On understanding software quality evolution from a defect perspective: A case study on an open source software system

Abstract

Software systems need to continually evolve during their life cycle. It is, therefore, important to monitor how their quality evolves so that quality assurance activities can be properly planned. In this paper, we analyze empirically the quality evolution of an open source software system (Apache Tomcat). We address software quality from an external perspective. We used the number of defects as a quality indicator. We wanted to investigate if the Qi (Quality Assurance Indicator) metric, which we proposed in a previous work, can be used to observe how quality, measured in terms of defects, evolves in the presence of changes. We performed an empirical analysis using historical data collected from the subject system covering a period of more than seven years (thirty-one versions). Results are reported and discussed in the paper.