Hello, Noah! > I misunderstood, and I was mistaken to see this as a bug fix. The > isolationtester is acting per its definition, and this would be a definition > change.
Yes, you are right, but I think it is better to clarify it somehow because from my point of view that definition feels like logic in patched version. But maybe my non-native English prevents me from understanding it correctly. In my opinion, it is better to add some clarification like "this marker affects only other steps which were launched before this step or during its execution". > I'd need to review the motivating test to form my own opinion on whether the > new definition makes it easier to write tests. > You could bundle this in the thread that wants this to stabilize that thread's CI results. I'll think about it, but currently it is stabilized using "notices" and, you know, more changes - harder to merge :) Also, I'll try one more time to stabilize it without "notices" - maybe I'll get some insight. Anyway - I'll ping you in that thread [0] (you're already in recipients) or bring it here. Best regards, Mikhail. [0]: https://www.postgresql.org/message-id/flat/ZnoZ6GNwkJmq-gTh%40paquier.xyz#4d13f826fb1e62860cc3ae30067bd23a
