Discover how Research Software Engineering is becoming a research field of its own, focused on improving the unique processes, tools, and challenges of building scientific software.
| Resource information | Details |
|---|---|
| Article title | Investigating Research Software Engineering: Toward RSE Research |
| Authors | Michael Felderer, Michael Goedicke, Lars Grunske, Wilhelm Hasselbring, Anna-Lena Lamprecht, and Bernhard Rumpe |
| Focus | Research Software Engineering |
Research software is usually created to address particular research questions, which means it often has special requirements that don’t align with typical commercial software. This article, Investigating Research Software Engineering: Toward RSE Research, brings into focus Research Software Engineering (RSE) as an emerging field that’s both a career path and an area of study focused on bridging the gap between software development and scientific research.
The article walks through several important themes: first, it looks back at decades of traditional software engineering research, highlighting how its goals and methods don’t always meet the needs of scientific software. Then, it introduces three major categories of research software, along with explaining the growing focus on sustainability. The article also defines Research Software Engineering (RSE) as a profession and outlines why scientific software projects have unique challenges, such as unclear initial requirements, difficulty in validation, and limited use of standard engineering practices.
What makes this article valuable is its introduction of the idea of RSE Research, as a complementary approach to RSE, which aims to improve the scholarly understanding of research software engineering and to develop and improve the tools, techniques, and skills needed to create research software that’s reliable, reusable, and easy to adapt. The article also outlines example questions that could potentially drive RSE research, such as what distinguishes it from other software disciplines, how to better support scientific workflows using traditional software engineering practices, which tools and skills RSEs need, and where technologies like generative AI might fit into this landscape.
In the end, the takeaway is simple but important: writing code isn’t enough. If we want research software to be robust and reusable, we need to rethink how it’s designed, developed, and maintained. The article also makes a strong case for international collaboration and improved funding to help this emerging field of RSE Research grow.
