If you read my post from a couple of days ago you’ll know I’m in Sydney listening to Ralph Kimball proclaiming all there is to know about ETL in Depth.
On Monday he made a point that I hadn’t really thought about before but which I think I’m absolutely in agreement with and that is
…if you want to write code, then this (ETL) might not be for you.
I think he is right on the money here and actually I wonder if writing code has really shifted from mainstream to a niche skill in the last few years?
I used to write code, in fact only yesterday I was reminiscing with someone about my BASIC skills which I learned on a Commodore VIC-20 and then refined on a BBC Model B back in the early 80’s. I then programmed in COBOL and ADABAS Natural for a living for many years so I do understand the passion for coding and the immense satisfaction that can be had from creating something with real business value from nothing more than keystrokes and coffee.
However, these days I notice that even within our own local IT department there are concerns about the rise of the integrator, and I note there is a more than a little disdain for a role where augmentation skills appear to be valued more highly than actually writing code.
So back to the point, ETL definitely involves writing code, but it isn’t just about writing code. You could spend (and I do) hours working on an issue, trying different approaches and testing performance impacts before deciding on a configuration setting which is made by simply clicking on a check box. So the problem solving challenge of coding is there but in a different form. Whether today’s coders are happy to jump ship and become ETL developers remains to be seen but I agree that they are different animals and ETL is most likely not for everyone.

I really like the image in this blog. Is it public domain? May I use it?
http://blog.une.edu.au/robbi/files/2009/03/computer-code.jpg
,..] blog.une.edu.au is other great source on this topic,..]