Re: RETURNING and DO INSTEAD ... Intentional or not?

From: Josh Berkus <josh(at)agliodbs(dot)com>
To: pgsql-hackers(at)postgresql(dot)org
Subject: Re: RETURNING and DO INSTEAD ... Intentional or not?
Date: 2007-09-13 20:02:35
Message-ID: 200709131302.35367.josh@agliodbs.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

All,

> I'll note that we currently prevent adding RETURNING to a *conditional* DO
> INSTEAD rule. This means that if we have a conditional DO INSTEAD rule
> which inserts into a different table than the final unconditional rule,
> we'll be RETURNING wrong or empty values. Mind you, that's a pretty
> extreme corner case.

FYI, after some tinkering around, I've found that RETURNING is 100%
incompatible with any table which has conditional DO INSTEAD rules; there's
just no way to make it work and return any intelligible data. This would be
a completely corner case, except that people use conditional DO INSTEAD rules
heavily with partitioning (and yes, real users are complaining).

I don't see this as super-urgent to fix for 8.3, but can we put it up as a
TODO?

-- Make it possible to use RETURNING together with conditional DO INSTEAD
rules, such as for partitioning setups.

--
Josh Berkus
PostgreSQL @ Sun
San Francisco

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2007-09-13 21:46:26 Another HOT thought: why do we need indcreatexid at all?
Previous Message Heikki Linnakangas 2007-09-13 19:15:01 Re: Problem