Re: [PATCH] Add EXPLAIN (ALL) shorthand

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Евгений Шишкин <itparanoia(at)gmail(dot)com>
Cc: David Christensen <david(at)endpoint(dot)com>, PostgreSQL-development Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: [PATCH] Add EXPLAIN (ALL) shorthand
Date: 2016-05-19 22:12:47
Message-ID: 49766.1463695967@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

=?utf-8?B?0JXQstCz0LXQvdC40Lkg0KjQuNGI0LrQuNC9?= <itparanoia(at)gmail(dot)com> writes:
>> On 19 May 2016, at 22:59, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> I'm not sure this is well thought out. It would mean for example that
>> we could never implement EXPLAIN options that are mutually exclusive
>> ... at least, not without having to redefine ALL as all-except-something.
>> Non-boolean options would be problematic as well.

> Maybe EVERYTHING would be ok.

That's not really getting at the substance of my complaint, which is that
I foresee regretting inventing such an option later, when we wish to invent
some option that it's not reasonable for ALL/EVERYTHING/WHATEVER to turn
"on" (if indeed the new option is "off"/"on" in the first place). We'll
either have to accept an arbitrary/inconsistent definition of what ALL
does to that option, or have to explain that ALL doesn't mean all.

My own experience is that I seldom want ANALYZE and VERBOSE at the same
time, so I'm doubtful that I'd find this proposal helpful even just
considering the current option set.

I'm also not convinced about the semantics of, say, EXPLAIN (ALL, COSTS
OFF). I see what the patch will do, but I'm not convinced I like that
(and I am convinced that the patch's documentation is misleading about
it).

I'm a bit inclined to think that what this is really about is that we
made the wrong call on the BUFFERS option, and that it should default
to ON just like COSTS and TIMING do. Yeah, that would be an incompatible
change, but that's what major releases are for no?

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Gavin Flower 2016-05-19 22:16:18 Re: [PATCH] Add EXPLAIN (ALL) shorthand
Previous Message David G. Johnston 2016-05-19 22:11:41 Re: [PATCH] Add EXPLAIN (ALL) shorthand