Leading U.S. software scientists and software executives met to discuss progress on the Cooperstown Initiative during August 1994. Cooperstown II focused on the two national software initiatives that resulted from Cooperstown I (1993): the National Software Council (NSC) and the National Software Data and Information Repository (NSDIR).
The focus of the Cooperstown Initiative is on planning, building, and assessing the progress towards the creation of a national vision and force in the U.S. software industry through software measurement-based mechanisms. The Cooperstown Initiative and workshops are sponsored by the Office of the Deputy Assistant Secretary of the Air Force for Computers, Communications and Support Systems (SAF/AQK) in cooperation with Rome Laboratory.
Highlights of Cooperstown II included a keynote address by Mr. Lloyd Mosemann II, Deputy Assistant Secretary of the Air Force, and a talk by Dr. Barry Boehm of the University of Southern California (USC). In addition, announcements were made regarding plans to establish a National Center for Software Data and Information and a National Software Data and Information Respository User's Group.
Presentations
Dr. Boehm spoke on "Software Measurement: Looking Forward." He argued that software, embodying context-dependent intelligence, has become the central integrative discipline in many systems and will continue to grow in importance. As a result, software metrics will also continue to grow in importance. Metrics are needed for planning, control, and process improvement. Effective measurement practices must adapt to changing software practice, particularly the movement towards application programming and the growing importance of systems integration, application composition, and application generators. Dr. Boehm described how USC's COnstructive COst MOdel (COCOMO) 2.0 Project is handling these challenges and concluded with some comments on the timeliness and importance of the Cooperstown Initiative.
Mr. Mosemann introduced the vision guiding Cooperstown I in his keynote address. He described the software problems plaguing two headline grabbing projects: the Denver Airport and the FAA's Advanced Automation System. He argued that the software problems of these two efforts are overshadowed by the successes of the software. In Denver, for example, three airplanes can simultaneously land at the airport and be guided to their gates by a sophisticated lighting system; both advances made possible by software.
Mr. Mosemann addressed U.S. national capabilities in software and stated that a "software engineering mentality" is needed in the nation's software community. Building better software requires control of the development process, engineering, and measurement. The Department of Defense's (DoD) promotion of software engineering university curricula, Ada 9X, and software acquisition "best practices" are likely to improve U.S. software capabilities.
Mr. Mosemann concluded his remarks by announcing the availability of Air Force software guidelines: "Guidelines for Successful Acquisition and Management of Software Intensive Systems: Weapons Systems, Command & Control Systems, and Management Information Systems," available from the Software Technology Support Center (STSC).
Mr. John Marciniak of Kaman Sciences Corporation provided a progress report on the formation of the National Software Council (NSC). He reported that the NSC Planning Task Force formed at Cooperstown I briefed Mr. Mosemann on 10 March 1994. Subsequently, the National Security Industrial Association (NSIA) was invited to continue the process of forming the NSC.
Ms. Judith Brodman and Ms. Donna Johnson of LOGOS International, Inc. presented the results of a survey to identify sources of Return On Investment (ROI) data for software measurement and process improvement programs. They found that organizations with higher Software Engineering Institute (SEI) Capability Maturity Model (CMM) ratings tended to collect data usable in calculating ROI. The speakers noted the need for additional work in collecting and building a database of ROI data.
Ms. Lorraine Martin and Mr. Scott Hissam of Unisys presented a status report on the development of the National Software Data and Information Repository (NSDIR) Order Prototype. The prototype was constructed by integrating Commercial Off The Shelf (COTS) tools and initially populated with data from the NASA Software Engineering Laboratory (NASA/SEL). The Unisys team surveyed the community for input into the project during development of the Prototype.
Dr. Richard Selby of the University of California - Irvine presented the draft "National Software Data and Information Repository (NSDIR) Strategic Plan." The draft plan was developed in conjunction with the NSDIR OrderĐI Prototype. Dr. Selby presented the goals and stakeholders in the project, the framework for the Repository, and issues for consideration within Cooperstown II's Working Group discussions.
Dr. Herbert Hecht of SoHaR Inc. and Dr. Barry Boehm discussed the motivation and likely issues of interest to an NSDIR User's Group. Dr. Boehm presented some interesting data from surveys performed by affiliates of the COCOMO 2.0 Program. The data indicated that affiliates preferred that access to the COCOMO 2.0 database be limited to Repository staff members. Affiliates will only be able to obtain restricted statistics, thereby protecting the identity of data contributors.
Mr. Errol Shim of U.S. Shoe discussed "lessons learned" in building the International Function Points User Group (IFPUG). Mr. Shim represented a unique perspective at the workshop as a non-high-technology, non-software-based commercial business. The emphasis of Mr. Shim's talk was devoted to IFPUG. His major point is that we should think of ourselves as in the "change business," not the "technology business" or the "measurement business." Mr. Shim emphasized the need for an appropriate level of funding, noting that "You get what you pay for."
Mr. Andrew Chruscicki of Rome Laboratory spoke on "A Concept of and Strategy for Building the National Center for Software Data and Information." Mr. Chruscicki contends that the NCSDI will address the need for a measurement-based infrastructure to facilitate software acquisition, production, and utilization in the United States.
Working Groups' Recommendations
Workshop attendess participated in one of several Working Groups. A number of recommendations regarding the Cooperstown Initiative were generated by the groups. Consensus appeared to exist in the following areas:
In addition, trial applications on systems already using metrics for decision-making and forming an interim User's Group were proposed. Other suggestions included a variety of organizational and institutional structures in which to embed the NSDIR, particularly a National Center for Software Data and Information (NCSDI). In general, workshop participants appeared to favor expanding participation in the NSDIR to include commercial users other than defense contractors.
In Conclusion
Mr. Mosemann thanked the Working Groups for their contributions. He tasked the group to produce minutes of Cooperstown II and pledged to continue the Cooperstown Initiatives in directions recommended by the Working Groups. He promised to initiate the NCSDI, begin an NSDIR User's Group, and brief key players in government on the Cooperstown Initiative.
Additional Information on the Cooperstown Initiative can be obtained from: