By Carlo Montangero

Dept. of Informatics

University of Pisa

Bid name An inquiry on non-functional requirements vs software architecture.

In the original plan the visit should end on the 10. The added day allowed me to partecipate to the initial phase of another visit by RENOIR partners, namely Gerti Kappel and Werner Retschitzegger of the University of Linz, Austria. This gave the opportunity to discuss the outcome of my visit in a broader context, and get some feedback.

The main outcome of the visit follows from interactions with Anthony Finkelstein and Wolfgang Emmerich. It is a research agenda in the novel field of architecture stability, bridging the areas of requirements engineering and software architecture. Such an investigation stems from the need to understand when a representation of a software system is stable in front of the unavoidable changes in the requirements. We expect the software industry to benefit by the results of such an investigation, since they should support better practices in the early design phases. Indeed, these practices will lead to the development of systems that are more robust with respect to the anticipated changes in the requirements.  

There are three hypothesis that we want to verify with this research. First, that non-functional requirements, which express the requested quality of service, play an essential role in the choice of the architecture. Second, that architectural representations can be be devised, which are suitable for annotations that lend to a characterization of the quality of service that the system can deliver. Third, that the resulting characterizations of the requested and delivered quality of service can be compared, in order to assess the stability of an architecture, with respect to the most likely changes in the requirements.

The research agenda has been generated by iterating through brain-storming meetings, and drafting and reviewing of the agenda. These activities account for most of my time during the visit. I also had the opportunity to discuss research directions with other people in Anthony's team, like Ernst Ellmer and others.