A comparison of software project architectures : agile, waterfall, spiral, and set-based

Abstract

Engineers and managers of software projects have the daunting task of successfully delivering the right product at the right time for the right price. The path to achieving these lofty goals is commonly not a straightforward endeavor. Due to the dynamic nature of software development, varying organizational circumstances, and situational idiosyncrasies of each project this can be a very difficult and sensitive process. Ideally, software development methodologies bring order to the chaos of software development. An ill-fitting development strategy, however, can create unnecessary friction and further complicate the prospect of a successful product delivery. Researchers and private organizations alike spend substantial resources to understand the strengths and weaknesses of commonly used development practices - the validation of which is highly problematic due to conflicting variables. This research ventures to bring clarity to the question: "Which development methodology is right for a particular situation?" Treating the software development project life-cycle as a socio-technical system, it can be decomposed to the most fundamental elements. Using these elements as the architectural building blocks of a project, Agile, Waterfall, Set-Based, and Spiral are each compared at the molecular level. This thesis evaluates these comparisons and how subsequent research applies to today's software projects.