Simon Riggs <simon(at)2ndquadrant(dot)com> writes:
> Tom has not spoken against checking for UNIQUE constraints: he is just
> pointing out that there never could be a constraint in the case I was
> identifying.
More generally, arguing for or against any system-wide change on the
basis of performance of compute_minimal_stats() is folly, because that
function is not used for any common datatypes. (Ideally it'd never be
used at all.)
regards, tom lane