RE: Some database discussions...

Yet another of the double-edged swords in the IBM midrange world... :(




-----Original Message-----
From: Rob Berendt [mailto:rob@xxxxxxxxx]
Sent: Thursday, May 18, 2017 8:45 AM
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
Subject: Re: Some database discussions...

<snip>
The second issue is one of general error handling. It is my experience that few RPG programmers write code to detect errors, much less handle them.
</snip>

Amen to that. I've seen where a write error was assumed to be that the row already existed and then a chain and update pair were attempted. The theory was that so few existing rows were supposed to be there it cost more in I/O's to do the checking before the write.


Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1 Group Dekko Dept 1600 Mail to: 2505 Dekko Drive
Garrett, IN 46738
Ship to: Dock 108
6928N 400E
Kendallville, IN 46755
http://www.dekko.com