> But this line of thinking does strengthen my feeling that throwing an error is the right thing to do for the moment. If we allow v10 to accept such cases but do something different from what we used to, that > will greatly complicate any future attempt to try to restore the old behavior.
> regards, tom lane Agreed. The other side benefit of throwing an error instead of just doing something different is you'll find out how rampant the old behavior is :). People are more likely to know to complain when their apps break than they are if it just silently starts doing something different. My main concern in these cases is the short-circuiting not happening. Because in these cases, the code goes into areas that it shouldn't which is likely to mess up some logic in hard to troubleshoot ways. I think erroring out is the best compromise. Thanks, Regina -- Sent via pgsql-hackers mailing list (firstname.lastname@example.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers