Received: from sss2.sss.pgh.pa.us (sss.pgh.pa.us [209.114.166.2]) by hub.org (8.10.1/8.10.1) with ESMTP id e673TrW68580 for ; Thu, 6 Jul 2000 23:29:54 -0400 (EDT) Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1]) by sss2.sss.pgh.pa.us (8.9.3/8.9.3) with ESMTP id XAA08480; Thu, 6 Jul 2000 23:29:42 -0400 (EDT) To: Paul McGarry cc: pgsql-sql@postgresql.org Subject: Re: MAX() of 0 records. In-reply-to: <396535EF.8132CF80@opentec.com.au> References: <396535EF.8132CF80@opentec.com.au> Comments: In-reply-to Paul McGarry message dated "Fri, 07 Jul 2000 11:44:15 +1000" Date: Thu, 06 Jul 2000 23:29:42 -0400 Message-ID: <8477.962940582@sss.pgh.pa.us> From: Tom Lane X-Archive-Number: 200007/71 Paul McGarry writes: > However there will be situations where there are no records for > a given item_entry_id with item_live='t'. Currently when I try > do update/insert a record so that this occurs I get the following > error 'ERROR: ExecutePlan: (junk) 'ctid' is NULL!' and the > insert/update attempt is rolled back. This seems like a backend bug to me, but being an overworked hacker I'm too lazy to try to reconstruct the scenario from your sketch. Could I trouble you to submit a formal bug report with a specific, hopefully compact script that triggers the problem? > Is there a good way of going about this or should I just be wrapping > the whole thing up in an Until I've isolated the bug I don't want to speculate about whether it'll be reasonable to try to back-patch a fix into 7.0.*. Usually we don't risk back-patching complex fixes into stable releases, but the fix might be simple once we know the cause. regards, tom lane PS: I trust you're using 7.0.* ?