Stephen Frost wrote:
> I can perhaps see a special case for SECURITY DEFINER functions but if
> we're going to special case them I'd think we'd need to make them only
> be creatable/modifiable at all by superusers or add another flag to the
> role to allow that.
I agree that owner changes of SECURITY DEFINER functions seem dangerous. I
would follow Stephen's idea that SECURITY DEFINER functions should only be
creatable/modifiable by superusers.
This would be similar to unix, where setting the suid/sgid bits is usually
only allowed to root.
Best Regards,
Michael Paesold