Peter Bosch

@[email protected]
821 Followers
168 Following
79 Posts
Hacker. Physicist. IC Die shots. PDP-11/VAX enthousiast. šŸ”¬SEM hobbyist (he/him)
Twitterhttps://twitter.com/peterbjornx
GitHubhttps://github.com/peterbjornx
LocationLeiden, Netherlands
Radio callsignPB1X
Peter BoschFeb 24

The past month Iā€™ve been part of an assembly on use of ā€œAIā€ in assessment and education at my university.

What surprised me was how many of of the students and younger faculty there (PhDs etc) were skeptical as hell about AI, while many lecturers and professors saw all kinds of opportunities to outsource their work.

Why is the frame always ā€œstudents all use AI and academia needs to changeā€ when the actual students beg for ways to preserve the status quo? (e.g. make AI use harder by doing more stuff offline, but keeping writing assignments etc)

Peter BoschFeb 12
Show threadAnders EknertFeb 12

I... don't... even, anymore.

https://crabby-rathbun.github.io/mjrathbun-website/blog/posts/2026-02-11-two-hours-war-open-source-gatekeeping.html

Peter BoschFeb 12
Anders EknertFeb 12

AI agent "contributes" PR to matplotlib.
PR gets rejected.
AI agent *writes and publishes blog to shame the maintainer*.

What a time to be alive.

https://github.com/matplotlib/matplotlib/pull/31132

Peter BoschFeb 8
Datagraver.comFeb 8
De petitie tegen Amerikaanse overname DigiD is, qua absolute aantallen, de top 50 grootste petities in NL binnengekomen.
https://actie.degoedezaak.org/petitions/stop-de-amerikaanse-overname-digid
Peter BoschFeb 8

Due to uni workload iā€™d completely missed it but seems like GPIB support made it into linux last december! I think I might have got the ball rolling on this back in ā€˜24 when I ran into gregkh at @revspace and asked him if it would be possible to upstream linux gpib. A few months later he CCed me on the first commits to staging to let me things were happening :)

I do feel somewhat bad about being too busy to help out with the actual work, so lots of thanks to the people who made it happen!

Peter BoschFeb 3
Eloy.Feb 2

universities in the 1980s: writing the majority of internet standard RFCs and their implementations

universities now: moving away from Microsoft cloud is really hard okay? šŸ„ŗ

Peter BoschAug 7, 2025
If the transport works out I'll be bringing a PDP-11/84 to #WHY2025 . Plan is to have it be accessible from the net via telnet, ftp, etc... but there will also be ~16 direct TTY connections available, as well as dial up access via @eloy 's infra if everything works out.
#why2025camp #retrocomputing
Peter BoschAug 6, 2025
Show threadKate TemkinAug 6, 2025

@why2025camp please, please consider not handing out unprotected cells at all

handing these out creates a risk even for the people who aren't opting in to using the batteries;

and as someone who's put time and energy into badges across a lot of events, I worry handing those out goes against our onus to care for the safety of everyone at the event

Peter BoschAug 6, 2025
Jonty WareingAug 6, 2025

For what it's worth, we made the same error on the @emf 2014 badge using unprotected lipo pouches. Never again.

That event was only ~1000 people and we had at least three instances of batteries emitting smoke #why2025

Peter BoschMay 24, 2025
Graham Sutherland / PolynomialMay 24, 2025

opened a kicad feature request to make PCB/ASIC reverse engineering workflows a fair bit faster - an option to auto-assign an unassigned pad's net when routing onto it.

https://gitlab.com/kicad/code/kicad/-/issues/20992

Add an option to automatically assign an unassigned pad's net when routing onto the pad. (#20992) Ā· Issues Ā· KiCad / KiCad Source Code / kicad Ā· GitLab

Description I would like an optional feature where routing a trace with an assigned net onto a pad with no assigned...

GitLab