Software Language Evolution

Abstract

By abstraction and factoring out domain specific knowledge, model driven engineering addresses the problem of increasing software complexity. Both models and meta models are generally subject to evolution, yet evolution of a meta model can cause conforming models to no longer conform and thereby no longer be usable. Therefore, models need to be migrated to reflect changes to their meta models. As evolution is typically frequent and reoccurring, manual migration of models is cumbersome and holds back the development process, yet automatic support is generally lacking. In this research we identify the problems caused by meta model evolution and develop methodologies and tools to solve these by supporting meta model evolution generically and automatically.