public inbox for [email protected]  
help / color / mirror / Atom feed
From: Ronan Dunklau <[email protected]>
To: [email protected]
To: David Rowley <[email protected]>
Cc: Ranier Vilela <[email protected]>
Subject: Re: Add proper planner support for ORDER BY / DISTINCT aggregates
Date: Thu, 15 Jul 2021 15:02:00 +0200
Message-ID: <2746808.Lq6AjvN2yG@aivenronan> (raw)
In-Reply-To: <CAApHDvrqnL3U=H6KznCsqtROGTgVT0d3rT0p7zDbkbXb-Fy8kA@mail.gmail.com>
References: <CAApHDvpHzfo92=R4W0+xVua3BUYCKMckWAmo-2t_KiXN-wYH=w@mail.gmail.com>
	<CAEudQAq-H2yNzCni1N_PW7PfrWr4jYMhzRAiQiGyAQ4vt4-sTg@mail.gmail.com>
	<CAApHDvrqnL3U=H6KznCsqtROGTgVT0d3rT0p7zDbkbXb-Fy8kA@mail.gmail.com>

Le mardi 13 juillet 2021, 06:44:12 CEST David Rowley a écrit :
> I've attached the updated patches.

The approach of building a pathkey for the first order by we find, then 
appending to it as needed seems sensible but I'm a bit worried about users 
starting to rely on this as an optimization. Even if we don't document it, 
people may start to change the order of their target lists to "force" a 
specific sort on the lower nodes. How confident are we that we won't change this 
or that we will be willing to break it ?

Generating all possible pathkeys and costing the resulting plans would be too 
expensive, but maybe a more "stable" (and limited) approach would be fine, like 
generating the pathkeys only if every ordered aggref shares the same prefix. I 
don't think there would be any ambiguity here.

-- 
Ronan Dunklau







view thread (64+ messages)  latest in thread

reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: [email protected]
  Cc: [email protected], [email protected], [email protected], [email protected]
  Subject: Re: Add proper planner support for ORDER BY / DISTINCT aggregates
  In-Reply-To: <2746808.Lq6AjvN2yG@aivenronan>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox