@exchgr The old-fashioned way!
You reminded me of something else I wrote last year. https://www.overmorrow.tech/sora-and-the-cost-of-real-things/
@exchgr The old-fashioned way!
You reminded me of something else I wrote last year. https://www.overmorrow.tech/sora-and-the-cost-of-real-things/
Ongeveer honderd actievoerders kwamen vandaag voor het centraal station bijeen om te protesteren tegen de dienstplicht en tegen overheidsinvesteringen in Defensie. Scholieren en studenten waren opgeroepen om hun les te verlaten voor de demonstratie. Onder hen ook studenten van de Radboud Universiteit.
"Banken dringen steeds meer hun bankapps op aan klanten"
"De Consumentenbond wil dat banken alternatieven blijven aanbieden."
https://www.consumentenbond.nl/nieuws/2026/banken-dringen-steeds-meer-hun-bankapps-op-aan-klanten
yeah of course it is garbage, why else would there be a need to improve it?
In early October I wrote a note on how mathematicians should stop selling their credibility to AI companies, and kept it in the draft folder. Seems still very relevant now, so here it is:
giannipetrella.eu/blog/2026/11...
"We are disappointed that NameSilo has held our domain without informing our team or proactively forwarding the abuse report, which would have allowed our moderation team to act"
https://codeberg.org/Codeberg/changelog/issues/7#issuecomment-28100157
I am good at a few things but very bad at many others. For example I know basically nothing about plumbing or sewing.
The idea that I would be able to give "the correct specification" for those problems to have someone unskilled do it really does not make any sense.
This post is about vibecoding.
44 countries on Earth have no coastline at all.
Losing direct, reliable sea access costs a country years of economic growth. Losing that access, or never having had it, shapes everything from national identity to actual wars.
This week in #geoweirdness we take a tour through some of the oddest consequences of being landlocked.
https://en.wikipedia.org/wiki/Landlocked_country
1/
9/ Nearby Liechtenstein 🇱🇮 is one of only two countries on Earth are doubly landlocked, meaning every single one of their neighbours is also landlocked. The other is Uzbekistan 🇺🇿 in central Asia.
To reach the ocean from either, you have to cross at least two international borders no matter which direction you travel.
What I want are tools that explicitly work against the slippery slope.
That break the chat-like interface that drives so much of the pain I see around me.
Tools that help me to work against the (often false) urgency I feel. That breaks.
Things breaking is good, actually.
Breakage leads to 'stopping and thinking' - and I think we all should be doing more of that, for our own sake
[4/4]
RE: https://floss.social/@gnome/117638482708824656
GNOME is kicking off their yearly #fundraiser today! Help keep all free #desktops running with your #donation!
https://violetpixel.com/blog/2026-11-17-we-need-smart-people-who-give-a-shit/
I haven't been so inspired by a rant in a long time
I want to *perform* this rant at an open mic in the belly of the beast — at a pop-up performance on Sand Hill road, with a backing band of villagers with pitchforks and torches
nooooo, compliance.conversations.im/ has shut down
does anybody know whether there is something similar that people who don't have an android¹ phone can use to check the configuration of their xmpp server(s)?
¹ my phone runs debian, so basically it will have to be some tool usable from the desktop?
If the Sun was the size of a lentil, the nearest star would be 140 km away. Voyager 1 is now 90 m from the Sun 😬
Douglas Adams said it best: "Space is big. You just won't believe how vastly, hugely, mind-bogglingly big it is. I mean, you may think it's a long way down the road to the chemist's, but that's just peanuts to space."
Dear friends,
We're reaching out to the people who have been part of Varia's life over the years, because now we could really use your support.
After nine years in Rotterdam, we have lost structural funding for 2027 and our members' contributions are no longer enough to cover the rising costs of keeping our space and shared digital infrastructure running.
We're inviting people who value what we've built together, to help us keep Varia going. Even a small donation can make a difference!
