Does it hurt to update a lot more often?
It is now after April 2023, should that be updated?
I wouldn’t. Once in a while the update will have something broken about it (even though it has passed tests.) My approach is to wait for a day or two, check these forums for bugs or support issues, then update on a weekday afternoon. I pick the time, because if there is a problem, I hope to see a fix pretty soon. Updating in the middle of the night on the weekend would be more risky.
It’s not often that the latest is broken, and it’s generally fixed quickly, but - in my view - there’s no need to amplify the risk by updating more often or at bad times.
So Wednesday at about this time would be a good time to update?
I use afternoon in UK or EU time, on the grounds that the US is also awake at that time, or will be soon.
If you use any Pavilion plugins it’s worth checking our Plugin Status Dashboard for current compatibility in the event theres been a breaking change from core.
That’s cool. Is there a delay between a new test-passed branch being announced here on Meta and it being the tests-passed branch mentioned on your dashboard?
tests-passed
is a continuous stream of merges.
Our canary servers (one for tests-passed and one for stable) rebuild every 24 hours.
Let’s say I use the Custom Wizard plugin and am on 3.1.0.beta6. Would I be able to use the form to decide whether to upgrade to 3.1.0.beta8? Should I leave 24 hours after the upgrade announcement on Meta?
No. If it says compatible just do it.