These need to run at about the same times, and each runs quickly.
Doesn't seem worthwhile to fire up a VM for each of them individually.
<!--
If you know who should review your pull request, please assign it to
that
person, otherwise the pull request would get assigned randomly.
If your pull request is for a specific language, please add the
appropriate
lang label.
-->
...
<!--
If you know who should review your pull request, please assign it to
that
person, otherwise the pull request would get assigned randomly.
If your pull request is for a specific language, please add the
appropriate
lang label.
-->
It check status on a pr, and when some set are successful it makes itself successful.
If any that it's watching fail, it fails.
If the force-merge label is applied, it succeeds.
With the plan: we make EasyCLA and probably Sanity, and this one required statuses for merging (i.e. anything that should not be bypassed still gets verified by GitHub)
We make every other status that we believe in (so everything but the Google3 presubmits because they're fresh?) something that this script verifies.
That lets us do automerge (all things are passing)
And if all things are not passing we can use human judgement and apply the force-merge label to bypass these checks... But importantly not the CLA check