Received: from dvl by smile.2scale.net with local (Exim 3.12 #1) id 1AnzMp-0005eE-00; Tue, 03 Feb 2004 13:09:23 +0100 Date: Tue, 3 Feb 2004 13:09:23 +0100 From: Dirk Vleugels To: "Schroeder, Alexander" Cc: 'Dirk Vleugels' Subject: Re: SELECT Problem Message-ID: <20040203130923.M19592@smile.2scale.net> References: <52F518B40AE66944A6F62699ADE3F04103BA314B@dewdfx2g.wdf.sap.corp> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <52F518B40AE66944A6F62699ADE3F04103BA314B@dewdfx2g.wdf.sap.corp>; from alexander.schroeder@sap.com on Tue, Feb 03, 2004 at 12:45:29PM +0100 On Tue, Feb 03, 2004 at 12:45:29PM +0100, Schroeder, Alexander wrote: > Aside from the issue in the optimizer - why would a change in the JVM change an SQL Statement? Different iterator strategies maybe? The generation of the statements is hidden away in the persistance framework. But the result _should_ be the same. > The hint is obviously specifying the tables in the same order, and of course keeping the table statistics up-to-date. The statistics are up-to-date. Why should the order of the tables (the statement is equivalent, dont you agree?) matter? If the EXPLAIN Plan is the same, i expect both statements to produce the same behaviour ... Puzzled, Dirk