Jarek Potiuk
Jarek Potiuk
Yeah. Let me re-open it then. Yes. Flaky tests with very low probability of failing are the worst to track. Parallel runs have this property that they are running in...
Yes. But question is how to make them unique
This is not happenning recently. Likely fixed by other flaky test fixes.
> @potiuk , I think, "lower bound" was given as an example in the warnings.~~ > > For example, asgiref is currently set to be `"asgiref>=2.3.0",` in hatch_build.py. > >...
BTW. I see those warnings as a great opportunity to guide us with reviewing some of our dependencies - because like in this case we might have some surprising and...
> @potiuk , can you assign me this issue? sure :) BTW. I can add you to the triage team so that you can do it yourself (and help with...
https://github.com/apache/airflow/pull/43074 ?
> I don't see an issue with batching them. Me neither. > @potiuk I'm slightly hesitant to set the minimum purely based on "~6 months old" though. Are we planning...
BTW. We already have a number of some arbitrary lower-bounds - not documented - and I can't recall a single case where it cause problems.
cc: @ofek -> would love to get some comments on that.