The Tool Generation Challenge for Executable Domain-Specific Modeling Languages

Executable domain-specific modeling languages (xDSMLs) have the potential of bringing major benefits to the development of complex software-intensive systems as they provide abstractions of complex system behaviors and allow for early analyses of that behavior. However, in order to be useful, xDSMLs have to be equipped with model analysis tools supporting domain engineers in comprehending, exploring, and analyzing modeled behaviors. Hand-crafting such tools in an ad hoc manner imposes significant efforts and costs on the development process and is, hence, mostly done for broadly adopted xDSML only. Executable metamodeling approaches seek to overcome this limitation by providing formalisms to define the execution semantics of xDSMLs in a systematic way building the basis for automatically generating model analysis tools. While significant advances towards achieving this vision have been achieved in recent years, there are still many challenges to be solved for generating out-of-the-box analysis support for xDSMLs. In this paper we revisit the tool generation challenge introduced by Bryant et al. [3] seven years ago reflecting on recent achievements and identifying open challenges.