So much for the regularity of updates. The second is a week too late already. After finding a brief moment to install and test-run PalaestrAI (as well as browsing through the code to get a feeling for the terminology of agent, brain, muscle and such) we started with the scenario for the next semester. Alas, there will be no time to migrate to a newer cluster computer. But, I have finished grading of all written exams last week. Now there are only 14 papers from the masters course to be graded and three days of oral exams to then mop up and finalize last semester. But, aside of reading up on Ologs, and building a first model for the running project, I could invest a few too many hours in polishing the LabSceneTool, which is our setup and control-tool for running scenarios in the laboratory. (There will be more soon, after I finished this step modularization and cleaning up, but think of it as a simpler (at least for us) way to utilise KVM.) And I got to do a first step towards setting up a local LLM-based assistant (or "agent" if you please). I might get rid of all the annoyed faces around me if I stop talking to myself and talk to some unruly AI. On the other hand, the few times I actually ask some AI for some assistance I normally end up being disappointed. Last week I asked for having the WAC ACL turned from Turtle into a graph in dot notation. After a few tries and requests for corrections I ended up loading the RDF into vim and regexping the thing into an acceptable graph within 20 Minutes. I also got some more understanding of the defined classes and mappings, that might have escaped me if I didn't have to review my own work. (Plus, I did not wrongly reverse edges or hallucinate additional nodes.) The UN finally dropped the Mercator-projection and chose the Equal Earth projection. Seemingly only one country had problems with an equiareal depiction. Congratulations Africa, hopefully you now have a little less confused tourists when they realize how big that continent really is. ;-) And unless you can actually calculate the great arc from Paris to New York on a Mercator map I assume you could not actually make use of the properties of the old map.