i* or Not i*, This Was the Question

We applied goal-oriented modelling with i* in several projects in our 20-year or more research activities. In this chapter, we propose an a posteriori reflection and revisit our choice to use or not i* in representative projects, in light of modelling language’s qualities, which are well accepted in research literature. When performing requirements engineering for software systems, goal-modelling provided us conceptual tools for the analysis of alternative ways to reach domain stakeholders’ goals and for making explicit their mutual dependencies. Thus we were able to experience the pragmatic qualities of the models we built. The integration of goal-oriented modelling with other elicitation and reasoning techniques was relevant to consolidate and validate requirements. In some cases, to increase the language’s domain appropriateness quality, we extended the language adding new constructs. We shortly report about evaluation studies that have performed to assess quality properties of the proposed language’s extensions. Concluding remarks on our experience are proposed.