Received: from localhost.localdomain (postgresql.org [64.49.215.8]) by localhost (Postfix) with ESMTP id EDFED4763E9 for ; Thu, 9 May 2002 10:35:35 -0400 (EDT) Received: from relay.uni-heidelberg.de (relay.uni-heidelberg.de [129.206.100.212]) by postgresql.org (Postfix) with ESMTP id CC34B4765E3 for ; Thu, 9 May 2002 10:27:42 -0400 (EDT) Received: from ix.urz.uni-heidelberg.de (mail.urz.uni-heidelberg.de [129.206.119.234]) by relay.uni-heidelberg.de (8.12.1/8.12.1) with ESMTP id g49ERiZi002510 for ; Thu, 9 May 2002 16:27:44 +0200 (MET DST) Received: from heiko1 (ppp439.urz.uni-heidelberg.de [129.206.203.239]) by ix.urz.uni-heidelberg.de (8.8.8/8.8.8) with SMTP id QAA28686 for ; Thu, 9 May 2002 16:27:44 +0200 Message-ID: <000601c1f767$17b34000$010aa8c0@heiko1> From: "Heiko" To: Subject: Bad time external representation '' Date: Thu, 9 May 2002 16:37:31 +0200 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2600.0000 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000 X-Archive-Number: 200205/95 X-Sequence-Number: 7731 hi, currently upgraded SuSE linux 7.0 to 8.0 including an upgrade from postgres 7.0 to 7.2. almost everything went very well, all the data intact, very peachy. only one seemingly stupid problem: queries like 'select case when a.dtmstart is null then '' else a.dtmstart end from a;' where 'dtmstart' is a time-field result in a 'bad time external representation ''' - error. (tried various casting tricks, none worked). it works fine if i just leave the entire 'case when' - section out; if i remember correctly, i put the section in because otherwise i got explicit 'null' values in my application; this does not happen anymore, so the 'select case' workaround has become superfluous. for various reasons it would be a major pain in the butt for me, however, to change the sql syntax in my applications; i would either have to produce and control twice the amount of versions or change my interface so that it catches the 'null's that 7.0 produces. is there a way to make postgres 7.2. accept my old 'case when'-queries? and on a logical basis: why should a 'case when' expect one explicit data-type? shouldn't it theoretically be possible to use a 'case when' also to hand back different types depending on whether the condition comes true or not? or am i overlooking something basic (would not be entirely untypical of me :-) thanks a lot for any help, h. jakubzik