21.5 KB
Newer Older
## GitLab core team & GitLab Inc. contribution process
2 3 4


5 6 7 8
<!-- START doctoc generated TOC please keep comment here to allow auto update -->
**Table of Contents**  *generated with [DocToc](*

9 10 11 12 13 14
- [Purpose of describing the contributing process](#purpose-of-describing-the-contributing-process)
- [Common actions](#common-actions)
  - [Merge request coaching](#merge-request-coaching)
- [Assigning issues](#assigning-issues)
- [Be kind](#be-kind)
- [Feature freeze on the 7th for the release on the 22nd](#feature-freeze-on-the-7th-for-the-release-on-the-22nd)
  - [Feature flags](#feature-flags)
  - [Between the 1st and the 7th](#between-the-1st-and-the-7th)
    - [What happens if these deadlines are missed?](#what-happens-if-these-deadlines-are-missed)
  - [On the 7th](#on-the-7th)
19 20
    - [Feature merge requests](#feature-merge-requests)
    - [Documentation merge requests](#documentation-merge-requests)
  - [After the 7th](#after-the-7th)
  - [Asking for an exception](#asking-for-an-exception)
Mek Stittri's avatar
Mek Stittri committed
- [Bugs](#bugs)
  - [Regressions](#regressions)
Mek Stittri's avatar
Mek Stittri committed
  - [Managing bugs](#managing-bugs)
26 27 28 29 30 31 32 33 34 35
- [Release retrospective and kickoff](#release-retrospective-and-kickoff)
- [Copy & paste responses](#copy--paste-responses)
  - [Improperly formatted issue](#improperly-formatted-issue)
  - [Issue report for old version](#issue-report-for-old-version)
  - [Support requests and configuration questions](#support-requests-and-configuration-questions)
  - [Code format](#code-format)
  - [Issue fixed in newer version](#issue-fixed-in-newer-version)
  - [Improperly formatted merge request](#improperly-formatted-merge-request)
  - [Accepting merge requests](#accepting-merge-requests)
  - [Only accepting merge requests with green tests](#only-accepting-merge-requests-with-green-tests)
36 37 38

<!-- END doctoc generated TOC please keep comment here to allow auto update -->

40 41 42

## Purpose of describing the contributing process

43 44 45 46 47 48
Below we describe the contributing process to GitLab for two reasons:

1. Contributors know what to expect from maintainers (possible responses, friendly
  treatment, etc.)
1. Maintainers know what to expect from contributors (use the latest version,
  ensure that the issue is addressed, friendly treatment, etc.).

50 51
- [GitLab Inc engineers should refer to the engineering workflow document](

## Common actions

### Merge request coaching

Several people from the [GitLab team][team] are helping community members to get
their contributions accepted by meeting our [Definition of done][done].
Mark Pundsack's avatar
Mark Pundsack committed

Mark Lapierre's avatar
Mark Lapierre committed
What you can expect from them is described at

61 62 63 64 65 66 67 68 69 70 71 72
### Milestones on community contribution issues

The milestone of an issue that is currently being worked on by a community contributor
should not be set to a named GitLab milestone (e.g. 11.7, 11.8), until the associated
merge request is very close to being merged, and we will likely know in which named
GitLab milestone the issue will land. There are many factors that influence when
a community contributor finishes an issue, or even at all. So we should set this 
milestone only when we have more certainty.

Note this only applies to issues currently assigned to community contributors. For
issues assigned to GitLabbers, we are [ambitious in assigning milestones to issues](

73 74 75 76
## Assigning issues

If an issue is complex and needs the attention of a specific person, assignment is a good option but assigning issues might discourage other people from contributing to that issue. We need all the contributions we can get so this should never be discouraged. Also, an assigned person might not have time for a few weeks, so others should feel free to takeover.

77 78
## Be kind

79 80 81
Be kind to people trying to contribute. Be aware that people may be a non-native
English speaker, they might not understand things or they might be very
sensitive as to how you word things. Use Emoji to express your feelings (heart,
82 83 84 85
star, smile, etc.). Some good tips about code reviews can be found in our
[Code Review Guidelines].

[Code Review Guidelines]:

## Feature freeze on the 7th for the release on the 22nd

89 90 91 92 93
After 7th at 23:59 (Pacific Time Zone) of each month, stable branch and RC1 
of the upcoming release (to be shipped on the 22nd) is created and deployed to 
The stable branch is frozen at the most recent "qualifying commit" on master.
A "qualifying commit" is one that is pushed before the feature freeze cutoff time
and that passes all CI jobs (green pipeline).
94 95

Merge requests may still be merged into master during this
96 97
period, but they will go into the _next_ release, unless they are manually
cherry-picked into the stable branch.

