Bom Dia, Rio!
Looking forward to the 1st International Workshop on Software Engineering and Research Software #SERS2026 at #ICSE2026 today, where #ResearchSoftware meets #SoftwareEngineeringResearch and vice versa!
Remote
Stephan Druskat
@sdruskat@fediscience.org
mastodon 4.6.9#SoftwareEngineering researcher with a focus on #ResearchSoftware 🚀 at @ubbraunschweig@openbiblio.social (@tubraunschweig@wisskomm.social) 📚🦁 • @SoftwareSaved@mastodon.social Fellow 🧑💻 • Co-lead #CitationFileFormat 📑 • #RSEng • opinions my own
0 Followers
0 Following
7 Posts
Joined April 26, 2025
pronouns:
he/him
homepage:
research interests:
Replying to
@SoftwareSaved@mastodon.social Just learned about @SoftwareSaved@mastodon.social #ResponsibleAI Study Group's
⚠️ Generative AI Risk Register
🔗 https://jshng-glasgow.github.io/SSI-Responsible-AI-Risk-Register/
This resource is open to, e.g., the #ResearchSoftware and #RSEng community for registering and researching risks of using #GenAI tools in research software development and engineering, and discover mitigations.
📢 The program for the 1st International Workshop on Software Engineering and Research Software #SERS2026 is now online!
🔗 https://conf.researchr.org/home/icse-2026/sers-2026#program
Our #ICSE2026 workshop aims to bring the particular needs of research software engineering to, and those who write that software into, the #SoftwareEngineering community at #ICSE.
It's jam-packed with 19 short presentations and a collaborative session around overarching topics such as understanding research software engineering, quality and maintainability of #ResearchSoftware, software in the context of research and research software ecosystems and communities.
Replying to
@float13 @theorangetheme @alice Thanks for sharing this! Had to try of course, interesting (to me) result: this works for English entries, German ones end up circling around languages and linguistics. What that has to do with anything is , however, a philosophical question?
Replying to
@andrewnez@mastodon.social Great read, thanks for writing it!
It reminded me partly of this:
https://indieweb.social/@tg/116522563497941266 without the nostalgia; the analogy being, I think, that we tend to get metrics for "platform-like" packages that are seen and used by the person in front of the screen, while the "boring" infrastructure bits are missed cos they don't (need to) follow every wave of "disruption"? ¯\_(ツ)_/¯
When prompting a locally run Llama 3.2 1B Instruct to explain the difference between #Fibonacci tree and a Fibonacci sequence for a 11-year-old in German (as requested by K2), the response started off with "Ze Fibonacci-Treier und ze Fibonacci-Sequenz ..." 😂
#ZeGerman in front of the screen coudn't help but subvocalize (in an approximation of the accent of #Snorri, a character from the great #Maniac series) and LOL a bit.
(The rest of the response was quite unintelligible, but not in a funny way. Definitely worked to discourage K2 from using LLMs much for school-related stuff.)