Evolutionary togetherness: how to manage coupled evolution in metamodeling ecosystems

In Model-Driven Engineering (MDE) metamodels are cornerstones for defining a wide range of related artifacts interlaced with explicit or implicit correspondences. According to this view, models, transformations, editors, and supporting tools can be regarded as a whole pursuing a common scope and therefore constituting an ecosystem. Analogously to software, metamodels are subject to evolutionary pressures too. However, changing a metamodel might compromise the validity of the artifacts in the ecosystem which therefore require to co-evolve as well in order to restore their validity. Different approaches have been proposed to support at different extent the adaptation of artifacts according to the changes operated on the corresponding metamodels. Each technique is specialized in the adaptation of specific kind of artifact (e.g., models, or transformations) by forcing modelers to learn different technologies and languages. This paper discusses the different relations occurring in a typical metamodeling ecosystem among the metamodel and the related artifacts, and identifies the commonalities which can be leveraged to define a unifying and comprehensive adaptation process. A language and corresponding supporting tools are also proposed for the management of metamodel evolution and the corresponding togetherness with the related artifacts.