Requirements engineering is a core activity in software engineering. However, formal requirements engineering methodologies and documented requirements are often missing in scientific computing projects. We claim that there is a need for methodologies, which capture requirements for scientific computing projects, because traditional requirements engineering methodologies are difficult to apply in this domain.

Many research approaches in software engineering propose models or extensions to existing models. This includes system models, such as UML models, and project management artifacts, such as bug reports or tasks. For evaluating a research approach these models have to be implemented. Existing tools often lack the required extensibility and do not support data collection and analysis. Therefore, many researchers have to spend significant effort on the implementation. In this paper, we identify common requirements for a platform to build research tools on and we evaluate them based on potential use for recent publications. To evaluate the relevancy of the requirements we reviewed 274 research papers from the software engineering category of two major conferences over the last two years. We found 38 research approaches among those that would have benefited from a common platform with the presented requirements.

In recent years, models are increasingly used throughout the entire lifecycle in software development projects. In effect, the need for collaborating on these models emerged, requiring change tracking and versioning. However, many researchers have shown that existing methods and tools for Version Control (VC) do not work well on graph-like models, such as UML, SysML or domain-specific modeling languages. To alleviate this, alternative techniques and methods have been proposed which can be classified into state-based and operation-based approaches. Existing research shows advantages of operation-based over state-based approaches in selected use cases, such as conflict detection or merging. However, there are only few results available on the advantages of operation-based approaches in the most common use case of a VC system: review and understand change. In this paper, we present and discuss both approaches and their use cases. Moreover, we present the results of an empirical study to compare a state-based with an operation-based approach for the use case of reviewing and understanding change. For this study, we have mined an operation-based model repository and interviewed users to assess their understanding of randomly selected changes. Our results indicate that users better understand complex changes in the operation-based representation.

Planning iterations in software projects requires considering artifacts from different aspects such as requirements, specifications, tasks or even bug reports. UNICASE is a unified CASE tool integrating these relevant artifacts into one model. We demonstrate how the tool supports planning and executing iterations.

Models need to be put under version control to facilitate collaboration and to control change. EMFStore is a Software Configuration Management system tailored to the specific requirements for versioning models. It employs operation-based change tracking, conflict detection and merging.