Re: [GENERAL] UPDATE ... ON CONFLICT DO NOTHING - Mailing list pgsql-general

From Peter Geoghegan
Subject Re: [GENERAL] UPDATE ... ON CONFLICT DO NOTHING
Date
Msg-id CAH2-Wzn=HoJc9qi2L1639U52t=pAeVfzxSfsVRLbzGWHUDbt4Q@mail.gmail.com
Whole thread Raw
In response to [GENERAL] UPDATE ... ON CONFLICT DO NOTHING  (Alexander Farber <alexander.farber@gmail.com>)
List pgsql-general
On Tue, Mar 14, 2017 at 4:55 AM, Alexander Farber
<alexander.farber@gmail.com> wrote:
> I wish there would be possibility to add ON CONFLICT DO NOTHING /* and drop
> the review */ to the both UPDATE's above, but there is no such thing
> described at https://www.postgresql.org/docs/9.5/static/sql-update.html

Heikki wanted to keep the door open to a feature like this around the
time that speculative insertion (the underlying mechanism used by
UPSERT) went in. I eventually convinced him that it is drastically
more complicated than it appears. This is essentially because it's
much harder to back out of an UPDATE than it is to back out of an
INSERT. If you're really interested, search through the -hackers
archives from around April of 2015.

--
Peter Geoghegan


pgsql-general by date:

Previous
From: "David G. Johnston"
Date:
Subject: Re: [GENERAL] UPDATE ... ON CONFLICT DO NOTHING
Next
From: Patrick B
Date:
Subject: [GENERAL] index on search - pg 9.2