In https://github.com/ciena-blueplanet/ciena-devops-testbed when using the CODEOWNERS entries of
* @notmessenger @juwara0 @job13er README.md @notmessenger @juwara0
and making a PR changing just the .THROWAWAY file all three users (@notmessenger @juwara0 @job13er) are added as reviewers and it only takes one approver for the PR to be mergeable (https://github.com/ciena-blueplanet/ciena-devops-testbed/pull/45)
When the .THROWAWAY file and the README.md file are changed it requires one of the three users (@notmessenger @juwara0 @job13er) as well as one of the two (@notmessenger @juwara0) to be mergeable (https://github.com/ciena-blueplanet/ciena-devops-testbed/pull/46)
This is exactly what I expected to happen per https://help.github.com/articles/about-codeowners/
However when I attempt to do similar via using teams things get all messed up and I’m not sure why.
I created the https://github.com/orgs/ciena-blueplanet/teams/reviewers-tooling/members team and added @notmessenger, @juwara0, and @job13er to it.
I created the https://github.com/orgs/ciena-blueplanet/teams/owners-architecture team and added @notmessenger and @juwara0 to it.
I then used these CODEOWNERS entries:
* @ciena-blueplanet/reviewers-tooling README.md @ciena-blueplanet/owners-architecture
and got this result for only the .THROWAWAY file changed (https://github.com/ciena-blueplanet/ciena-devops-testbed/pull/45):
and this result for the .THROWAWAY file and README.md file changed (https://github.com/ciena-blueplanet/ciena-devops-testbed/pull/46):
Have I set the teams and/or CODEOWNERS up correctly?
Sorry that I only have screenshots for the failed state - I am using the same repo to work through these issues so I made changes to the files for different scenarios I’m testing. It may even be likely that the current content of the CODEOWNERS file are different than what I’ve pasted here as I’m still making changes but the content pasted here is accurate and I will not be making any changes to the teams I have setup so they can be looked at.
You can ignore the double travis-ci required status checks - I know why there are two and it is unrelated to this issue.
I also now realize that my screen captures did not capture the reviewers that were assigned. For PR 45 it did not assign everyone in the https://github.com/orgs/ciena-blueplanet/teams/reviewers-tooling/members team as I expected, and like it does when individual users are listed. For PR 46 it assigned no one at all.