99 100
By freezing the stable branches 2 weeks prior to a release, we reduce the risk
of a last minute merge request potentially breaking things.

102 103 104 105 106 107 108 109 110
Any release candidate that gets created after this date can become a final
release, hence the name release candidate.

### Feature flags

Merge requests that make changes hidden behind a feature flag, or remove an
existing feature flag because a feature is deemed stable, may be merged (and
picked into the stable branches) up to the 19th of the month. Such merge
requests should have the ~"feature flag" label assigned, and don't require a
111 112 113 114 115
corresponding exception request to be created. 

A level of common sense should be applied when deciding whether to have a feature 
behind a feature flag off or on by default.

The following guideliness can be applied to help make this decision:

118 119 120 121 122 123
* If the feature is not fully ready or functioning, the feature flag should be disabled by default.
* If the feature is ready but there are concerns about performance or impact, the feature flag should be enabled by default, but 
disabled via chatops before deployment on environments. If the performance concern is confirmed, the final release should have the feature flag disabled by default.
* In most other cases, the feature flag can be enabled by default.

For more information on rolling out changes using feature flags, read [through the documentation](

125 126 127 128 129 130
In order to build the final package and present the feature for self-hosted
customers, the feature flag should be removed. This should happen before the
22nd, ideally _at least_ 2 days before. That means MRs with feature
flags being picked at the 19th would have a quite tight schedule, so picking
these _earlier_ is preferable.

131 132 133 134
While rare, release managers may decide to reject picking a change into a stable
branch, even when feature flags are used. This might be necessary if the changes
are deemed problematic, too invasive, or there simply isn't enough time to
properly test how the changes behave on

136 137
### Between the 1st and the 7th

These types of merge requests for the upcoming release need special consideration:
139 140 141 142 143 144

* **Large features**: a large feature is one that is highlighted in the kick-off
  and the release blogpost; typically this will have its own channel in Slack
  and a dedicated team with front-end, back-end, and UX.
* **Small features**: any other feature request.

145 146
It is strongly recommended that **large features** be with a maintainer **by the
1st**. This means that:
147 148 149 150 151 152 153 154 155 156 157 158 159

* There is a merge request (even if it's WIP).
* The person (or people, if it needs a frontend and backend maintainer) who will
  ultimately be responsible for merging this have been pinged on the MR.

It's OK if merge request isn't completely done, but this allows the maintainer
enough time to make the decision about whether this can make it in before the
freeze. If the maintainer doesn't think it will make it, they should inform the
developers working on it and the Product Manager responsible for the feature.

The maintainer can also choose to assign a reviewer to perform an initial
review, but this way the maintainer is unlikely to be surprised by receiving an
MR later in the cycle.

161 162
It is strongly recommended that **small features** be with a reviewer (not
necessarily a maintainer) **by the 3rd**.
163 164 165 166 167 168

Most merge requests from the community do not have a specific release
target. However, if one does and falls into either of the above categories, it's
the reviewer's responsibility to manage the above communication and assignment
on behalf of the community member.

Marcia Ramos's avatar
Marcia Ramos committed
169 170
Every new feature or change should be shipped with its corresponding documentation
in accordance with the
[documentation process](
172 173
and [structure]( guides.
Note that a technical writer will review all changes to documentation. This can occur
174 175 176
in the same MR as the feature code, but [if there is not sufficient time or need,
it can be planned via a follow-up issue for doc review](,
and another MR, if needed. Regardless, complete docs must be merged with code by the freeze.
Marcia Ramos's avatar
Marcia Ramos committed

178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200
#### What happens if these deadlines are missed?

If a small or large feature is _not_ with a maintainer or reviewer by the
recommended date, this does _not_ mean that maintainers or reviewers will refuse
to review or merge it, or that the feature will definitely not make it in before
the feature freeze.

However, with every day that passes without review, it will become more likely
that the feature will slip, because maintainers and reviewers may not have
enough time to do a thorough review, and developers may not have enough time to
adequately address any feedback that may come back.

A maintainer or reviewer may also determine that it will not be possible to
finish the current scope of the feature in time, but that it is possible to
reduce the scope so that something can still ship this month, with the remaining
scope moving to the next release. The sooner this decision is made, in
conversation with the Product Manager and developer, the more time there is to
extract that which is now out of scope, and to finish that which remains in scope.

For these reasons, it is strongly recommended to follow the guidelines above,
to maximize the chances of your feature making it in before the feature freeze,
and to prevent any last minute surprises.

