Change CODEOWNERS to use teams rather than individuals

[mkavulich]

Because our development workflow requires approvals representing each host model rather than individuals, we are setting up teams to allow for pings/reviews/approvals to be tied to these teams rather than individuals

User interface changes?: No

Testing: None, Github interface changes only

[gold2718]

When I look at the diffs to review, I see an error message that none of the new groups is recognized. This triggered a couple of questions because I do not know how this works.

• When a new PR is open under this system, who received a review request? Is it everyone in the group (i.e, has this been tested in other contexts)?
• Is there any way for us peons to see who is in which group? Under the repo settings, I see people (collaborators) but no groups.
• Do the groups need to get explicitly added as collaborators with write permission?

[climbfuji]

See Steve's comment above, this is not working as expected (yet):

[mkavulich]

@climbfuji @gold2718 The tags should be fixed now, I needed to include the organization as well as the team name.

@gold2718 I realize I haven't attached you to any of the existing teams yet. Would it make sense for you to be added to @NCAR/ccpp-framework-ncar-contingent or would you rather stay as an "independent" code owner (I can add your name back manually).

Reviews should work the same as in the past for reviewers who are a team member, they should get a "ping" the same way they would get if they were tagged individually. I'm not actually sure what the visibility is for teams; I am able to mouse over and click them in the top-right list of reviewers but I'm not sure if that applies to all, please try it out. The groups did need write permission for the repository to be reviewers; this has already been done.

[gold2718]

@gold2718 I realize I haven't attached you to any of the existing teams yet. Would it make sense for you to be added to @NCAR/ccpp-framework-ncar-contingent or would you rather stay as an "independent" code owner (I can add your name back manually).

I would like to hear from the NCAR folks on this (@peverwhee, @cacraigucar, @nusbaume). The question is whether my review should be considered a valid NCAR review (because a review from me would then clear the NCAR flag). I'm sort of leaning away from this since I have not been going to the CAM-SE or CAM-SIMA meetings, I am not closely following all the necessary details. Still I'd like to be useful so I'm okay being on that team if folks think it would help.

[climbfuji]

I would be in favor of creating a NorESM team for Steve, but your call.

@mkavulich I noticed that you are in all teams. I wonder if that will tick the box for all of them if you approve.

[mkavulich]

@mkavulich I noticed that you are in all teams. I wonder if that will tick the box for all of them if you approve.

I noticed that as well...as a repo maintainer I can always override approvals so it doesn't make any technical difference, but this is a practical problem for tracking approvals at a glance. That said, I typically don't give official approvals unless Dustin is unavailable. We can discuss at the next meeting.

[peverwhee]

@gold2718 I realize I haven't attached you to any of the existing teams yet. Would it make sense for you to be added to @NCAR/ccpp-framework-ncar-contingent or would you rather stay as an "independent" code owner (I can add your name back manually).

I would like to hear from the NCAR folks on this (@peverwhee, @cacraigucar, @nusbaume). The question is whether my review should be considered a valid NCAR review (because a review from me would then clear the NCAR flag). I'm sort of leaning away from this since I have not been going to the CAM-SE or CAM-SIMA meetings, I am not closely following all the necessary details. Still I'd like to be useful so I'm okay being on that team if folks think it would help.

I'll also vote for a NorESM group for @gold2718 . I want your feedback on PRs, Steve. I just also want to look at things!

[gold2718]

I'm okay with a NorESM team, fingers crossed that it improves our workflow (or at least workflow visibility).

[climbfuji]

Can you please create and add the NorESM team?

[climbfuji]

... and remove yourself from the teams that you shouldn't be on, please?

[gold2718]

... and remove yourself from the teams that you shouldn't be on, please?

Who are these messages to? I do not have any ability to see teams, never mind change them.

[climbfuji]

... and remove yourself from the teams that you shouldn't be on, please?

Who are these messages to? I do not have any ability to see teams, never mind change them.

This was meant for the creator of the PR (@mkavulich) - sorry if that was not clear.

[mkavulich]

So I've run into a complication in creating the NorESM team... @gold2718 is not a member of the NCAR organization, so I can't add him. I have reached out to request that he be added to the NCAR organization.

[gold2718]

I used to be a member of the NCAR org but they stripped me when I left (probably automatic from the registration system). If that doesn't work, can we just make a team called gold2718 for now that just has me in it (i.e., just include me)? There is no I in team but there is me 😉

[mkavulich]

@gold2718 I got the response from CISL that they do restrict NCAR organization membership to those who have an active CIT login (which I assume you do not). Since Github Teams must be tied to an organization (and NCAR doesn't seem to allow external teams), I think it's best just to leave you as an individual reviewer rather than try to make some sort of team at this time. I've added you back to the CODEOWNERS file.

[gold2718]

I got the response from CISL that they do restrict NCAR organization membership to those who have an active CIT login

Well, now I'm confused as my goldy CIT login is still active (at least I can log into and run on Derecho with that login and my CIT credentials). goldy was also my NCAR login so I'm not sure what they are telling you is the whole story. I'm not sure it matters for now as I'm the only one here currently working on CCPP stuff.

On the other hand, I was talking about future plans with one of our recent refugee hires (from EPA and USGS) and he independently brought up the CCPP so he has at least heard about it.

[mkavulich]

@gold2718 We can definitely re-visit this if other NorESM personnel end up wanting access in the future. From some further correspondence with CISL, it seems like they are considering further restricting NCAR organization in the future, which would be unfortunate. I'll keep my eyes open for any future policy changes.

[mkavulich]

@dustinswales Could you give this a quick review as the other NOAA person?