X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id B2736D1B1BE for ; Wed, 16 Jun 2004 05:42:20 -0300 (ADT) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024) with ESMTP id 17259-08 for ; Wed, 16 Jun 2004 08:42:16 +0000 (GMT) Received: from gatefox.foxi.nl (foxi.xs4all.nl [194.109.246.192]) by svr1.postgresql.org (Postfix) with ESMTP id 68DAAD1B193 for ; Wed, 16 Jun 2004 05:42:12 -0300 (ADT) Received: from techfox.foxi (techfox.foxi [192.168.0.79]) by gatefox.foxi.nl (8.10.2-20030922/8.10.2) with ESMTP id i5G8gAu19362; Wed, 16 Jun 2004 10:42:10 +0200 From: Frank van Vugt Organization: Foxi Consultants in Technology BV To: Tom Lane Subject: Re: why is a constraint not 'pushed down' into a subselect when this subselect is using a 'group by' ?? Date: Wed, 16 Jun 2004 10:42:10 +0200 User-Agent: KMail/1.5.4 References: <200406150849.29470.ftm.van.vugt@foxi.nl> <200406151730.40464.ftm.van.vugt@foxi.nl> <3139.1087322681@sss.pgh.pa.us> In-Reply-To: <3139.1087322681@sss.pgh.pa.us> Cc: pgsql-performance@postgresql.org MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200406161042.10259.ftm.van.vugt@foxi.nl> X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.0 tagged_above=0.0 required=5.0 tests= X-Spam-Level: X-Archive-Number: 200406/159 X-Sequence-Number: 7217 > We don't attempt to make every possible inference (and I don't think > you'd like it if we did). I wasn't really asking you to, either ;)) Just trying to achieve a more in-depth understanding of the way things work. > This example doesn't > persuade me that it would be worth expending the cycles to do so. In the real thing I can easily get good processing times by adding an extra join with article inside in the group by and simply use the constraint on that as well, so I'm ok with any choice you make on this. I thought this might have been some kind of special case though, given its occurence on the use of group by. > for example, zero and minus zero in IEEE-standard float arithmetic. Good example, brings back a few memories as well ;) Thanks for your explanation, Tom! -- Best, Frank.