Open Source Research Opinion: Explorative Tools and Process Research

We will be participating in UC Irvine’s workshop on the future of open source research. The organizers asked participants to provide opinions on three research areas that warrant further attention. Here is our second one:

2. Improved Open Source Process and Tooling

Quantitative analyses of open source projects can only show how things are done today. But what about tomorrow? Where do groundbreaking new practices and tools come from? Open source software development, when compared with other approaches like plan-driven and agile methods, has many advantages but also is clearly lacking in many dimensions. Most of these problems surface, when developers are no longer “scratching their own itch”. In such situations, more traditional methods of product and project management can and should be applied.

Thus, we need to see which practices of plan-driven and agile methods can be applied and carried over to open source software development. Some agile methods practices, for example, seem easy to apply: Refactoring, test-driven development, and continuous integration, just to name a few, seem compatible with an open source value system. Nevertheless, they pose research questions, for example, why open source projects tend to prefer code review or user testing over test-driven development. Other practices like consistent product management are virtually unheard of in open source software development. Here open source can learn and benefit a lot.

Thus, the research question is: Which existing traditional plan-driven and agile methods practices are compatible with open source software development, would be beneficial to be integrated, and how should they be integrated? Such research work is highly empirical qualitative hypothesis driven and can build on the prior mentioned quantitative analyses. The end result could be a new process framework combining the best of plan-driven, agile methods, and open source software development.