You need ID to drive a car, which is essential in modern America. Worse still you need ID to rent a house and that’s normally getting fed straight into a massive insecure database. The advantage of Linux is that we could theoretically choose who we give our ID to (whether that’s Red Hat, Ubuntu, OpenSUSE, Debian, Arch, etc). Handing over your ID is necessary for some essential parts of modern life, and while I wouldn’t want to hand it over to access my operating system, I would be able to accept it.

Thinking critically, let’s imagine that only government approved companies could verify your ID and those companies are Google, Apple, Microsoft, and Persona. At that point I’d … really hate it but I’d hand over my ID. Then I’d double check my operating system isn’t logging and sharing my internet traffic.

There’s no indication that our online traffic will be required by law to be linked with our proven ID. If such a thing does happen, then firstly we are totally screwed, and secondly it would likely involve all major websites participating. We fundamentally won’t be able to get around it in that case.

Your response again doesn’t really follow from what I wrote. It retains some key words but not the ideas.

Browser fingerprinting which exists because the average person can’t be bothered concealing it and the theoretical sharing of your ID with the sites you visit due to a government mandate are two entirely different things. The relevant difference is that the government doesn’t mandate browser fingerprinting, it exists because it is technologically possible and the mitigation measures are more inconvenient than the average user is willing to deal with.

As for normalizing OS-level ID checks as a slippery slope towards sharing your full ID as part of a HTTP request … firstly that is not something you can get around with an alternative distro anyway, because it would involve all major websites. Secondly, that is a hypothetical within a hypothetical. Thirdly, if that really is the path that we’re on, now is not is not the most effective time to oppose it, because the slippery slope argument is far more persuasive from the bottom of the slope.

EDIT: I think I just did the same thing I accused you of, talking past you. My response basically just rejects your core conceit, that being a distinction between the private power-user experience and the non-private normie experience, and nothing else. I’ll need to edit this.

EDIT 2: Okay, fixed.

Parts of what you just said are not really a proper response to what I said, either because of accuracy or relevance. So I’m just going to address the one important part of what you said, metadata.

I didn’t consider metadata because I treat proof of age as what it is, proof of age with proof of identity being incidental. If visiting a website requires handing over my full birthday, “hardware ID”, or real identity then I would be concerned, but we’re not there yet.

It’s a widely held view in the general public that you should be able to browse the internet privately just like you should be able to browse a library without the government seeing a log of every book you read, and I hope that would be enough to resolve this. The general public is not very concerned about browser fingerprinting, which effectively erases user privacy, but government mandated sharing of your identity online would be a red line that would get the normies involved.

I think it’s helpful to put some thought into why you use Linux and what you really need from it. I use it primarily for choice, privacy, and to just not be using anything by Microsoft/Apple/Google. Security is nice to have but it’s not the reason I’m using Linux, so handing over my photo ID to a third party I trust is an acceptable if disappointing risk.

Sure, my OS will be tied to my ID, but as long as my online traffic isn’t that should be fine. If they wanted to monitor my online traffic it would make far more sense to do it at the VPN level instead. Not by having my open source operating system redirect my traffic so that it’s associated with my ID.

The big risk is social media requiring proof of ID. Bots are becoming more and more common and proof of ID available at the OS level on Windows, Mac, and Android would be very tempting for social media. That’s a different concern though.

The stereotype of pedophiles in cop shows is that they use desktop computers anyway, not phones. Don’t know how true to reality that is though.

I think both approaches are too extreme. Supporting every device leads to poor security, poor stability, and therefore a poor user experience, but only supporting just Google devices (while there is a good reason for that) is a step too far for most people.

If I were in the position of e/os I’d just support probably three manufacturers. Going through the major ones that I know of: Motorola and Google are obvious picks. Next would need to be something cheap and popular. Samsung is way out of the question. Xiaomi and Vivo I’ve never seen their phones mentioned outside of China (which is a country that generally doesn’t have the same privacy considerations as people in the west do). That leaves Oneplus and Tecno Mobile for the third model.

They’re two sides of the same coin. Can’t have privacy without security and can’t have security without privacy.

Looking at the post though he’s specifically talking about advanced security as a means of preserving privacy, security you’d need if (based on his model) targeted by a government (whether foreign or your local police forensics team). I don’t think his model is correct though because while extra hardened security is useful to protect privacy in such an instance, it’s also just best practice because it’s better to have too much security than not enough, just to keep your bank account secure at least.

The fact is we need cameras. I know being in favour of any form of law enforcement is heresy here but there are murderers who were arrested only because of CCTV. The question is not “cameras or no cameras” because cameras will win for the average person 9 times out of 10 (to the extent that Ring doorbell cameras sell like hotcakes), the question is whether the cameras are internet connected.

I find iPhones themselves dystopian.

I tried to think of an exception, but no dice.

Hypothetically though, if Canada invaded and annexed the United States then gave them free universal healthcare, that would probably be completely justifie. Basically any action against the US government is acceptable if I think about it.

The Democrats would not do something as self-destructive as invading Iran.

Even if everyone did, not every business would. Shell would buy American oil in a heartbeat.

CCTV isn’t a problem. It’s mass surveillance that’s the problem. As the saying goes, “show me the man and I’ll show you the crime”. With enough information the state can always find something to turn their political enemy into a criminal. The government should not be able to do that with a mere flick of the wrist.

If anything a higher number of CCTV’s is a good thing, assuming the number of all types of camera combined remains constant between cities.

• Searching a large dataset with a vague search criteria.
• Real-time feedback when studying a foreign language (since accuracy is less important than quantity).
• Apparently in medicine they’re using generative AI for something meaningful, but I’m not entirely convinced it is actually generative AI and I’d need to do more research.
• Sometimes it can help in learning to program and in sanity-checking code security.

I think part of the problem is that Linux basically supports every use case. Sometimes the scope needs to reduce, older hardware needs to be abandoned, etc.

The article cites some hardware being incompatible with Rust as part of the issue, but perhaps what needs to change is that a line in the sand needs to be drawn over which hardware is accepted. The legacy hardware can be left to the forks developed by paid employees for their businesses own personal use.

Linux doesn’t need to become as hardline on hardware as Windows 11, but the distros where the maintainers are at risk of burnout can certainly afford to abandon at least some hardware. Apparently Rust doesn’t support alpha (1992-2007), hppa (1986-2008), ia64 (2001-2019), m68k (1979-1994), or s390 (1990-2004). All of these are at least 18 years old, with the exception of ia64, but apparently the Linux kernel already dropped support in 2024.

To combat global AI memory shortage

More likely, to combat Cxmt.

If the system stores the actual age, that means they could automatically change said boolean.

Also the California law requires users be divided into one of four age brackets. Not just adult/not-adult.

And besides it’s just common sense that systemD have an age field. They already have a name field.

They’re also bombing luxury hotels and fossil fuel facilities.

Iran isn’t so much defending itself, as it is establishing consequences for the West by bombing everyone they can in the region (while still holding back to some extent).

If I remembered it, it wouldn’t be overrated.