the state of Django's progress, we acknowledge that this simply will not happen. By distributing the load of Trac maintenance to the community, we accept that there will be mistakes. Trac is "mostly accurate", and we give allowances for the fact that sometimes it will be wrong. That's okay. We're perfectionists with deadlines.
We rely on the community to keep participating, keep tickets as accurate as possible, and raise issues for discussion on our mailing lists when there is confusion or disagreement.
Django is a community project, and every contribution helps. We can't do this without YOU!
What the various statuses "really" mean...
Django uses Trac as its sole official issue tracker. All known bugs, desired features, and ideas for changes are logged there.
However, Trac can be quite confusing even to veteran contributors. Having to look at both flags and statuses isn't immediately obvious, and the statuses themselves have names that can be misinterpreted.
What Django's statuses "really mean":
- Unreviewed: The ticket has not been reviewed by anyone who felt qualified to make a judgment about whether the ticket contained a valid issue, a viable feature, or ought to be closed for any of the various reasons.
- Accepted: The big grey area! The absolute meaning of "accepted" is that the issue described in the ticket is valid and is in some stage of being worked on. Beyond that there are several considerations:
- Accepted + No Flags: The ticket is valid, but no one has submitted a patch for yet. Often this means you could safely start writing a patch for it.
- Accepted + Has Patch: The ticket is waiting for people to review the supplied patch. This means downloading the patch and trying it out, verifying that it contains tests and docs, running the test suite with the included patch, and leaving feedback on the ticket.
- Accepted + Has Patch + (any other flag): This means the ticket has been reviewed, and has been found to need further work. "Needs tests" and "Needs documentation" are self-explanatory. "Patch needs improvement" will generally be accompanied by a comment on the ticket explaining what is needed to improve the code.
- Design Decision Needed: This status is for issues which may be contentious, may be backwards incompatible, or otherwise involve high-level design decisions. These decisions are generally made by the core committers, however that is not a requirement. See the FAQ below for "My ticket has been in DDN forever! What should I do?"
- Ready For Checkin: The ticket was reviewed by any member of the community other than the person who supplied the patch and found to meet all the requirements for a commit-ready patch. A core committer now needs to give the patch a final review prior to being committed. See the FAQ below for "My ticket has been in RFC forever! What should I do?"
- Someday/Maybe: Generally only used for vague/high-level features or design ideas. These tickets are uncommon and overall less useful since they don't describe concrete actionable issues.
- Fixed on a branch: Used to indicate that a ticket is resolved as part of a major body of work that will eventually be merged to trunk. Tickets in this stage generally don't need further work. This may happen in the case of major features/refactors in each release cycle, or as part of the annual Google Summer of Code efforts.
Here's an example workflow:
- Alice creates a ticket, and uploads an incomplete patch (no tests, incorrect implementation).
- Bob reviews the patch, marks it "Accepted", "needs tests", and "patch needs improvement", and leaves a comment telling Alice how the patch could be improved.
- Alice updates the patch, adding tests (but not changing the implementation). She removes the two flags.
- Charlie reviews the patch and resets the 'patch needs improvement flag' with another comment about improving the implementation.
- Alice updates the patch, fixing the implementation. She removes the "patch needs improvement" flag.
- Daisy reviews the patch, and marks it RFC.
- Jacob reviews the RFC patch, applies it to his checkout, and commits it.
Advice for new contributors
New contributor and not sure what to do? Want to help but just don't know how to get started? This is the section for you.
- Pick a subject area that you care about, that you are familiar with, or that you want to learn about. You don't already have to be an expert on the area you want to work on; you become an expert through your ongoing contributions to the code.
- Triage Tickets. If a ticket is unreviewed and reports a bug, try and duplicate it. If you can duplicate it and it seems valid, make a note that you confirmed the bug and accept the ticket. Make sure the ticket is filed under the correct component area. Consider writing a patch that adds a test for the bug's behavior, even if you don't fix the bug itself.
- Look for tickets that are accepted and review patches to build familiarity with the codebase and the process. Mark the appropriate flags if a patch needs docs or tests. Look through the changes a patch makes, and keep an eye out for syntax that is incompatible with older but still supported versions of Python. Run the tests and make sure they pass on your system. Where possible and relevant, try them out on a database other than SQLite. Leave comments and feedback!
- Keep old patches up to date. Oftentimes the codebase will change between a patch being submitted and the time it gets reviewed. Make sure it still applies cleanly and functions as expected. Simply updating a patch is both useful and important!
- Trac isn't an absolute; the context is just as important as the words. When reading Trac, you need to take into account who says things, and when they were said. Support for an idea two years ago doesn't necessarily mean that the idea will still have support. You also need to pay attention to who *hasn't* spoken -- for example, if a core team member hasn't been recently involved in a discussion, then a ticket may not have the support required to get into trunk.
- Start small. It's easier to get feedback on a little issue than on a big one.
- If you're going to engage in a big task, you should make sure that your idea has support first. This means getting someone else to confirm that a bug is real before you fix the issue, and ensuring that the core team supports a proposed feature before you go implementing it.
- Be bold! Leave feedback! Sometimes it can be scary to put your opinion out to the world and say "this ticket is correct" or "this patch needs work", but it's the only way the project moves forward. The contributions of the broad Django community ultimately have a much greater impact than that of the core developers. We can't do it without YOU!
- Err on the side of caution when marking things Ready For Check-in. If you're really not certain if a ticket is ready, don't mark it as such. Leave a comment instead, letting others know your thoughts. If you're mostly certain, but not completely certain, you might also try asking on IRC to see if someone else can confirm your suspicions.
- Wait for feedback, and respond to feedback that you receive. Focus on one or two tickets, see them through from start to finish, and repeat. The shotgun approach of taking on lots of tickets and letting some fall by the wayside ends up doing more harm than good.
- Be rigorous. When we say "PEP8, and must have docs and tests", we mean it. If a patch doesn't have docs and tests, there had better be a good reason. Arguments like "I couldn't find any existing tests of this feature" don't carry much weight--while it may be true, that means you have the extra-important job of writing the very first tests for that feature, not that you get a pass from writing tests altogether.
Questions that come up regularly, anything that's been overlooked, doesn't fit in its own section, etc....
- This ticket I care about has been ignored for days/weeks/months! What can I do to get it committed?
- First off, it's not personal. Django is entirely developed by volunteers (even the core devs), and sometimes folks just don't have time. The best thing to do is to send a gentle reminder to the Django Developers mailing list asking for review on the ticket, or to bring it up in the #django-dev IRC channel.
- I'm sure my ticket is perfect, can I mark it as RFC myself?
- Short answer: No. It's always better to get another set of eyes on a ticket. If you're having trouble getting that second set of eyes, see question 1, above.
- My ticket has been in DDN forever! What should I do?
- Design Decision Needed requires consensus about the right solution. At the very least it needs consensus among the core developers, and ideally it has consensus from the community as well. The best way to accomplish this is to start a thread on the Django Developers mailing list, and for very complex issues to start a wiki page summarizing the problem and the possible solutions.