The recent success of open source software has meant that there is a need to further understand the way it is devel oped, and in doing so identify best practice in open source software development. Agile software development is the term used to describe and promote a variety of software de velopment methodologies designed for use in commercial projects with unstable requirements and tight time frames. Agile methods are often termed "lightweight methods" in that they downplay the importance of traditional labor in tensive software engineering practices such as exhaustive requirements specifications, detailed design and exhaustive documentation. This paper argues that the open source software development process is in fact a special type of agile method.
Software development is a knowledge intensive process and the information generated in open source software development projects is typically housed in a central Internet repository. Open source repositories typically contains vast amounts of information, much of it unstructured, meaning that even if a question has previously been discussed and dealt with it is not a trivial task to locate it. This can lead to rework and confusion amongst developers and possibly deter new developers from getting involved in the project in the first place. This paper will present the case for an open source software development ontology. Such an ontology would enable better categorization of information and the development of sophisticated knowledge portals in order to better organize community knowledge and increase efficiency in the open source development process.