Andrew Dunstan <andrew@dunslane.net> writes:
> On 3/22/22 12:58, Tom Lane wrote:
>> I only suggested removing the error check in _PG_init, not
>> changing the way the test works.
> Mark and I discussed this offline, and decided there was no requirement
> for the module to be preloaded. Do you have a different opinion?
No, I was actually about to make the same point: it seems to me there
are arguable use-cases for loading it shared, loading it per-session
(perhaps via ALTER USER SET or ALTER DATABASE SET to target particular
users/DBs), or even manually LOADing it. So the module code should
not be prejudging how it's used.
On reflection, I withdraw my complaint about changing the way the
test script loads the module. Getting rid of the need for a custom
.conf file simplifies the test module, and that seems good.
So I'm on board with Mark's patch now.
regards, tom lane