Re: ON CONFLICT does not support deferrable unique constraints - Mailing list pgsql-general

From Francisco Olarte
Subject Re: ON CONFLICT does not support deferrable unique constraints
Date
Msg-id CA+bJJbz_XRAgfc=6bsuzMtO6iynOoqDDCL1UHiW1mzrcWUkd-Q@mail.gmail.com
Whole thread Raw
In response to Re: ON CONFLICT does not support deferrable unique constraints  (Andreas Joseph Krogh <andreas@visena.com>)
Responses Re: ON CONFLICT does not support deferrable unique constraints  (Peter Geoghegan <pg@bowt.ie>)
List pgsql-general
On Wed, Aug 24, 2016 at 9:22 PM, Andreas Joseph Krogh
<andreas@visena.com> wrote:
> As a developer I want it to "just work", if there's an error of any kind then abort the transaction, just as it was
non-deferrable.

Everybody wants everything to "just work", for their own ( in a lot of
cases unspecified even to the themselves ) definition of JW.
Developers normally want to be told what "just work" means when
developing something.

That been said, I'm not sure making it ( deferred constraint act like
immediate ones during upserts ) work is even a good idea. If it can be
conditionally enabled with a simple set and implemented in very few (
< 20 ) lines of code, ok for me , otherwise I would prefer the reduced
bug surface.

Francisco Olarte.


pgsql-general by date:

Previous
From: Alexander Farber
Date:
Subject: Re: Forward declaration of table
Next
From: Russell Keane
Date:
Subject: corruption in indexes under heavy load