Automated Software Metrics, Repository Evaluation and Software Asset Management: New Tools and Perspectives for Managing Integrated Computer Aided Software Engineering (I-Case)

Abstract

Automated collection of software metrics in computer aided software engineering (CASE) environmentsopens up new avenues for improving the management of software development operations, as well asshifting the focus of management's control efforts from "software project� to "software assets" stored in acentralized repository. Repository evaluation, a new direction for software metrics research in the 1990s,promises a fresh view of software development performance for a range of responsibility levels. We discussthe automation of function point and code reuse analysis in the context of an integrated CASE (I-CASE)environment deployed at a large investment bank in New York City. The development of an automatedcode reuse analysis tool prompted us to re-think how to measure and interpret code reuse in the I-CASEenvironment. The metrics we propose describe three dimensions of code reuse -- leverage, value andclassification -- and we examine the value of applying them on a project and a repository-wide basis.