A policy investigation model for long-term software evolution processes
Paul Wernick · 2004
We describe a system dynamics simulation of a long-term software evolution process. The model has been deliberately kept as simple as possible, to allow for easier calibration. The simulation model is intended to allow the examination of potential effects on software system size of proposed software process changes. The model structure incorporates a number of policy variables whose values can be set to represent the effects of such process changes. Results are presented showing the calibration of the modelled trend in system size growth over 13 years against an actual trend. The sensitivity of the model trend to changes in the values of the policy inputs is also shown. An example of a process with improved fault detection is presented.