Open source software products have recently garnered a lot of attention from both academia and industry. As the authors indicate, there are many unproven hypotheses on open source software products: that they foster faster system growth; foster more creativity; lead to a high success rate, due to their simplicity; typically have fewer defects, since defects are found and fixed more rapidly; and are more modular. These common beliefs, in fact, might have misled practitioners when planning software development strategies. To quantitatively investigate these common perceptions, an empirical study, using three open source and three closed source software systems, was conducted.
The authors first propose a set of metrics for their quantitative study, by revising existing ones. They then design their research by analyzing the source code archives of three closed source projects. Their findings support two of these five common beliefs. The validated perceptions are that open source projects foster more creativity, and open source software products lead to fewer defects. The other three beliefs fail to hold in their investigation.
In my opinion, this study is questionable: the revised metrics, sample data, and sample size are either superficial or insufficient. If one has the same project implemented using both open source and closed source software products, one could use the metrics to measure that product’s complexity, creativity, simplicity, and modularity. Note that the authors also recommend, when implementing an open source project, that practitioners not take these common beliefs for granted, since they might be false under certain circumstances.