I wrote:
> Heikki Linnakangas <hlinnaka@iki.fi> writes:
>> On 01/11/2020 22:47, Tom Lane wrote:
>>> With that, we don't actually need aggregate_dummy() to exist at
>>> all, because it's never referenced. Having "aggregate_dummy"
>>> as the prosrc value for an aggregate function is now just a
>>> random convention; any other string would do as well. (We could
>>> save a few bytes in pg_proc by choosing a shorter string, but
>>> probably it's better to stick to the existing convention.)
>> NULL would seem like the natural value for that.
> I wouldn't be in favor of that unless we changed the prolang value
> as well. Which could certainly be considered, but it makes the
> patch rather more invasive, and I'm not sure it's worth it.
Looking closer, I see that pg_proc.prosrc is marked NOT NULL,
so this couldn't work anyway unless we wish to remove that marking.
Which doesn't seem particularly wise. I pushed this without any
change in the catalog contents.
regards, tom lane