Re: [HACKERS] Index created in BEFORE trigger not updated during INSERT

On Sun, Jun 4, 2017 at 7:23 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> I'm unsure whether to back-patch or not; the main argument for not doing
> so is that if any extensions are calling DefineIndex() directly, this
> would be an API break for them.  Given what a weird case this is, I'm not
> sure it's worth that.

Knowing that it has taken 20 years to get a report for this problem,
It seems to me that one report does not win against the risk of
destabilizing extensions that would surprisingly break after a minor
update.

> A possible end-run around the API objection would be to not add an extra
> argument to DefineIndex() in the back branches, but to use !is_alter_table
> as the control condition.  That would work for the core callers, although
> we might need a special case for bootstrap mode.  On the other hand,
> thinking again about hypothetical third-party callers, it's possible that
> that's not the right answer for them, in which case they'd be really in
> trouble.  So I'm not that much in love with that answer.

Yes, that's not worth the trouble I think for back-branches.

The patch looks good to me, could you add a regression test? This
could be used later on as a basis for other DDL commands if somebody
looks at this problem for the other ones.
-- 
Michael