Heh. If you know that the input file contains strictly new records, why
bother with that logic at all?
One of my clients wanted to create a couple of new companies and duplicate
the records of what I'll call his master company for the two new companies.
His software provider quoted him 40 hours for writing code to accomplish
this task. I did it in 4 hours just by knowing a few back door tricks I
learned back in the late 70's. He now has a SEQUEL script that prompts him
for the original company number and his new company number. The script takes
care of the rest.
It uses zero lines of code. One of the programmers at the software provider
asked me to show him how I did it. He starts a new job in January at a
company that already has SEQUEL, and wanted to get a head start. :-))
[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of rob@xxxxxxxxx
Sent: Thursday, December 26, 2013 10:56 AM
To: Midrange Systems Technical Discussion
Subject: RE: SQL created table slows reporting tool
I hear you Paul. I have a tough time explaining to the OVER 40 crowd why
their updates take so long.
They read this huge input file. Instead of summarizing it by the
subtotals and only writing at subtotal time they do a
if no record found,
Thus having over a million extra I/O's.
It's like they've never heard of control breaks, total time, etc.