It is inevitable that most software systems will undergo continuous changes, increasing the risk of architectural degeneration. Several software architectural evaluation processes attempt to address architectural issues from various perspectives and assist to mitigate the risk of architectural degeneration. This paper describes enhancements made to an existing software architecture evaluation process and presents the results of a case study conducted using the improved process. The results suggest that the enhanced software architecture evaluation process is more broadly applicable than the previous version. It is believed that this process would be effective with systems utilizing various architectural styles and design patterns.

Industry often complains that current university curricula fail to address the practical issues of real software development. This paper outlines a proposal for an innovative core curriculum for a Bachelor of Science in computer science. The proposed core curriculum contains elements of traditional computer science programs combined with software engineering via a team-oriented, hands-on approach to large-scale software development. In addition to traditional lecture/project/exam courses, students are required to take an eight-semester sequence of "Software Factory" courses. Software Factory courses put the students' newly acquired skills to work in a real software organization staffed and managed by all students in the program. Students from all courses in the Software Factory sequence meet simultaneously to fulfil their roles in the software organization. We expect the students will be better-prepared software engineering practitioners after completing a curriculum that combines traditonal courses with practical Software Factory experience.

Reducing software development cycle time without sacrificing quality is crucial to the continued success of most software development organizations. Software companies are investing time and money in reengineering processes incorporating improvements aimed at reducing their cycle time. Unfortunately, the impact of process improvements on the cycle time of complex software processes is not well understood. The objective of our research has been to provide decision makers with a model that will enable the prediction of the impact a set of process improvements will have on their software development cycle time. This paper describes our initial results of developing such a model and applying it to assess the impact of software inspections. The model enables decision makers to gain insight and perform controlled experiments to answer "What if?" type questions, such as, "What kind of cycle time reduction can I expect to see if I implement inspections?" or "How much time should I spend on inspections?".