Murilo dos Santos Vianna
From a new RSE in IBG3...
The deRSE25 was my first time attending an RSE conference and I really enjoyed it. My background is on environmental modelling focusing on climate change impacts in agriculture, food production and land use. The conference was held in the KIT (Germany), in parallel with the software engineering conference SE25 from 25th-27th of February 2025. Several presentations covering the main topics of RSE were given, followed by parallel workshops, discussion sessions (BoF), and poster presentations.
During the conference I could meet former colleagues and get to know new people from completely different disciplines who share various common questions on how to better develop, document, ship and maintain software for research applications. At the same time, I had the opportunity to meet software engineers and computer scientists who are also keen to understand the demands and find the tools that could be suited in a research setup. Although RSE has gained traction in the last years and the community is apparently growing, a clear path and instruments to support career development and recognition of professionals in this area are still a work in progress. Particularly within the current academic scheme that is mainly oriented towards fundings acquisition, publication and tenure-track systems. On the other hand, it seems clear that an RS engineer would benefit from a deeper understanding of its domain area, which could be obtained through a PhD. Having experience in academia can also facilitate communication and expectation management between scientists and software engineers in research projects. It was also interesting to see how the Helmholtz Association is developing some metrics to classify RS based on FAIR principles criteria. During the deRSE25 conference I had the opportunity to attend and participate in many BoF sessions and presentations addressing these topics in more detail.
The demand for FAIR principles appeared as a common ground for all disciplines. During the conference I could understand not only the demands for different domain areas but also learnt some tools to support the FAIR principles in scientific workflows. Some examples were the HERMES and FACILE-RS that could be integrated with CI-pipelines of GitLab/GitHub for automated meta-data creation and software archiving in repository with persistent identifiers (e.g., Zenodo). Software Management Plans also appeared as a tool to improve the documentation and management of research software, while some high-level interfaces for interoperability of scientific models and data were also presented (e.g., ComIn, RUBIX). Review studies, tools and workflows to analyze static code and git repositories’ meta-information were also presented aiming to find patterns in scientific software development, their developers (“personas”) and demands. There were also dedicated sessions and workshops on the usage of LLMs for RSE, and ML-assisted data workflows.
In summary, I really learned a lot and connected to many new colleagues at the deRSE conference. It was a great opportunity to update my skillset that could directly be used on my daily work in the IBG3. I’d definitely recommend the conference and look forward to future ones.