Hello,
>> This patch was marked as ready for committer, but clearly there's an
>> ongoin discussion about what should be the default behavoir, if this
>> breaks existing apps etc. So I've marked it as "needs review" and moved
>> it to the next CF.
>
> The issue is that root (aka Tom) seems to be against the feature, and would
> like the keep it as current. Although my opinion is that the previous
> behavior is close to insane, I'm ready to resurect the guc to control the
> behavior so that it would be possible, or even the default.
>
> Right now I'm waiting for a "I will not veto it on principle" from Tom (I'm
> okay with a reject based on bad implementation) before spending more time on
> it: Although my time is given for free, it is not a good reason to send it
> down the drain if there is a reject coming whatever I do.
>
> Tom, would you consider the feature acceptable with a guc to control it?
Tom, I would appreciate if you could answer this question.
For me, the current behavior is both stupid and irritating: why would pg
decide to drop the result of a query that I carefully typed?. It makes the
multi-query feature basically useless from psql, so I did not resurrect
the guc, but I will if it would remove a veto.
In the meantime, here is a v9 which also fixes the behavior when using
\watch, so that now one can issue several \;-separated queries and have
their progress shown. I just needed that a few days ago and was
disappointed but unsurprised that it did not work.
Watch does not seem to be tested anywhere, I kept it that way. Sigh.
--
Fabien.