From lindah@epocrates.com Fri, 07 Sep 2001 15:42:19 -0700 From: "Hagedorn, Linda" Date: Fri, 07 Sep 2001 15:42:19 -0700 Subject: RE: Egregious coding In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: text/plain Title: Egregious coding Significant documentation has already been produced showing the overhead associated with the code.  The effect of shutting off 'literal only' sql on the QA system would centralize the problem in a single location, and allow the parties involved to correct the code before moving to staging or production.  Setting cursor_share=force is a workaround.  Thanks.      What about something that would interfere with the parser?  What about language code pages?  Replace the  ' and ''  with something else that would cause the parse to fail?  Has anyone coded an exit on the parser?             All ideas are welcome.    Linda Finished reviewing only 10% of 95M of non-reusable SQL, and in the process of creating bug tickets.          -----Original Message-----From: Greg Moore [mailto:sqlgreg@pacbell.net]Sent: Thursday, September 06, 2001 6:00 PMTo: Multiple recipients of list ORACLE-LSubject: Re: Egregious coding >  I would love to hold a >  valid threat over their heads   The key word is valid.  Threatening to throw a switch and blow up any SQL that contains literals is a threat, but is it a valid threat?   I mean, what was your previous job, suicide bomber?   Seriously, providing accurate measurement of how much the extra parsing now costs, along with an evaluation of current CPU use and an estimate of how much longer they have before the system bottlenecks might be a touch more civilized.