Re: invalidating cached plans

Neil Conway <neilc@samurai.com> writes:
> (BTW, another thing to consider is how the rewriter will effect a plan's 
> dependencies: I think we should probably invalidate a plan when a 
> modification is made to a view or rule that affected the plan.

This issue goes away as long as you follow the rule that any change to a
table's schema invalidates all plans that mention the table.  Views and
tables that have rules will still be mentioned in the rangetable of the
resulting plan, even if they aren't part of the active plan.  (We use
that for access rights checking.)

Too tired to consider the other details at the moment...

			regards, tom lane