Received: from sss.pgh.pa.us ([216.151.103.158]) by postgresql.org (8.11.3/8.11.1) with ESMTP id f34Fsa847368 for ; Wed, 4 Apr 2001 11:54:36 -0400 (EDT) (envelope-from tgl@sss.pgh.pa.us) Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1]) by sss.pgh.pa.us (8.11.3/8.11.3) with ESMTP id f34FsNw10794; Wed, 4 Apr 2001 11:54:23 -0400 (EDT) To: Kyle cc: pgsql-sql@postgresql.org Subject: Re: Query broken under 7.1RC2 In-reply-to: <3ACB3CB9.CB5F0C46@actarg.com> References: <3ACB3CB9.CB5F0C46@actarg.com> Comments: In-reply-to Kyle message dated "Wed, 04 Apr 2001 09:24:41 -0600" Date: Wed, 04 Apr 2001 11:54:22 -0400 Message-ID: <10791.986399662@sss.pgh.pa.us> From: Tom Lane X-Archive-Number: 200104/47 X-Sequence-Number: 2386 Kyle writes: > The SQL is a little strange because the subquery tries to reference > individual records from the outer query and then sum them. > The more I look at it, I wonder if it is not just bad SQL. It is bad SQL, but pre-7.1 Postgres failed to catch it. Here's the CVS log entry: 2000-09-25 14:09 tgl * src/backend/optimizer/plan/planner.c: System neglected to complain about ungrouped variables passed to sublinks when outer query contained aggregates but no GROUP clause. Since you have a SUM(), the outer query is a grouped query, and so references to the ungrouped h.* columns in the SELECT targetlist are not well defined. Lord knows what result you were getting from 7.0 ... regards, tom lane