In model-driven development (MDD), domain-specific modeling languages (DSMLs) are used as tailor-made software languages targeting dedicated application domains. Due to the narrow domain coverage of DSMLs, demands to integrate their individual functionality into a consolidated DSML arise (e.g., developing a software product combining two or more pre-existing DSMLs). However, in order to realize the benefits of integrated DSMLs, it must be ensured that the integrated DSML is correctly implemented and behaves as specified. To support the integration and the test of DSMLs, this thesis presents an approach targeting the Meta Object Facility (MOF) and the Unified Modeling Language (UML)- a metamodeling infrastructure frequently employed for the MDD of software systems. The integration of DSMLs is based on a rewriting technique for model-to-text (M2T) transformations. This method allows for the reuse as well as for the automatic refactoring of M2T transformation templates to fix important syntactical mismatches between templates and the integrated DSML. To test an integrated DSML, scenarios are used to define domain requirements on an abstract level (via structured text descriptions). In a subsequent step, executable scenario tests are derived from the requirements-level scenarios. These executable scenario specifications are then employed to test the integrated DSML for compliance with corresponding domain requirements. Empirical evaluations of the approach (case studies, controlled experiment) demonstrate its successful application, collect evidence for its usefulness, and quantify its benefits. The integrated proof-of-concept implementations build on the Eclipse Modeling Framework (EMF), making use of and extending well-known Eclipse-based projects. All accompanying developments are placed into the public domain as free/libre open source software. Within the framework of this thesis, research results were originally published as individual contributions (workshop, conference, and journal articles). All research contributions are results of applying a design science research approach. (author's abstract)

In model-driven development (MDD), domain-specific modeling languages (DSMLs) act as a communication vehicle for aligning the requirements of domain experts with the needs of software engineers. With the rise of the UML as a de facto standard, UML/MOF-based DSMLs are now widely used for MDD. This paper documents design decisions collected from 90 UML/MOF-based DSML projects. These recurring design decisions were gained, on the one hand, by performing a systematic literature review (SLR) on the development of UML/MOF-based DSMLs. Via the SLR, we retrieved 80 related DSML projects for review. On the other hand, we collected decisions from developing ten DSML projects by ourselves. The design decisions are presented in the form of reusable decision records, with each decision record corresponding to a decision point in DSML development processes. Furthermore, we also report on frequently observed (combinations of) decision options as well as on associations between options which may occur within a single decision point or between two decision points. This collection of decision-record documents targets decision makers in DSML development (e.g., DSML engineers, software architects, domain experts). (authors' abstract) / Series: Technical Reports / Institute for Information Systems and New Media

In model-driven development (MDD), domain-specific modeling languages (DSMLs) act as a communication vehicle for aligning the requirements of domain experts with the needs of software engineers. With the rise of the UML as a de facto standard, UML/MOF-based DSMLs are now widely used for MDD. This paper documents design decisions collected from 90 UML/MOF-based DSML projects. These recurring design decisions were gained, on the one hand, by performing a systematic literature review (SLR) on the development of UML/MOF-based DSMLs. Via the SLR, we retrieved 80 related DSML projects for review. On the other hand, we collected decisions from developing ten DSML projects by ourselves. The design decisions are presented in the form of reusable decision records, with each decision record corresponding to a decision point in DSML development processes. Furthermore, we also report on frequently observed (combinations of) decision options as well as on associations between options which may occur within a single decision point or between two decision points. This collection of decision-record documents targets decision makers in DSML development (e.g., DSML engineers, software architects, domain experts). / Series: Technical Reports / Institute for Information Systems and New Media

In model-driven development (MDD), domain-specific modeling languages (DSMLs) act as a communication vehicle for aligning the requirements of domain experts with the needs of software engineers. With the rise of the UML as a de facto standard, UML/MOF-based DSMLs are now widely used for MDD. This paper documents design decisions collected from 90 UML/MOF-based DSML projects. These recurring design decisions were gained, on the one hand, by performing a systematic literature review (SLR) on the development of UML/MOF-based DSMLs. Via the SLR, we retrieved 80 related DSML projects for review. On the other hand, we collected decisions from developing ten DSML projects by ourselves. The design decisions are presented in the form of reusable decision records, with each decision record corresponding to a decision point in DSML development processes. Furthermore, we also report on frequently observed (combinations of) decision options as well as on associations between options which may occur within a single decision point or between two decision points. This collection of decision-record documents targets decision makers in DSML development (e.g., DSML engineers, software architects, domain experts). (authors' abstract) / Series: Technical Reports / Institute for Information Systems and New Media

