EVENTS

Wilhelm Hasselbring from Christian-Albrechts-Universität zu Kiel 

For large, complex software systems, the design of the overall system structure, known as software architecture, presents a critical challenge. Software architecture outlines the system in terms of components and their interconnections, distinguishing itself from the details of a fine-grained system description. It establishes a link between requirements and the final system, providing a rationale for design decisions and encompassing decisions about quality attributes, which can then be evaluated early in the development process. For instance, system performance is heavily impacted by the complexity of coordination and communication, especially when components are distributed across a network.

The reusability of components and services is significantly influenced by their coupling within the software architecture. When a new research software engineer begins work on an existing software system, the architecture is typically the first element they review. This can be challenging if the architecture is undocumented. In such cases, reverse engineering the architecture from the code can aid in understanding the program.

As the embodiment of initial design decisions, software architecture includes choices that are difficult to alter later, making them crucial for careful consideration. Research software is often developed with limited resources and without a long-term maintenance vision, which can lead to architectural decay.

The research software engineering community holds the belief that better software fosters better research. Similarly, the software engineering research community asserts that better architecture results in better software. Therefore, I advocate for enhanced consideration of architectures in research software to enhance the quality of research.

Institutions

• Helmholtz Platform for Research Software Engineering - Preparatory Study