Ronan Dunklau <ronan@dunklau.fr> writes:
> That wouldn't solve the first problem mentioned, which is that for some
> regression tests one may want to test the costs themselves, which is now
> impossible with the new planning time feature.
That's a bogus argument, because it was impossible before too. We have
no such tests now, and it's unlikely we will ever add any, because costs
inherently are platform-dependent. The reason we invented COSTS OFF in
the first place was to make it possible to do EXPLAIN in regression tests
without getting platform-dependent output.
I have no great objection to making both COSTS OFF and TIMING OFF suppress
the "planning time" output, if that's the consensus. I would object to
taking away that behavior of COSTS OFF, because of the implications for
back-patching EXPLAIN queries in regression tests.
Another possibility, which would introduce less non-orthogonality into
the switch design, is to remove the connection to COSTS OFF but say that
planning time is only printed when execution time is also printed (ie,
only in EXPLAIN ANALYZE). This seems to me that it would not be removing
much functionality, because if you just did a plain EXPLAIN then you can
take the client-side runtime (psql \timing) as a close-enough estimate
of planning time.
regards, tom lane