Keywords:
Astract: Predefined composite operations are handy for efficient modeling, e.g., for the automatic execution of refactorings, and for the introduction of patterns in existing models. Some modeling environments provide an initial set of basic refactoring operations, but hardly offer any extension points for the user. Even if extension points exist, the introduction of new composite operations requires programming skills and deep knowledge of the respective metamodel. In this paper, we introduce a method for specifying composite operations within the user´s modeling language and environment of choice. The user models the composite operation by-example, which enables the semi-automatic derivation of a generic composite operation specification. This specification may be used in various modeling scenarios, like model refactoring and model versioning. We implemented the approach in the Operation Recorder and performed an evaluation by defining multiple complex refactorings for UML diagrams.

Keywords:
Astract: The standardized QVT Relations language, one cornerstone of Model-Driven Architecture (MDA), has not yet gained widespread use in practice, not least due to missing tool support in general and inadequate debugging support in particular. Transformation engines interpreting QVT Relations operate on a low level of abstraction, hide the operational semantics of a transformation and scatter metamodels, models, QVT code, and traces across different artifacts. We propose a model-based debugger representing QVT Relations on bases of TROPIC, a model transformation framework which utilizes a variant of Colored Petri Nets (CPNs) providing an explicit runtime model and a homogenous view on all artifacts of a transformation.

Keywords:
Astract: The status of current model-driven engineering technologies has matured over the last years whereas the infrastructure supporting model management is still in its infancy. Infrastructural means include version control systems, which are successfully used for the management of textual artifacts like source code. Unfortunately, they are only limited suitable for models. Consequently, dedicated solutions emerge. These approaches are currently hard to compare, because no common quality measure has been established yet and no structured test cases are available. In this paper, we analyze the challenges coming along with merging different versions of one model and derive a first categorization of typical changes and the therefrom resulting conflicts. On this basis we create a set of test cases on which we apply state-of-the-art versioning systems and report our experiences.

Keywords:
Astract: Traditionally, models are considered as pretty pictures supporting merely the documentation of a software development project. With the rise of model-driven engineering (MDE) this viewpoint has to be reconsidered. Models become first-class artifacts which yield the basis for the generation of executable program code. In modern university curricula of computer science and related fields this paradigm shift must not be ignored. At the Business Informatics Group (BIG) of the Vienna University of Technology we offer an advanced modeling course called Model Engineering where we elaborate current trends, development, and state-of-the-art techniques necessary to realize the visions of MDE. In this paper we report which concrete concepts and approaches we teach and how we structure and organize a practical hands-on lab where the students have to build their own model-driven development environment consisting of their own modeling languages, certain types of model transformations, and code generation facilities.

Keywords:
Astract: Approximately 1000 students of computer science and business informatics attend the course Introduction to Object- Oriented Modeling (OOM) offered by the Business Informatics Group (BIG) of the Vienna University of Technology each year in order to learn the basic concepts of the Unified Modeling Language (UML) and to obtain a certificate. For finishing the course successfully, the students must pass three small tests where they have to prove their theoretical knowledge about UML concepts as well as their ability to apply this knowledge in practical exercises. In this paper we report our experiences in assessing the modeling knowledge of our students and we reveal how we design the tests.