Django uses Trac for managing the work on the code base. 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!
Unfortunately, not all bug reports and feature requests in the ticket tracker provide all the required details. A number of tickets have patches, but those patches don’t meet all the requirements of a good patch.
One way to help out is to triage tickets that have been created by other users.
Most of the workflow is based around the concept of a ticket’s triage stages. Each stage describes where in its lifetime a given ticket is at any time. Along with a handful of flags, this attribute easily tells us what and who each ticket is waiting on.
Since a picture is worth a thousand words, let’s start there:
We’ve got two roles in this diagram:
By way of example, here we see the lifecycle of an average ticket:
Some tickets require much less feedback than this, but then again some tickets require much much more.
Below we describe in more detail the various stages that a ticket may flow through during its lifetime.
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.
The big gray 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 it yet. Often this means you could safely start writing a patch for it. This is generally more true for the case of accepted bugs than accepted features. A ticket for a bug that has been accepted means that the issue has been verified by at least one triager as a legitimate bug - and should probably be fixed if possible. An accepted new feature may only mean that one triager thought the feature would be good to have, but this alone does not represent a consensus view or imply with any certainty that a patch will be accepted for that feature. Seek more feedback before writing an extensive patch if you are in doubt.
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 + Needs …
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.
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 committer now needs to give the patch a final review prior to being committed. See the New contributors’ FAQ for “My ticket has been in RFC forever! What should I do?”
This stage isn’t shown on the diagram. It’s used sparingly to keep track of high-level ideas or long term feature requests.
These tickets are uncommon and overall less useful since they don’t describe concrete actionable issues. They are enhancement requests that we might consider adding someday to the framework if an excellent patch is submitted. They are not a high priority.
A number of flags, appearing as checkboxes in Trac, can be set on a ticket:
This means the ticket has an associated patch. These will be reviewed to see if the patch is “good”.
The following three fields (Needs documentation, Needs tests, Patch needs improvement) apply only if a patch has been supplied.
This flag is used for tickets with patches that need associated documentation. Complete documentation of features is a prerequisite before we can check them into the codebase.
This flags the patch as needing associated unit tests. Again, this is a required part of a valid patch.
This flag means that although the ticket has a patch, it’s not quite ready for checkin. This could mean the patch no longer applies cleanly, there is a flaw in the implementation, or that the code doesn’t meet our standards.
Tickets that would require small, easy, patches.
Tickets should be categorized by type between:
Tickets should be classified into components indicating which area of the Django codebase they belong to. This makes tickets better organized and easier to find.
The severity attribute is used to identify blockers, that is, issues which should get fixed before releasing the next version of Django. Typically those issues are bugs causing regressions from earlier versions or potentially causing severe data losses. This attribute is quite rarely used and the vast majority of tickets have a severity of “Normal”.
It is possible to use the version attribute to indicate in which version the reported bug was identified.
This flag is used for tickets that relate to User Interface and User Experiences questions. For example, this flag would be appropriate for user-facing features in forms or the admin interface.
You may add your username or email address to this field to be notified when new contributions are made to the ticket.
With this field you may label a ticket with multiple keywords. This can be useful, for example, to group several tickets of a same theme. Keywords can either be comma or space separated. Keyword search finds the keyword string anywhere in the keywords. For example, clicking on a ticket with the keyword “form” will yield similar tickets tagged with keywords containing strings such as “formset”, “modelformset”, and “ManagementForm”.
When a ticket has completed its useful lifecycle, it’s time for it to be closed. Closing a ticket is a big responsibility, though. You have to be sure that the issue is really resolved, and you need to keep in mind that the reporter of the ticket may not be happy to have their ticket closed (unless it’s fixed, of course). If you’re not certain about closing a ticket, just leave a comment with your thoughts instead.
If you do close a ticket, you should always make sure of the following:
A ticket can be resolved in a number of ways:
If you believe that the ticket was closed in error – because you’re still having the issue, or it’s popped up somewhere else, or the triagers have made a mistake – please reopen the ticket and provide further information. Again, please do not reopen tickets that have been marked as “wontfix” and bring the issue to django-developers instead.
The triage process is primarily driven by community members. Really, ANYONE can help.
To get involved, start by creating an account on Trac. If you have an account but have forgotten your password, you can reset it using the password reset page.
Then, you can help out by:
Note
The Reports page contains links to many useful Trac queries, including several that are useful for triaging tickets and reviewing patches as suggested above.
You can also find more Advice for new contributors.
However, we do ask the following of all general community members working in the ticket database:
A regression is a bug that’s present in some newer version of Django but not in an older one. An extremely helpful piece of information is the commit that introduced the regression. Knowing the commit that caused the change in behavior helps identify if the change was intentional or if it was an inadvertent side-effect. Here’s how you can determine this.
Begin by writing a regression test for Django’s test suite for the issue. For
example, we’ll pretend we’re debugging a regression in migrations. After you’ve
written the test and confirmed that it fails on the latest master, put it in a
separate file that you can run standalone. For our example, we’ll pretend we
created tests/migrations/test_regression.py
, which can be run with:
$ ./runtests.py migrations.test_regression
Next, we mark the current point in history as being “bad” since the test fails:
$ git bisect bad
You need to start by "git bisect start"
Do you want me to do it for you [Y/n]? y
Now, we need to find a point in git history before the regression was
introduced (i.e. a point where the test passes). Use something like
git checkout HEAD~100
to checkout an earlier revision (100 commits earlier,
in this case). Check if the test fails. If so, mark that point as “bad”
(git bisect bad
), then checkout an earlier revision and recheck. Once you
find a revision where your test passes, mark it as “good”:
$ git bisect good
Bisecting: X revisions left to test after this (roughly Y steps)
...
Now we’re ready for the fun part: using git bisect run
to automate the rest
of the process:
$ git bisect run tests/runtests.py migrations.test_regression
You should see git bisect
use a binary search to automatically checkout
revisions between the good and bad commits until it finds the first “bad”
commit where the test fails.
Now, report your results on the Trac ticket, and please include the regression test as an attachment. When someone writes a fix for the bug, they’ll already have your test as a starting point.
Jun 22, 2017