Neil Conway <neilc(at)samurai(dot)com> writes:
> Jaime Casanova wrote:
>> Or are any reasons to keep that argument?
> None that I can see. I'll apply the patch later today, barring any
> objections.
Looking at the 7.3 sources, the argument was used to support the
then-existing convention that an error message include the name
of the reporting routine. This does seem obsolete now.
regards, tom lane