Selection of Open Source Components – A Qualitative Survey in Norwegian IT Industry

Open-source-software (OSS) becomes more and more used in software products developed by the industry. The motivation for this can e.g. be to reduce development/maintenance costs and to shorten time-to-market. There are several technical methods for selecting and evaluating OTS (Off-the-Shelf) components, but research shows that these methods are rarely used. OTS components come in two types: COTS (Commercial-Off-The-Shelf) and OSS (Open-Source-Software); this report focuses on OSS. The amount of systems with open source components is large and there are many questions to be asked such as: Where and how to find OSS components? How to evaluate them? How to learn the components? How to take care of knowledge about the chosen component? How to make best use of OSS? What component's version should be chosen and what could be done with respect to the maintenance and the new versions of the component? Without an effective way to select and evaluate OSS components, the time spent choosing the correct components may offset the advantages of using them. This report presents the current state-of-the-art, a research design, and then presents the results of a survey performed in collaboration with Norwegian IT companies. The survey is based on the results of an explorative survey performed in the autumn of 2006; the interview guide was improved and we performed a larger survey. The state-of-the-art focuses on investigating existing research related to software development processes for component based-systems and OTS selection processes with the goal of finding research questions for research interviews. The results of the interviews may help software companies to understand the existing practices and improve their own processes. This thesis presents two exploratory research goals and seven research questions. The research goals are: 1) to explore the company role: integrator of open source components, and 2) to explore their development processes with focus on OSS selection processes. The contributions of this work include a literature study, and new empirical knowledge about OSS selection. This knowledge refers to the qualitative answers to the research questions. We have performed a qualitative descriptive study.Preface This report has been written during spring 2007, as part of the course TDT4900 Software engineering, master thesis at the Norwegian University of Science and Technology (NTNU). The context of the work is within the European ITEA project, COSI (Co-development using inner & Open source in Software Intensive products). The Norwegian COSI wants to enable the Norwegian IT sector to fully …