By Karl G. Rogers

It is interesting and informative to reflect back over 25 years of software development. Twenty-five years ago the Software Crisis was raging and there were many engaged in trying to convert software code production from art to engineering science. The DoD was actively funding the pursuit of a solution. The term Software Crisis was first coined at a NATO Software Engineering Conference in 1968. It was the result of dramatic increases in computing power outpacing the ability of developers to produce working software. It is no wonder the DoD was interested in improving the odds of software being successful; at the time, approximately one in eight finished software projects were considered successful. It was this DoD effort to improve software development that originally funded the creation of CrossTalk as an information exchange forum.

Watts Humphrey published the CMM® 25 years ago in 1988 and as a book, “Managing the Software Process” the following year. This was the beginning of a lot of great work on software process improvement. CMM would later be followed by other great works by Watts Humphrey such as Team Software Process (TSP) and Personal Software Process (PSP). All along the way, CrossTalk has been there covering the transformation of the software industry from crisis to manageable and predictable software development. CrossTalk has published articles about many types of process improvement, some of which have been successful and others not so much. Many of us have witnessed firsthand this transformation of the software industry. We have seen the transformation from very limited process control to process control being the rule, not the exception. We have seen the progression from CMM to the CMMI®.

Today we continue to strive to improve quality and predictability while at the same time reducing cost. Unlike 25 years ago, we now have data and processes that support controlled predictable high-quality software development. We have all probably participated in the debates about what amount of process improvement/control is enough. As the Software Maintenance Group Director, I don’t know the ultimate answer to the question; however we continue to pursue improved software predictability, quality and price. This issue of CrossTalk is focused on just how things have changed over the last 25 years. I hope you enjoy the perspectives provided in this issue of CrossTalk.

Karl Rogers

Software Maintenance Group Director 

309th Software Maintenance Group


CMMI® and CMM® are registered in the U.S. Patent and Trademark Office by Carnegie Mellon University

Next »