On Thu, Jun 9, 2011 at 12:48 AM, Pavel Golub <pavel@microolap.com> wrote:
>>> it's a feature addition I approve of. I think serious consideration
>>> ought to be given to locking down returned results so PQsetvalue refuses
>>> to touch them, instead. Otherwise we're likely to find ourselves unable
>>> to make future optimizations because we have to support this
>>> barely-used-by-anybody corner case.
>
> Do I understand correctly that there is no any chance at all to have function
> like PQdeleteTuple in libpq? (see my message "PQdeleteTuple
> function in libpq" on Wed, 1 Jun 2011)
It means the feature is on thin ice. I'm personally in favor of
keeping it, and perhaps cautiously exploring ways to go further. I'm
waiting for Tom to make a call...I see three options:
1) patch bug and leave behavior alone (apply andrew C's patch)
2) discuss behavior change in -hackers
3) patch bug and behavior immediately (get a new patch with doc change as well)
merlin