Measuring Globally Distributed Software Development

Over the last few years software project support tool repositories such as version control and problem tracking systems are increasingly being used for empirical studies of software. This has been primarily driven by the interest in open source projects and the wide availability of their repositories.  The course exemplifies this approach with two industry applications developed for globally distributed development teams. The first application is a method to find parts of source code that tend to be changed independently by distributed teams and the second application is a tool that shows relationships among developers, organizations and the source code they work on. The expected outcome of this course is to expose participants to the potential capabilities of such a measurement framework and to provide examples of industry-relevant software engineering scenarios.

Readings:

Mockus, A. and Herbsleb, J. D. (2002). Expertise Browser: A Quantitative Approach to Identifying Expertise. In Proceedings, International Conference on Software Engineering, Orlando, FL, pp. 503-512.

Mockus, A. and Weiss, D. M. Globalization by Chunking: A Quantitative Approach. IEEE Software, January – March (2001).

Guest Lecturer: Audris Mockus