A summery cartoon for @newscientist.com

Meet the AITO Dahl-Nygaard Prize Winners at ECOOP'25:

Mira Mezini, TU Darmstadt (Senior Prize)
Amir Shaikhha, University of Edinburgh (Junior Prize)

Check out their awards-citation:
https://2025.ecoop.org/track/ecoop-2025-awards

It's also the first time that the award named after the two eminent Norwegian pioneers will be presented in 🇳🇴!

#ecoop2025 #aito #ProgrammingLanguages

Frage an die #diy Bubble hier - bin über jede Idee dankbar - Boost welcome! 😀👍

Habt ihr einen Trick, wie man ein gußeisernes Fallrohrstück aus dem Standrohr herausbekommt? Das steckt dort, mit einer Wulstdichtung versehen, schon seit vermutlich 50 Jahren.

Habe schon einen Schlupf versucht, um es herauszuziehen. Wackeln und ziehen. Kreisförmig bewegen und ziehen. Dichtung versucht herauszuziehen. WD40 an die Dichtung.

Bisher vllt. einen Zentimeter herausgekommen, aber nun geht nichts mehr.

With both #Eurovision and the Papal conclave happening in May, it’s gonna be a big month for Europeans who love overly complex, confusing and dramatic ways to select a winner that are run by international bureaucracies, include smoke for dramatic effect and feature excessively extra outfits

@jzohren Excellent to see you back in academia! On the badging system: Maybe student representation can come up with a system of criteria and hand out these badges according to this system.

Habe am Wochenende geträumt, dass ein neues EU-Gesetz vorschreibt, Warnhinweise auf wiederverwendete PowerPoint-Folien anbringen zu müssen 😱

If an engineering team is a delivery machine, culture is the lubricant. The smoother it runs, the more you ship.

Just returning home from meeting with the OPAL folks! OPAL is a highly-configurable static program analysis framework for JVM languages and I work on this for more than a decade now. This spring Dominik was so kind to host us for two days at UDE in Essen.

https://sse.cs.tu-dortmund.de/news-list/details/opal-spring-research-meeting-in-essen-49642/

Ich erwarte von jedem CDU- oder CSU-Mitglied, das Anstand besitzt, entweder auszutreten oder in massiven innerparteilichen Protest zu treten. https://ard.social/@tagesschau/113912721184068594

3....2.....1....Go! 🚀

Starting now, our application platform is open and we're excited to get to know brilliant minds within the field of #ComputerScience - and relevant fields - like you. ✨

In the upcoming days more information will follow, and if you join our online event on Wednesday, you'll have the chance to meet us, and your potential collaboration partner. 😉

Stay tuned and check out our website!

#ApplyNow
SoftwareCampus2025

https://softwarecampus.de/en/aktuelles/january-22nd-2025-information-event-on-the-application-phase/

We are seeking nominations for the 2025 Dahl-Nygaard junior and senior prizes!

** Deadline: 31 January 2025 **

Established by AITO in 2004, these annual prizes are named after Ole-Johan Dahl and Kristen Nygaard, whose pioneering conceptual and technical work in the sixties shaped that view of programming and modeling which is now known as object-orientation.

The winners will be given the opportunity of giving an invited talk at #ECOOP2025 in Bergen, Norway.

https://2025.ecoop.org/track/ecoop-2025-technical-papers#Nominations-for-the-2025-Dahl-Nygaard-junior-and-senior-prizes

#ecoop

I don't understand the point of publishing a paper about a great innovative tool that you developed if you're not going to make that tool available to anyone. And I don't understand the point of citing this paper as a example of a great innovative tool, either. If it's not available, it might as well be fiction. #AcademicRant #AcademicPublishing #OpenScience

If you think your codebase is old, I was talking to a team yesterday that just refactored out references to Hollerith cards (aka the reason why 80 characters is convention for line lengths) #researchsoftware

@jonasfranz Ah verstehe.

@jonasfranz Gegen was rechnest du es denn? Egal wie hoch oder niedrig es ist es bleibt ja eine Ausgabe.

Since a foundation model tries to predict the most likely label/completion, it can't do things like find the best thing. You get the average code from the training set, not the best code. For some tasks that is fine - everyone probably does address forms the same - but for other tasks - the cognitively challenging ones - that is potentially fatal.

And if you don't write code yourself, (or essays/cover letters/short stories/…), how will you know when you are seeing the average, vs the best?