A decisional framework for legacy system management

Making a decision about how to evolve a legacy system cannot be made spontaneously; rather, it requires a decisional framework that takes into account several factors including software value, risk analysis, and cost estimation. We present a decisional framework to manage legacy systems that exploits an assessment model and a taxonomy of maintenance interventions a legacy system can undergo during its life-cycle. The decisional framework has been defined within a pilot project involving a major international software enterprise. The project aims at assessing and improving the current practices of the organization and at experimenting software maintenance processes conducted by teams distributed at different sites in a cooperative networking environment.