Sunday, June 01, 2014

If you worked in a mainframe shop after about 1990, you followed the rules

During my own 31 year career in information technology, mostly in legacy mainframe systems, I did learn that you have to build up procedures to guarantee the quality of your own work. 
The biggest areas where major issues could occur included the handling of production files, the handling of elevations (also called “promotions” or just “moves”), and, particularly, making sure that relatively infrequently used capabilities of a system (like an end-of-month or end-of-year process, or particularly data saving, whether by modern data warehousing or older systems like microfiche) are actually looked at occasionally by end users.

Since the late 1980s, most major mainframe shops had security systems (whether RACF or Top Secret) which prevented programmers from routinely having access to production files.  Also, since about 1990, most shops have implemented automated source and version management systems that, when followed, guarantee source-load module integrity. 

Yet, it was amazing to me that some programmers found these safeguards annoying, and thought that instead all programmers should be bonded.
I do wonder how these security concepts play out in modern Internet companies, most of all in workplaces like Facebook and Google, where systems testing has to be difficult and multi-staged. Good topic for more articles?
When I worked in this kind of environment, it became “my world”.  It wasn’t until I “retired” at the end of 2001 that I began to grasp the scope of the issues of workplace and career for most people.
I’ll be covering some quirks in these issues fairly soon in my new Wordpress blogs.

No comments: