I wonder how this relate to Arch Linux.

The IG Nobel Prize is satire. Many aspects of the ceremony are excentric and/or humorous.

All of the 36 ceremonies were the first annual ceremony. That’s just the way it is. Enjoy the show.

Le comité éthique a approuvé une procédure très proche qui l’autorisait à menacer des serpents d’une façon un peu différente, avec une barre de fer.

Il a modifié la méthode pour utiliser son pieds au lieu d’une barre de fer sans re-soumettre cela au comité éthique. Il s’est défendu en disant qu’il craignait de blesser le serpent avec la barre.

Il a effectivement utilisé une méthode qui n’est ethniquement approuvé. Mais pas certain que la méthode approuvé était beaucoup mieux.

Oui. La publication a été retiré pour raison éthique. https://retractionwatch.com/2025/03/07/snakes-soft-stepping-retraction-butantan-institute/

2023 IPCC report: https://www.ipcc.ch/report/ar6/syr/ For a TL;DR see page 27 of Summary for Policymakers

The Shift Projects publications on decarbonisation: https://theshiftproject.org/en/publications/ This NGO publishes sector-specific analysis often focusing on France

Plans for energy and transport sectors typically include rapid electrification and move to renwables, varying levels of nuclear, and a residual amount of fossiel fuel or biofuel. Combined with efficiency, sobriety in energy and resources uses.

There was so much delaying and lobbying against change, that avoding the worse now require accelerating change in basically every sector, because every sector is dependent in some way on fossiel fuels. No single person can understand and know all sectors at a deep level. Analysing this require large group of people with expert in different fields. For instance the IPCC and The Shift Project.

Let’s not rush the transition to clean energy, it might disrupt part of the economy that need fossil fuel. /s

The world isn’t just paying for the US’ inability to honor the 2015 Iran nuclear deal. It’s paying for shortsightedness w-r-t energy and fossiel fuel.

Je découvre Gentside et remarque une pointe d’ironie

Gentside fait partie du groupe Prisma Media, acteur majeur des médias en France, et est également disponible en plusieurs langues, dont l’anglais et l’allemand. Prisma Media est une société du groupe Vivendi, société par actions simplifiée au capital de 3 000 000 euros, présidée par Arnaud Lagardère.

Bluetooth range is too short. AM/FM frequencies are regulated so that individual cannot legally broadcast. So these probably aren’t used to send data to manufacturers.

They’re probably relying on existing channels that are IP capable, have good coverage, have relatively cheap antenna/MoDem available on the market. Such as 3/4/5G, Wi-Fi, LoRa.

Disabling those likely mean loosing features, ie emergency calls, anti-theft localization… But it may be the easiest way and can be an acceptable compromise for some. Flashing devices to keep using the hardware and features with more trustworthy software is potentially better. I fear regulators/vendors would put roadbloacks. All onboard systems may require some kind of certification to be street legal. If software certification is necessary, then hobbyist couldn’t possible pay for that, which means in practice hacking you own device may not be legal. If there’s an accident, except insurances to use any unauthorized car modification to put the blame on you and deny coverage.

Chinese manufacturers, amongst others, won’t reduce such risks

Let’s regulate them. Car have regularions for safety equipments including seatbelt, lights, brakes, … There could a regulation against onboard systems that are always connected, or connected by default.

The article rightly point out politics are lagging behind. Cathing up should include regulation.

The Dutch warning does not identify China or Chinese manufacturers as the source of the threat. It addresses the vulnerabilities of connected vehicles in general.

That makes sense. Adversairies, whoever they are, can exploit vulnerabilities in car regarless of the factory’s location.

Fixing individuals vulnerabilities is not enough. Manufacturers need to rethink and reduce the amount of connectivity to reduce risks. An always-connected onboard system with mic, camera, and complex software almost certainly has flaws that can be abused.

High power chargers have specific uses cases, such as public charging stations in high traffic areas. They are useful for quick charge when it’s the only option. They cost more to purchase, install, and it’s going to cost more for people charging at those stations.

Drivers who can wait a few more minutes would be better off using cheaper lower-power charging stations, or ideally charging overnight at home.

Any insight into the most severe ones, eg is remote exploitation possible and likely?

The Linux project request CVEs for any memory safety bug just in case some turn out to be vulnerabilities, in which case they may be exploited given enough effort. Many are likely not exploitable or unlikely to be, and/or low severity.

There are some higher severity ones given the following message. But this doesn’t mean all CVEs listed are severe issues.

Several vulnerabilities have been discovered in the Linux kernel that may lead to a privilege escalation, denial of service or information leaks.

Large AI companies know their tools are dangerous, and are shouting it from rooftop.

Knowing that, they keep developing them, and ran agents with little supervision, insufficient sandboxing, and ended up attacking third parties.

I hope there’s liability.

The Academie is out of touch with contemporary french language. They got laughted at when they released their latest dictionary which stated FRANC was the official currently, despite the EURO replacing it 20+ years ago.

Commercial dictionaries like Robert or Larousse are more useful or up-to-date. The Academie dictionnary seems mostly useless sadly.

violations of state consumer protection law

Then presidential pardon is not an option.

This will keep happening until there’s an appropriate price/tax on carbon, to make polluters pay.

VC mostly care about gambles and profits. The only insentives they understand are money related.

Love that headline.

In such hypothetical situation, it may be a bit more ethical to do like in the Alias series finale. Don’t kill the vilain, let him to starve in a remote place. If he happens to be immortal for some reason, ensure he’s burried deep.

What does “compute” represent in this graph?

Computer parts have gotten much more expansive in the last few year, in large part due to disproportionate investments in LLMs, and LLM companies hoarding hardware.

People are causing confusion when using account to both repost messages, which are sometimes written in the first person, and writing their own posts in the first person.

It seems fair to label them as bot or similar if the amount of (presumably automatic) reposts cross a threshold. It would be nice to make it possible to remove the label after they stop mixing both activities within the same account.

If the platform is too accomodating with accounts that mix artificial and human content, expect bots and LLMs account to abuse the rules and claim they are partially human curated. It’s going to be impossible to verify and a nightmare to moderate.