We've written a bit more about the situation and what your support would make possible here: https://varia.zone/en/support/
Support us: https://opencollective.com/variazone
Please boost, share, let this message drift into distant networks...
Thanks a lot!
Varia
There's a lot said about LLM-generated vulnerability reports, but LLM-generated bug reports in general are a huge problem because of the underlying economics. Every bug fix has an obvious benefit (the bug is gone now) but has some additional costs:
First, there is the opportunity cost. You have to spend time reviewing the bug, creating the fix, and reviewing the fix. That is time that is spent on that is not spent on other things. If a bug has no security implications and is almost impossible to trigger, then it should be low priority. Writing bug-free code is impossible. Even the best formal verification techniques don't save you when your specification is buggy (to give a concrete example: EverCrypt was formally verified with the property for temporal memory safety that no object was accessed after being deallocated. The spec forgot to mention that objects should be deallocated at some point). Fixing bugs is always a tradeoff and there are almost certainly bugs that will impact more users that you could have been fixing while you were instead triaging and fixing something an LLM produced.
Equally importantly, the bug fix may introduce additional regressions. This is especially true for LLM-generated fixes. I've had several bug reports where someone has had the LLM generate a test case and then use that (plus the rest of the test case) as the fitness function for generating a fix. This has a number of common failure modes. Sometimes the test case triggers a specific failure path, so you can fix that path without fixing the underlying root cause. Sometimes the test case is entirely spurious (for example: API misuse) and the fix introduces some significant performance regression, where the original failure mode was a deliberate choice to crash in particularly unusual cases rather than taking a 10% performance hit in the common path.
This isn't new to LLMs. We saw exactly the same thing with static analysis. If you ran a new static analyser on a large existing C codebase, you'd typically see thousands of reports. Some were security issues. Some were real bugs that would likely affect users. Some were real bugs that would be unlikely to ever be triggered in the real world. Some were false positives. Going through all of them and finding the ones that actually mattered was incredibly hard.
The problem is never 'find more bugs', it's always 'prioritise the bugs that actually matter'. Some of the LLM-based approaches are helpful here for providing an exploit path, but LLMs are typically too flexible here. They'll often generate test cases that are outside of your threat model, or which misuse APIs in a way that allows callers of a library to break themselves (which they can easily do without the library). So even triaging from the test case is tricky.
Unfortunately, determining the real-world impact of a bug depends on a lot of extrinsic knowledge. LLMs will try to infer this but it varies hugely between codebases and so a statistical approach often gives a very wrong impression.
De huidige politiek.
In de afgelopen tien jaar heeft Nederland acht staatssecretarissen voor belastingen gekend. Tien ministers van Volksgezondheid. En veertien ministers van Buitenlandse Zaken.
(...)
Toch goed om even bij stil te staan, mochten mensen in hun ergernis over de strapatsen van het minderheidskabinet alweer naar een verlossende wissel verlangen. (...)
in die tien jaar zaten we een derde van de tijd met een demissionair kabinet. Over stilstand gesproken. Willen we dat weer?
codeberg.org is unfortunately down right now due to a yet undefined issue with our domain. It was not expired and not transferred, the domain is completely gone from our account and we did not receive any email notifications, and it is currently in the status clientHold. We are currently trying to contact our registrar to get more information right now & trying to resolve this issue, but this is probably going to take some time.
We are sorry for the inconveniences & are working on a solution.
UPDATE2: clientHold is gone, codeberg should be back soon! RESOLVED.
Oh. Oops. It seems that someone convinced the .org registrar that #codeberg is bad and so the domain name codeberg.org is currently on clientHold and thus removed from DNS. They are trying to get the domain name restored and are in parallel checking if they can switch to codeberg.eu.
UPDATE: "We have received the confirmation that our domain was suspended and our case got forwarded to the abuse team."
https://social.anoxinon.de/@codebergstatus/117636437840688845