**9. Formulating system dynamics as a process**

Implementing an effective measurement program as a meta-structure of its parent system is a big challenge. And overcoming these challenges is worthwhile, because such measures provide insights into the system behavior. Such implementation also improves the ability to plan and track systems evolutionary progress and needed for addressing risks/problems much earlier (Clark, 2002). Therefore, we need to have an underlying methodology that can be applied to formulate a process for evolvable systems. Typically this would include the study and observation of selected set of metrics, development track history, changes in usage track history, and so on. In the previous sections of this chapter, we attempted to explain the necessary elements of such a process in detail. Although the resulting evolution of software mostly appears to be driven and controlled by human decisions, managerial edict and developer's judgement, yet a high-level generalized pseudo code for such a process can be developed, as presented in fig. 8. This is in conformance with the system dynamics approach suggested by Richardson (Richardson, 1999).
