Tom Lane wrote:
> Bruce Momjian <pgman@candle.pha.pa.us> writes:
> > SET autocommit to 'off' is going away in 7.4 so the SET USER command
> > will not work anymore in that release, and I can't think of a
> > workaround. There will be interface-specific settings, I assume, but I
> > am not sure how that would be controlled per-user.
>
> As I was just telling some Red Hat cohorts, I intend to provide
> autocommit support in psql for 7.4 --- probably controlled by a psql
> variable, though I've not gotten round to designing the details yet.
> In any case it will be possible to turn autocommit off in your ~/.psqlrc
> file if you have the desire to do so. I would also say that the setting
> will not apply to psql -c invocations, thus avoiding the worst effects
> for scripts ...
Youch --- a .psqlrc that doesn't effect psql -c. First we had SET
doesn't get rolled back if it is the first command of a transaction,
now, with that gone, we have psql -c not reading autocommit settings in
.psqlrc. The warts seem to follow autocommit where ever it goes! :-)
-- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610)
359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square,
Pennsylvania19073
