Lots of little comments on the PRs; thanks for cleaning up various little messes along the way.
But overall, I have to say I'm really torn about this:
- I'm afraid I don't think the Task adds much value; there's just so little shared logic (much of which really belongs in PhotoCalib), and eventually there will be even less. It would still be a small positive addition, but...
- There's a ton of config churn, and while I think our rules and precedent don't require an RFC for that (some of the configs you are deprecating here were added a few weeks ago on
DM-21308, and that was not accompanied by an RFC), I think in the future, they probably should: config changes are at least as visible as code interface changes, and often much more.
That said, adding the testing to the ci_hscs and of course adding recalibration to ForcedPhotCcd is useful new functionality that we absolutely want.
So, I'd feel a lot more comfortable with this if people whose daily work is more affected by config deprecations than mine (e.g. Yusra AlSayyad, Lauren MacArthur) chimed in on this. And I'd be interested to get your opinion on how tough it would be to take the config options out of the subtask and put them back in the parent task(s), presumably by passing them down as bools (sorry!).