Re: SQLPLUS Question

From: Mladen Gogala <gogala.mladen_at_gmail.com>
Date: Sat, 16 Jan 2010 23:15:31 +0000 (UTC)
Message-ID: <hithaj$d19$2_at_solani.org>



On Sat, 16 Jan 2010 23:08:18 +0100, Gerard H. Pille wrote:

> Everyone knows you know a bit about it (Oracle). You know near to
> nothing about ksh.

OK. You write me a script doing something related to Oracle in ksh and I will write the same script in Perl. Let's compare the execution times and the memory footprint. Also, as a minimum security measure, let's pack the passwords in hexadecimal form, so that they're not visible from the scripts at first glance. This is what I have in mind:

[mgogala_at_medo ~]$ perl -e 'print unpack("H*","scott/tiger"),"\n";' 73636f74742f7469676572
[mgogala_at_medo ~]$

If your script is faster and more secure than mine, I will publicly apologize. If, on the other hand, my script performs better than yours, you will publicly apologize to me in this forum. That's a fair bet, will you take it? We both will have the fullest freedom of expression in writing the script.
I will accept Jonathan Lewis, Tanel Poder, Nial Litchfield or Nuno Suoto as judges in this matter.
Let me specify things a bit further: You create a data model and the script to populate the tables. You define the task and create a ksh script to do the task. I will write the same thing using Perl and will not change the data model by adding indexes or triggers, clustering tables or doing anything that would be incompatible with the existing data model. I will have the full freedom to add PL/SQL objects, specifically, functions, procedures and packages but not triggers. Username and password combination must not be hard coded as an ASCII text in the script. In other words, things like "connect scott/tiger_at_local" are not allowed.

What say you?

-- 
http://mgogala.freehostia.com
Received on Sat Jan 16 2010 - 17:15:31 CST

Original text of this message