> This style of "DROP COLUMN" would change the attribute
> numbers whose positons are after the dropped column.
> Unfortunately we have no mechanism to invalidate/remove
> objects(or prepared plans) which uses such attribute numbers.
> And I've seen no proposal/discussion to solve this problem
> for DROP COLUMN feature. We wound't be able to prevent
> PostgreSQL from doing the wrong thing silently.
That issue got me confused now (There was a previous mail where
you suggested using logical colid most of the time). Why not use the
physical colid in prepared objects/plans. Since those can't currently change
it seems such plans would not be invalidated.
Andreas