The Short note on the Does Distributed Development Affect Software Quality?
An Empirical Case Study of Windows Vista. Is as follows.
In there study we divide binaries based on the level of geographic dispersion of their commits. We studied the post release failures for the Windows Vista code base and concluded that distributed development has little to no effect. We posit that this negative result is a significant finding asit refutes, at least in the context of Vista development, conventional wisdom and widely held beliefs about distributed development. When coupled with prior work, our results support the conclusion that there are scenarios in which distributed development can work for large software projects. Based on earlier work our study shows that Organizational differences are much stronger indicators of quality than geography. An Organizational compact but geographically distributed project might be better than an geographically close organizationally distributed project. We have presented a number of observations about the development practices at Microsoft which may mitigate some of the hurdles associated with distributed development, but no causal link has been established. There is a strong similarity between these practices and those that have worked for other teams in the past as well as solutions proposed in other work. Directly examining the effects of these practices is an important direction for continued research in globally distributed software development.
In this paper, we have presented a case study on the communication of distributed projects. The study shows that there is no significant difference in the amount of communication of two-location projects and three-location projects. The results show a trend that the amount of communication in two location projects is higher than the amount of communication in three-location projects. The study also analyzes the effect of time zone differences classifying the projects in three time ranges: large, medium, and small. The result also shows no significant difference in the communication of projects in these time zone ranges; however, the data shows a trend towards more communication in the small time zone range. We also analyzed the reply time for e-mails in projects located in different time zones. We found that in projects located in the small time zone range, the reply time for emails was faster than in projects located in the large time zone range. Our results indicate a trend on the effect of communication, however, the analysis reveal that the differences are not significant. As future work, we plan to collect more data in the next editions of DOSE in order to have more reliable data. We also plan to extend the study and compare the distributed projects with projects developed in a single location. Additionally, we want to study the quality of the produced software and compare the number of failures in local projects to two-location projects and three-location projects. For future studies, we will keep the classification by time zone ranges. Acknowledgments: We would like to thank all the people involved in DOSE: Do Le Minh, Franco Brusatti, Giordano Tamburrelli, Huynh Quyet Thang, Lajos Kollar, Mei Tang, Natalia Komlevaja, Nazareno Aguirre, Peter Kolb, Raffaela Mirandola, Sergey Karpenko, Sungwon Kang, Victor Krisilov, Viktor Gergel; and the students who took the course