Developing a usable metric toolkit for a commercial environment

Dictates of prevalent development cultures often prevent important progress in the development of software quality metrics being used in commercial software environments. Comparison of data between projects is discouraged due to differences in development environments. Programmers decline to count errors, employees regard error statistics as performance appraisal weapons, and costs to set up baselines are high. The author shows how to alleviate these problems and presents some experiences in two commercial projects conducted in separate corporations, thereby illustrating how simple quality metrics can be developed and used. >