Merge ~ines-almeida/launchpad:invert-bug-webhook-feature-flag into launchpad:master
Status: | Merged |
---|---|
Approved by: | Ines Almeida |
Approved revision: | 34b6034b26009fc3b6155fbe350f11c69f88b884 |
Merge reported by: | Otto Co-Pilot |
Merged at revision: | not available |
Proposed branch: | ~ines-almeida/launchpad:invert-bug-webhook-feature-flag |
Merge into: | launchpad:master |
Diff against target: |
234 lines (+17/-30) 10 files modified
lib/lp/bugs/interfaces/bugtarget.py (+2/-2) lib/lp/bugs/subscribers/webhooks.py (+3/-3) lib/lp/bugs/tests/test_subscribers.py (+2/-2) lib/lp/registry/browser/distribution.py (+2/-2) lib/lp/registry/browser/distributionsourcepackage.py (+2/-2) lib/lp/registry/browser/product.py (+2/-2) lib/lp/services/features/flags.py (+3/-2) lib/lp/services/webhooks/tests/test_browser.py (+0/-4) lib/lp/soyuz/stories/distribution/xx-distribution-packages.rst (+1/-0) lib/lp/soyuz/tests/test_packagecopyjob.py (+0/-11) |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Guruprasad | Approve | ||
Review via email: mp+445657@code.launchpad.net |
Commit message
Invert bug webhooks feature flag to be enabled by default
For feature flags that we intend to keep enabled (and particularly for the unit tests to run with the newly added feature) this change is inverting this feature flag so that the feature is enabled by default. To disable the feature, one just needs to set the "bugs.webhooks.
Description of the change
I haven't run the full test suite in this (I ran enough to feel confident to MP it), so we might discover a couple extra places that are missing permissions when all the full test suite runs after merge.
IMO, there might still be value in setting feature flags globally for testing, and I opened a ticket to potentially address that in the future (https:/
Inês, here is what our documentation on Feature Flags says about the type of change made in this MP
> The effect is typically 'enabled', 'visible', or 'timeout'. These should always be in the positive sense, not 'disabled'. If timeouts are given, they should be in seconds (decimals can be given in the value.)
I agree with this and also believe that the feature flags should be opt-in and not opt-out by default. Since your intent is to enable it globally to catch all permission issues, we should try to make the relevant test suite changes to that effect and not toggle the default behaviour of a feature flag, imho.