Seems generally fine to me (aside from jenkins failures. are those related?)

No jenkins has been failing for a while now, local test are passing (I'll start a new run to double-check)

So I can confirm the tests are still passing :)

@ralph just to be sure, are we good to go?

Is signing helpers the right term here?

Is it worth supporting gitolite2 for tech debt reasons? Just wondering if anything in Fedora uses v2.

Edit: NVM, it's very little code difference.

If you removed this if statement I think you could delete config.append(' RW+ = %s' % project.user.user), since they're both getting the same kind of access.

What is a deploykey?

scary

So I'm not too familiar with gitolite but overall the change looks ok. A few minor comments is all.

+0.5 (would +1 if I was more comfortable with this part of the codebase).

Adjusted, left over from copy/pasta ^^

Good call :)

They are SSH keys you can generate and give to a service to grant them commit access to your repo.

agreed though pretty contained as everything comes from the config file

Amusingly, turned out that line cannot be removed as main admins aren't listed in the committers. So I added a comment regarding this.

I forgot to mention that I merged this because @bowlofeggs also reviewed it in part of https://pagure.io/pagure/pull-request/2361