Re: operator dependency of commutator and negator, redux

Robert Haas <robertmhaas(at)gmail(dot)com> writes:
> On Thu, Dec 20, 2012 at 10:52 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> I was thinking a NOTICE at most. If it's a WARNING then restoring
>> perfectly valid pg_dump files will result in lots of scary-looking
>> chatter. You could make an argument for printing nothing at all,
>> but that would probably mislead people who'd fat-fingered their
>> COMMUTATOR entries.

> What about jiggering the dump so that only the second of the two
> operators to be dumped includes the COMMUTATOR clause?

Seems messy and fragile. In particular this'd represent a lot of work
in order to make it more likely that the restore malfunctions if someone
makes use of pg_restore's -l switch to reorder the entries. Also it
would not retroactively fix the problem for restoring dumps made with
existing pg_dump versions.

> Even a NOTICE in
> pg_dump seems like too much chatter (witness recent quieting of some
> other NOTICE messages we've all grown tired of)

pg_dump has included "set client_min_messages = warning" in its output
for quite some time now. So as long as we don't insist on making this
a WARNING, people won't see it in that usage.

regards, tom lane