Re: plpgsql_check_function - rebase for 9.3

Robert Haas <robertmhaas@gmail.com> writes:
> This is a very good point.  Annotating the function itself with
> markers that cause it to be more strictly checked will create a
> dump/reload problem that we won't enjoy solving.  The decision to
> check the function more strictly or not would need to be based on some
> kind of session state that users could establish but dump restore
> would not.

One would hope that turning off check_function_bodies would be sufficient
to disable any added checking, though, so I don't see this being a problem
for pg_dump.  But there might be other scenarios where an additional knob
would be useful.
        regards, tom lane