201 202
### On the 7th

Marcia Ramos's avatar
Marcia Ramos committed
Merge requests should still be complete, following the [definition of done][done].

Sean McGivern's avatar
Sean McGivern committed
205 206 207 208 209 210
If a merge request is not ready, but the developers and Product Manager
responsible for the feature think it is essential that it is in the release,
they can [ask for an exception](#asking-for-an-exception) in advance. This is
preferable to merging something that we are not confident in, but should still
be a rare case: most features can be allowed to slip a release.

211 212 213 214 215
All Community Edition merge requests from GitLab team members merged on the
freeze date (the 7th) should have a corresponding Enterprise Edition merge
request, even if there are no conflicts. This is to reduce the size of the
subsequent EE merge, as we often merge a lot to CE on the release date. For more
information, see
[Automatic CE->EE merge][automatic_ce_ee_merge] and
[Guidelines for implementing Enterprise Edition features][ee_features].

### After the 7th

221 222 223
Once the stable branch is frozen, the only MRs that can be cherry-picked into
the stable branch are:

* Fixes for [regressions](#regressions) where the affected version `xx.x` in `regression:xx.x` is the current release. See [Managing bugs](#managing-bugs) section.
225 226 227 228 229
* Fixes for security issues.
* Fixes or improvements to automated QA scenarios.
* [Documentation improvements]( for feature changes made in the same release, though initial docs for these features should have already been merged by the freeze, as required.
* New or updated translations (as long as they do not touch application code).
* Changes that are behind a feature flag and have the ~"feature flag" label.

231 232 233 234 235 236 237 238 239 240 241 242 243 244
During the feature freeze all merge requests that are meant to go into the
upcoming release should have the correct milestone assigned _and_ the
`Pick into X.Y` label where `X.Y` is equal to the milestone, so that release
managers can find and pick them.
Merge requests without this label will not be picked into the stable release.

For example, if the upcoming release is `10.2.0` you will need to set the
`Pick into 10.2` label.

Fixes marked like this will be shipped in the next RC (before the 22nd), or the
next patch release.

If a merge request is to be picked into more than one release it will need one
`Pick into X.Y` label per release where the merge request should be back-ported
to. For example:
246 247 248 249

- `Pick into 10.1`
- `Pick into 10.0`
- `Pick into 9.5`

Sean McGivern's avatar
Sean McGivern committed
251 252
### Asking for an exception

If you think a merge request should go into an RC or patch even though it does not meet these requirements,
you can ask for an exception to be made.

Check [this guide]( about how to open an exception request before opening one.

Mek Stittri's avatar
Mek Stittri committed
## Bugs

Victor Wu's avatar
Victor Wu committed
A ~bug is a defect, error, failure which causes the system to behave incorrectly or prevents it from fulfilling the product requirements.

262 263
The level of impact of a ~bug can vary from blocking a whole functionality
or a feature usability bug. A bug should always be linked to a severity level.
Refer to our [severity levels](

Whether the bug is also a regression or not, the triage process should start as soon as possible.
Mek Stittri's avatar
Mek Stittri committed
Ensure that the Engineering Manager and/or the Product Manager for the relative area is involved to prioritize the work as needed.

Mek Stittri's avatar
Mek Stittri committed
### Regressions

271 272
A ~regression implies that a previously **verified working functionality** no longer works.
Regressions are a subset of bugs. We use the ~regression label to imply that the defect caused the functionality to regress.
The label tells us that something worked before and it needs extra attention from Engineering and Product Managers to schedule/reschedule.

275 276
The regression label does not apply to ~bugs for new features for which functionality was **never verified as working**.
These, by definition, are not regressions.

A regression should always have the `regression:xx.x` label on it to designate when it was introduced.

Regressions should be considered high priority issues that should be solved as soon as possible, especially if they have severe impact on users.

Mek Stittri's avatar
Mek Stittri committed
282 283
### Managing bugs

284 285
**Prioritization:** We give higher priority to regressions on features that worked in the last recent monthly release and the current release candidates.
The two scenarios below can [bypass the exception request in the release process](, where the affected regression version matches the current monthly release version.
Mek Stittri's avatar
Mek Stittri committed
286 287
* A regression which worked in the **Last monthly release**
   * **Example:** In 11.0 we released a new `feature X` that is verified as working. Then in release 11.1 the feature no longer works, this is regression for 11.1. The issue should have the `regression:11.1` label.
Mek Stittri's avatar
Mek Stittri committed
   * *Note:* When we say `the last recent monthly release`, this can refer to either the version currently running on, or the most recent version available in the package repositories.
Mek Stittri's avatar
Mek Stittri committed
289 290
* A regression which worked in the **Current release candidates**
   * **Example:** In 11.1-RC3 we shipped a new feature which has been verified as working. Then in 11.1-RC5 the feature no longer works, this is regression for 11.1. The issue should have the `regression:11.1` label.
   * *Note:* Because runs release candidates of new releases, a regression can be reported in a release before its 'official' release date on the 22nd of the month.
Fabio Busatto's avatar
Fabio Busatto committed

