On May 20 and 21st the Nordic-RSE (Research Software Engineer) conference took place in Gothenburg, co-organized by GRIDH. After learning about the conference at a research engineer meetup at Humanisten in March, four research engineers from Språkbanken Text were among the participants from all over the Nordic countries and even the UK.
Besides a strong focus on high-performance computing (HPC) and consulting researchers on their computing needs, other topics included, as a selection:
- open data (Rebecca Willén (IGDORE) "Clarity, Not Consensus: Rethinking Unity in Open Science") and software (Richard Darst (Aalto University, Finland): "In the modern era, what can we learn from the history of free and open software?")
- reproducible, i.e. actually runnable, research software (Ina Pöhner (School of Pharmacy, Faculty of Health Sciences, University of Eastern Finland, Kuopio, Finland): "Unreviewed and Unrunnable? On the role of peer review in publishing reproducible research software")
- as well as best practices (Marine Guyot (CodingResearcher/freelance RSE): "Design Patterns: The secret to smarter, reproducible research")
Here are some slides from one of the best (and most frustrating) talks about reproducible software.
Slide from "Unreviewed and Unrunnable?" presenting the research question about the quality of code linked in scientific papers:
Slide from "Unreviewed and Unrunnable?" with the results showing that the majority of code cannot be run or is missing:
A final slide with steps towards more runnable code:
The conference dinner and coffee breaks gave space for networking and informal conversations, including the inevitable discussion about the usefulness and risks of large language models.
A career panel, tutorial sessions and posters rounded up the conference program. The [Nordic-RSE community](https://nordic-rse.org/) was very inviting and inclusive and we are happy to have been part of this conference and looking forward to staying in touch.
