Peter Eisentraut <peter(dot)eisentraut(at)2ndquadrant(dot)com> writes:
> When you dump out the regression test database and load it back in, a
> few tables end up with different column orders:
> ...
> This table is part of a lengthy inheritance chain, so this might be
> intentional or too hard to fix. This behavior goes back to 9.2 and
> possibly further.
Yeah, the variation in f_star and friends is intentional (and very
very old).
Can't say about the partition tests though.
regards, tom lane