Mek Stittri's avatar
Mek Stittri committed
293 294 295
When a bug is found:
1. Create an issue describing the problem in the most detailed way possible.
1. If possible, provide links to real examples and how to reproduce the problem.
296 297
1. Label the issue properly, using the [team label](,
   the [subject label](
Fabio Busatto's avatar
Fabio Busatto committed
   and any other label that may apply in the specific case
1. Notify the respective Engineering Manager to evaluate and apply the [Severity label]( and [Priority label](
Mek Stittri's avatar
Mek Stittri committed
300 301
The counterpart Product Manager is included to weigh-in on prioritization as needed.
1. If the ~bug is **NOT** a regression:
   1. The Engineering Manager decides which milestone the bug will be fixed. The appropriate milestone is applied.
1. If the bug is a ~regression:
304 305
   1. Determine the release that the regression affects and add the corresponding `regression:xx.x` label.
      1. If the affected release version can't be determined, add the generic ~regression label for the time being.
306 307
   1. If the affected version `xx.x` in `regression:xx.x` is the **current release**, it's recommended to schedule the fix for the current milestone.
      1. This falls under regressions which worked in the last release and the current RCs. More detailed explanations in the **Prioritization** section above.
   1. If the affected version `xx.x` in `regression:xx.x` is older than the **current release**
309 310
      1. If the regression is an ~S1 severity, it's recommended to schedule the fix for the current milestone. We would like to fix the highest severity regression as soon as we can.
      1. If the regression is an ~S2, ~S3 or ~S4 severity, the regression may be scheduled for later milestones at the discretion of the Engineering Manager and Product Manager.
Mek Stittri's avatar
Mek Stittri committed

312 313
## Release retrospective and kickoff

Victor Wu's avatar
Victor Wu committed
314 315
- [Retrospective](
- [Kickoff](

317 318 319 320
## Copy & paste responses

### Improperly formatted issue

321 322 323 324
Thanks for the issue report. Please reformat your issue to conform to the
[contributing guidelines](
325 326 327

### Issue report for old version

328 329 330 331 332 333 334
Thanks for the issue report but we only support issues for the latest stable version of GitLab.
I'm closing this issue but if you still experience this problem in the latest stable version,
please open a new issue (but also reference the old issue(s)).
Make sure to also include the necessary debugging information conforming to the issue tracker
guidelines found in our [contributing guidelines](
335 336 337

### Support requests and configuration questions

339 340
Thanks for your interest in GitLab. We don't use the issue tracker for support
requests and configuration questions. Please check our
David's avatar
David committed
[getting help]( page to see all of the available
support options. Also, have a look at the [contribution guidelines](
for more information.
345 346 347

### Code format

David's avatar
David committed
Please use \`\`\` to format console output, logs, and code as it's very hard to read otherwise.
351 352 353

### Issue fixed in newer version

354 355 356 357 358 359 360 361 362
Thanks for the issue report. This issue has already been fixed in newer versions of GitLab.
Due to the size of this project and our limited resources we are only able to support the
latest stable release as outlined in our [contributing guidelines](
In order to get this bug fix and enjoy many new features please
If you still experience issues at that time please open a new issue following our issue
tracker guidelines found in the [contributing guidelines](

### Improperly formatted merge request

366 367 368 369
Thanks for your interest in improving the GitLab codebase!
Please update your merge request according to the [contributing guidelines](

371 372
### Accepting merge requests

Is there an issue on the
David's avatar
David committed
[issue tracker]( that is
similar to this? Could you please link it here?
Please be aware that new functionality that is not marked
[`Accepting merge requests`](
might not make it into GitLab.
381 382 383

### Only accepting merge requests with green tests

Jeroen van Baarsen's avatar
Jeroen van Baarsen committed
385 386 387 388
We can only accept a merge request if all the tests are green. I've just
restarted the build. When the tests are still not passing after this restart and
you're sure that is does not have anything to do with your code changes, please
rebase with master to see if that solves the issue.
390 391

393 394