Browsing by Author "Äärilä, Mika"

Designing software architectures requires finding a balance between different architectural qualities. Architectures should therefore be assessed on a scale of suitability on how well the architecture corresponds to the requirements of the target system. This suitability can be subjected to critical evaluation. In this study, we explore software architecture evaluation methods with the intent of finding out what can be said about the methods in general, what kinds of resources they require, and what types of benefits they provide. We accomplish this by analysing research articles on a variety of software architecture evaluation case studies through the lens of a systematic literature review. We discover that software architecture evaluation is a process that predominantly consists of pre-evaluation, evaluation and post-evaluation phases, all of which have different requirements not only for the evaluator but also for the stakeholders participating in the evaluation. The benefits of the evaluations vary. The majority of evaluations finds issues on the architecture they are evaluating, and notably often the evaluations receive positive feedback on their innate quality of being a way to share information and enable discussion. Ultimately, software architectures evaluations are a human endeavor that focuses on discussions surrounding the object of the evaluation. Those running the evaluation should be skilled in facilitating this discussion with the help of the structures provided by the software architecture evaluation methods.