In the process of model-driven development (MDD) of software artifacts, domain-specific modeling languages (DSMLs) are an integral part. They act as the communication vehicle for aligning the requirements of the domain expert with the needs of the software engineer. With the rise of the UML as de facto standard for modeling software systems, MOF/UML-based DSMLs are now widely used for MDD. This paper documents design decisions from ten DSML projects which are based on the MOF/UML and which we conducted over the last years. We present our experiences in the form of reusable decision templates for all decision points detected in each phase of the DSML development process. Furthermore, we report also on identified decision dependencies which may occur within a single decision or between two decisions. (author's abstract) / Series: Technical Reports / Institute for Information Systems and New Media

Background: A key idea of feature orientation is to decompose a software product line along the features it provides. Feature decomposition is orthogonal to object-oriented decomposition it crosscuts the underlying package and class structure. It has been argued often that feature decomposition improves system structure (reduced coupling, increased cohesion). However, recent empirical findings suggest that this is not necessarily the case, which is the motivation for our empirical investigation. Aim: In fact, there is little empirical evidence on how the alternative decompositions of feature orientation and object orientation compare to each other in terms of their association with observable properties of system structure (coupling, cohesion). This motivated us to empirically investigate and compare the properties of three decompositions (object-oriented, feature-oriented, and their intersection) of 28 feature-oriented software product lines. Method: In an exploratory, observational study, we quantify internal attributes, such as import coupling and cohesion, to describe and analyze the different decompositions of a feature-oriented product line in a systematic, reproducible, and comparable manner. For this purpose, we use three established software measures (CBU, IUD, EUD) as well as standard distribution statistics (e.g., Gini coefficient). Results: First, feature decomposition is associated with higher levels of structural coupling in a product line than a decomposition into classes. Second, although coupling is concentrated in feature decompositions, there are not necessarily hot-spot features. Third, the cohesion of feature modules is not necessarily higher than class cohesion, whereas feature modules serve more dependencies internally than classes. Fourth, coupling and cohesion measurement show potential for sampling optimization in complex static and dynamic product-line analyses (product-line type checking, feature-interaction detection). Conclusions: Our empirical study raises critical questions about alleged advantages of feature decomposition. At the same time, we demonstrate how the measurement of structural attributes can facilitate static and dynamic analyses of software product lines. (authors' abstract) / Series: Technical Reports / Institute for Information Systems and New Media

Background: A key idea of feature orientation is to decompose a software product line along the features it provides. Feature decomposition is orthogonal to object-oriented decomposition it crosscuts the underlying package and class structure. It has been argued often that feature decomposition improves system structure (reduced coupling, increased cohesion). However, recent empirical findings suggest that this is not necessarily the case, which is the motivation for our empirical investigation. Aim: In fact, there is little empirical evidence on how the alternative decompositions of feature orientation and object orientation compare to each other in terms of their association with observable properties of system structure (coupling, cohesion). This motivated us to empirically investigate and compare the properties of three decompositions (object-oriented, feature-oriented, and their intersection) of 28 feature-oriented software product lines. Method: In an exploratory, observational study, we quantify internal attributes, such as import coupling and cohesion, to describe and analyze the different decompositions of a feature-oriented product line in a systematic, reproducible, and comparable manner. For this purpose, we use three established software measures (CBU, IUD, EUD) as well as standard distribution statistics (e.g., Gini coefficient). Results: First, feature decomposition is associated with higher levels of structural coupling in a product line than a decomposition into classes. Second, although coupling is concentrated in feature decompositions, there are not necessarily hot-spot features. Third, the cohesion of feature modules is not necessarily higher than class cohesion, whereas feature modules serve more dependencies internally than classes. Fourth, coupling and cohesion measurement show potential for sampling optimization in complex static and dynamic product-line analyses (product-line type checking, feature-interaction detection). Conclusions: Our empirical study raises critical questions about alleged advantages of feature decomposition. At the same time, we demonstrate how the measurement of structural attributes can facilitate static and dynamic analyses of software product lines. (authors' abstract) / Series: Technical Reports / Institute for Information Systems and New Media