This isn’t really the same kind of bug. Those bugs made instructions emit the wrong answer, which is obviously really bad, and they’re really rare. The bugs in the article make instructions take different amounts of time depending on what else the CPU has done recently, which isn’t something anyone would notice except that by asking the kernel to do something and measuring the time to execute affected instructions, an attacker that only had usermode access could learn secrets that should only be available to the kernel.

The $5 Windows keys have never been legitimate - either they’re just people selling keys they’ve generated with a keygen or bought with a stolen credit card, or it’s students reselling free keys they’ve got from Dreamspark or a sysadmin selling keys from their employer’s enterprise licence, which, in Microsoft’s eyes, are all piracy. An OEM copy of Windows 11 Pro is about €150 and can’t be transferred to a different motherboard, and a retail copy which can be transferred is about €300. A one-time purchase copy of Office is about €120 (it’s also available as a subscription). These machines either have at least €270 of software on them, or €0 worth of pirated software on them.

It can’t be legitimate because licences for the bundled software cost more than the machines are being sold for. Also, the hardware included isn’t officially Windows 11 compatible, so selling it with Windows 11 installed is misleading the customer into thinking they’re buying something much more recent than they really are. For a decent number of people buying these, they’re likely to own something just as new already, and could get a free upgrade to Windows 11 by doing the same configuration tweaks as the sellers did.

AGPL is a full-on FOSS licence with strong copyleft requirements, not a measly open-source licence like Apache, which could be pivoted to proprietary at a moment’s notice. We’re communicating through an AGPL-licensed system right now as it’s what Lemmy’s licensed as. If they were going for a corporate-friendly licence, AGPL is the last thing they’d choose as it forces you to share source code with even more people than the regular GPL does.

Wikipedia management shouldn’t be under that pressure. There’s no profit motive to enshittify or replace human contributions. They’re funded by donations from users, so their top priority should be giving users what they want, not attracting bubble-chasing venture capital.

Lots of the criticism of the .ml admins isn’t for being communists, it’s for promoting fascism as long as it wears a red hat. Some of the strongest critics are out and proud communists, but who have the sensible opinion that doing what Stalin did and what Putin is doing are not ethical or effective ways to build communism.

Point defence systems for ships and anti aircraft missiles like Patriot have all killed people in autonomous modes in friendly-fire incidents, so LAWS aren’t new and don’t require AI. Barriers to building them are reducing, but again, that’s more due to cheap drones rather than AI.

That kind of bioplastic tends not to biodegrade naturally, instead requiring a heated industrial composter with specially engineered enzymes added. If it’s disposed of properly, it’s great, otherwise it’s no better than traditional plastic but costs more. Also, not all bioplastics biodegrade at all as all the word means is that the source material is biomass rather than oil.

It’s not guaranteed that it’s interpreted as a platitude by the person it’s directed at, and when the mismatch between the task and the work done is big enough to make it obviously a platitude, it’s just patronising, and risks being more insulting than not saying it at all.

The feedback in the article was obviously far from perfect, but from the sound of it, “good attempt” could be an actively harmful thing to say. Lots of effort had gone into making the wrong thing and making it fragile, which isn’t good at all, it’s bad. If you’d asked an employee to make a waterproof diving watch, and they came back with a mechanical clock made from sugar, even though it’s impressive that they managed to make a clock from sugar, it’s completely inappropriate as it’d stop working the instant it got wet. You wouldn’t want to encourage that kind of thing happening again by calling it good, and it’s incompatible enough with the brief that acknowledging it as an attempt to fit the brief is giving too much credit - someone who can do that kind of sugar work must know it’s sensitive to moisture.

The manager can apologise for not checking in sooner before so much time had been spent on something unsuitable and for failing to communicate the priorities properly, and acknowledge the effort and potential merit in another situation without implying it was good to sink time into something unfit for purpose without double checking something complicated was genuinely necessary.

I was meaning that the blade count and detachability was the difference in definition between turboprop and propfan/open turbofan, not that it was necessarily the thing making the engine more efficient.

I’ve seen turboprops in museums and on the internet with around six or eight blades. When I looked on the Wikipedia page for propfan engines, which seems to be another name for an open turbofan, the distinction seemed to be mainly how the blades were shaped (like propellor blades or turbine blades) and how tightly-integrated everything is (you can swap the propeller out on a turboprop).

