On Tuesday, April 22, 2025, David G. Johnston <
david.g.johnston@gmail.com> wrote:
On Tuesday, April 22, 2025, David G. Johnston <david.g.johnston@gmail.com> wrote:
On Tuesday, April 22, 2025, Steve Chavez <steve@supabase.io> wrote:
> alter event trigger command which doesn’t need to be exercised here
That part does need to be tested, I modified `AlterEventTriggerOwner_internal` to allow altering owners to regular users. Before it was only restricted to superusers.
Ok. I missed this.
Sorry for the self-reply but this nagged at me.
It’s probably not a big deal either way, but the prior test existed to ensure that a superuser couldn’t do something they are otherwise are always permitted to do - assign object to whomever they wish. So event_trigger.sql had a test that errored showing this anomaly. You moved the test and now are proving it doesn’t error. But it is not expected to error; and immediately above you already show that a non-superuser can be an owner. We don’t need a test to show a superuser demonstrating their normal abilities.
IOW, select test cases around the feature as it is implemented now, not its history. A personal one-off test to ensure that no super-user prohibitions remained will suffice to make sure all such code that needed to be removed is gone.
David J.
