On Wed, Feb 3, 2016 at 10:48 AM, Robert Haas <robertmhaas@gmail.com> wrote:
> I don't feel qualified to have an opinion on whether this is an
> improvement. I'm a little skeptical of changes like this on general
> principle because sometimes one clientele wants error A to be reported
> rather than error B and some other clientele wants the reverse.
> Deciding who is right is above my pay grade.
Exclusion constraints can sometimes have one client deadlock, rather
than see an exclusion violation. The particular client that sees an
error is undefined either way, so I personally felt that that wasn't
very important. But that's a problem that I'm closer to, and I won't
express an opinion on this patch.
--
Peter Geoghegan