Towards migration-aware filtering in model differences application

Abstract

The management of consistency among modelling artefacts is of crucial importance in model-driven engineering. Especially in distributed development, refinements of both metamodels and models are usually performed in a concurrent and misaligned manner, thus breaking consistency among model versions. Inconsistency situations become harmful when propagating changes from a local workspace to the shared repository or vice versa. The changes propagation can be achieved through model differences detection and application, exploiting filtering mechanisms when migration is not permitted. Nevertheless, loss of information due to metamodel evolutions may occur when filtering differences between models conforming to different versions of the modelling language. In this work we propose to minimise this loss of information by enhancing the filtering mechanism to take into account metamodel evolution information.