Horton
Menu

Re: Error on failed COMMIT

Vladimir Sitnikov <sitnikov.vladimir@gmail.com>

View thread
From: Vladimir Sitnikov <sitnikov.vladimir@gmail.com>
To: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Cc: Shay Rojansky <roji@roji.org>, Robert Haas <robertmhaas@gmail.com>, "Haumacher, Bernhard" <haui@haumacher.de>, Dave Cramer <davecramer@postgres.rocks>
Date: 2020-02-23T12:14:25Z
Lists: pgsql-hackers 
Shay> Asking drivers to do this at the client have the exact same breakage
impact as the server change, since the user-visible behavior changes in the
same way

+1

Dave>While we can certainly code around this in the client drivers I don't
believe they should be responsible for fixing the failings of the server.

Application developers expect that the database feels the same no matter
which driver is used, so it would be better to avoid a case
when half of the drivers create exceptions on non-committing-commit, and
another half silently loses data.

Vladimir