How many blades do you have to add to a turboprop before it’s promoted to an open turbofan and touted as a major new innovation?

Arch is at least more likely to update to a fixed version sooner, and someone getting something with pacman is going to be used to the idea of it breaking because of using bleeding edge dependencies. The difference with the Flatpak is that most users believe that they’re getting something straight from the developers, so they’re not going to report problems to the right people if Fedora puts a different source of Flatpaks in the lists and overrides working packages with ones so broken as to be useless.

People fall off rooftops fitting solar panels, burn to death repairing wind turbines that they can’t climb down fast enough to escape, and dams burst and wash away towns. Renewable energy is much less killy than fossil fuels, but per megawatt hour, it’s comparable to nuclear, despite a few large incidents killing quite a lot of people each. At the moment, over their history, hydro is four times deadlier than nuclear, wind’s a little worse than nuclear, and solar’s a little better. Fission power is actually really safe.

The article’s talking about fusion power, though. Fission reactions are dangerous because if you’ve got enough fuel to get a reaction at all, you’ve got enough fuel to get a bigger reaction than you want, so you have to control it carefully to avoid making it too hot, which would cause the steam in the reactor to burst out and carry chunks of partially-used fuel with it, which are very deadly. That problem doesn’t exist with fusion. It’s so hard to make the reaction happen in the first place that any problem just makes the reaction stop immediately. If you somehow blew a hole in the side of the reactor, you’d just get some very hot hydrogen and very hot helium, which would be harmless in a few minutes once they’d cooled down. It’s impossible for fusion power, once it’s working, not to be the safest way to generate energy in history because it inherently avoids the big problems with what is already one of the safest ways.

PowerShell has a system to sign scripts, and with its default configuration, will refuse to execute scripts, and with the more sensible configuration you should switch to if you actually use PowerShell, refuses to execute unsigned scripts from the Internet.

I suspect that most of the scripts you’re referring to just set -ExecutionPolicy Bypass to disable signature checking and run any script, though.

That’s misleading in the other direction, though, as PhysX is really two things, a regular boring CPU-side physics library (just like Havok, Jolt and Bullet), and the GPU-accelerated physics library which only does a few things, but does them faster. Most things that use PhysX just use the CPU-side part and won’t notice or care if the GPU changes. A few things use the GPU-accelerated part, but the overwhelming majority of those use it for optional extra features that only work on Nvidia cards, and instead of running the same effects on the CPU if there’s no Nvidia card available, they just skip them, so it’s not the end of the world to leave them disabled on the 5000-series.

A billion dollars worth of bitcoin isn’t that much of the total amount of bitcoin any more, so any impact on the price would be because of people reacting emotionally to the news, not because they’re was suddenly a flood of extra bitcoin in circulation.

If you write cross-platform software, the easiest solution is usually to pretend everything’s Unix. You’ll hit some problems (e.g. assuming all filesystem APIs always use UTF-8 will bite you on Windows, which switched to UCS2 before UTF-8 or UTF-16 were invented, so now uses UTF-16 for Unicode-aware functions as that’s the one that’s ABI compatible with UCS2, and passing UTF-8 to the eight-bit-char functions requires you to opt into that mode explicitly), but mostly everything will just work. There’s no XDG_CONFIG telling you to put these files anywhere in particular, as Windows is Windows, so most things use ~ as a fallback, which Windows knows to treat as %USERPROFILE%.

You can jam the Windows UI by spawning loads of processes with equivalent or higher priority to explorer.exe, which runs the desktop as they’ll compete for CPU time. The same will happen if you do the equivalent under Linux. However if you have one process that does lots of small allocations, under Windows, once the memory and page file are exhausted, eventually an allocation will fail, and if the application’s not set up to handle that, it’ll die and you’ll have free memory again. Doing the same under every desktop Linux distro I’ve tried (which have mostly been Ubuntu-based, so others may handle it better) will just freeze the whole machine. I don’t know the details, but I’d guess it’s that the process gets suspended until its request can be fulfilled, so as long as there’s memory, it gets it eventually, but it never gets told to stop or murdered, so there’s no memory for things like the desktop environment to use.