On Tue, Feb 24, 2004 at 09:10:57PM -0500, Rudy Lippan wrote:
> Somewhat old, but:
>
> http://archives.postgresql.org/pgsql-general/2000-02/msg00376.php
The postgres guys sometimes seem to have a weird idea of what
application developers want. "Very minor programming shortcut" indeed!
It would be like unix terminating programs just because one system call
fails. :-)
> (musing aloud: I wonder if they would accept a patch that made it
> configurable?)
This message
http://archives.postgresql.org/pgadmin-support/2003-02/msg00043.php
makes it seem that it might be hard to get the transaction back into a
consistent state after an error. So it might be a lot of work to write
that patch. :-(
Andrew
