Re: Implementing RESET CONNECTION ...

Bruce Momjian <pgman@candle.pha.pa.us> writes:
> Hans-J�rgen Sch�nig wrote:
>> I completely agree with Karel. I think it is a bad idea to change the
>> protocol for such a minor feature - i tend to call it overkill.

> I think autocommit is a good example for comparison.

Indeed, it is an *excellent* example for comparison.  The real problem
with autocommit was that it changed the interface semantics without
making that change sufficiently visible at all levels.

If we try to pretend that RESET CONNECTION isn't a protocol change
then we will silently break code that needs to know about it.  Which is
pretty much exactly what happened with autocommit.

> Should we add something like SET
> CONNECTION that would set the reset values for RESET CONNECTION?

This is an even bigger compatibility-breaker, as now anyone who can
issue SET CONNECTION can not only break code layers that were trying to
track backend state, he can break code layers that thought they knew
what RESET CONNECTION would accomplish.  I definitely recommend against
this idea.

            regards, tom lane