suggestion/question: I understand why this has been changed to release-1.7, and I'd be fine with it staying this way... but for at least the next few weeks, am I correct that these tests aren't going to do anything useful in practice, because we won't be updating the release-1.7 branch?

We could just comment out this whole config file basically until we do our first alpha for 1.7 and we'd lose nothing, I think. Periodics run against master do the "release-next" job until we create our first alpha.

(We've talked about wanting to do alphas more often, but we're not there yet and I'm thinking we could save some resources on our test cluster here)

Am I missing something? Could we, for now, not do any of these tests, or else maybe only run them every 2 days or every week rather than every 2 hours?

(I fully accept that this is a tangent to your PR here, and if you'd rather not have this discussion in this PR, then fair enough!)

You're right. The release-1.7 branch will only ever be identical to, or behind master, so it's pointless to run these tests right now....but it's what we've always done so I propose not changing that in this PR.
We could create another PR to disable them

but it's what we've always done so I propose not changing that in this PR.

Totally reasonable to leave it for a separate PR, yeah. I might raise that PR after this merges 👍

Yeah I think in the past we have not created the release branch until we actually want to cut a release (i.e. just before the first alpha)

but it's what we've always done so I propose not changing that in this PR.

Totally reasonable to leave it for a separate PR, yeah. I might raise that PR after this merges 👍