Honestly if the resultsets are small-enough, I just dump them to JSON and diff the files. But it has to be fully deterministically sorted for that (in a sane w

Honestly if the resultsets are small-enough, I just dump them to JSON and diff the files. But it has to be fully deterministically sorted for that (in a sane world "order by *" would be valid ANSI SQL).