Software developing communities are shifting to open source software (OSS) because of the reason that software development takes place in successive releases, thereby improving its quality and reliability. Multi-release development of OSS can provide an opportunity to inculcate the dynamic needs of the user in a very short span of time to survive in the market. In spite of having these benefits, numerous challenges can be faced during the multi-release OSS development. Some of the challenges can be the generation of errors during the addition of new features. To address the changing fault detection process, a change point phenomenon is considered so as to give more practicality to the model. In this article, we present a general framework for multi-release OSS modelling incorporating imperfect debugging and change points. Parameter estimation and model validation is done on the three releases of Apache, an open source software project.
Modelling Failures Occurrences of Open Source Software with Reliability Growth
