Re: [patch] libpq one-row-at-a-time API

Marko Kreen <markokr(at)gmail(dot)com> writes:
> On Mon, Jul 16, 2012 at 4:33 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> Mm. I still think we should drop it, because it's still a dangerous API
>> that's not necessary for the principal benefit of this feature.

> Yes, it is a secondary feature, but it fits the needs of the actual target
> audience of the single-row feature - various high-level wrappers of libpq.

> Also it is needed for high-performance situations, where the
> single-row-mode fits well even for C clients, except the
> advantage is negated by new malloc-per-row overhead.

Absolutely no evidence has been presented that there's any useful
performance gain to be had there. Moreover, if there were, we could
probably work a bit harder at making PGresult creation cheaper, rather
than having to expose a dangerous API.

regards, tom lane