“Bumper sticker management” inflates the cost of software and lowers its quality, according to Lloyd K. Mosemann II, a senior vice president of SAIC and retired deputy assistant secretary of the Air Force.
In a recent telephone interview with the AdaIC News
, Mosemann defined “bumper sticker management” as following rather than leading the consumer. His personal example stems from buying a VCR and choosing a Beta model after comparing its technological sophistication against VHS.
“Of course,” as he said, “consumers went with VHS. That’s bumper sticker management.”
Today, he compares the Beta vs. VHS situation with Ada vs. other languages. “Most senior people have no idea what Ada is vs. why Ada should be chosen over anything else.”
He spoke from a personal viewpoint and not as an SAIC representative.
Mosemann led software management for long enough to stumble again and again across the many obstacles to change. After 25 years as deputy assistant secretary to the Air Force, he retired as head of
communications, computers and support systems. “In that capacity I oversaw a leadership role with Ada,” he said.
Even as an early proponent of Ada, he admitted that he “had a tough time selling it to senior management once C++ came on the scene. People heard that it was new and better. They didn’t consider its difficulty with multiple tasking and multiple inheritance and how that affected software’s lifecycle. Those are some of the things that it made attractive but also difficult to maintain.”
Software managers still depend on Ada for developing robust and portable software. Today, Mosemann sites Praxis SPARKAda as an example of using Ada for its reliability. Otherwise, most companies go for “what’s new and what’s popular.” Software managers often ignore Ada because of a sense that “not enough people are experienced programmers. Also, it’s not sexy so people didn’t want to use it.”
He “supported Ada for the same reasons” that he is a “big supporter of process improvement.” Yet popularity rather than intelligent comparisons have also overwhelmed managers’ decision about how their software is developed. “Process is emphasized and enforced either when the government insists or when a bank wants to lower its software’s error rates and increase its robustness,” Mosemann said.
Armed for his illustrious career with a Bachelors and Masters degrees from the University of Chicago, Mosemann worked with the Air Force after spending 11 years in the Navy. The software community and government have recognized his leadership in improving management processes with two Presidential Meritorious Rank Awards, five Air Force Exceptional Service Medals, the Defense Meritorious Civilian Service Medal, the Society of Logistics Engineers Founders Medal, three Federal Computer Week 100
Awards, the Government Computer News
Stetson Award, a 1996 Inductee in the Government Computer News
Hall of Fame, and the DOD Software Program Managers Network H Mark Grove Award for Excellence in Software Management.
Mosemann's accomplishments include writing the standard “Guidelines for Successful Acquisition of Software Intensive Systems” and founding both the National Software Data and Information Repository and the Software Technology Conference (STC). In the keynote speech
at the most recent STC, held in April 2002 in Salt Lake City, Mosemann called for software managers to insist on “formal methods programming,” which he defined as “sound management, established engineering processes, robust software development environment, model based architecture, and a reliable programming language.”
In the AdaIC News
interview, he referred to the foundation for successful software development as a “three-legged stool: a mature process; robust tools like Rational’s; and a language that’s reliable, like Ada.”