Along with the timeline in another comment; you should still shit on MS for breaking things around a SecureBoot "vuln". SecureBoot is just the advertising name for pushing DRM. It provides no security benefit against most attacks users would care about. (It helps against physical attacks and attacks that might try to persist through wiping the machine; ie the only real benefit is for companies that lock down corporate machines)

I mean, even if it is still pretty easy to switch to Firefox/Edge/etc, it's probably still reasonable to say that the market share of Chrome qualifies as monopoly (or separately, ads or search). AIUI, even if does qualify, that on its own is not illegal. What would be is then abusing that monopoly in certain ways. Like maybe banning ad blockers and privacy tools to support their ad business. Or possibly if they broke support for other browsers in search or other websites that they qualified as a monopoly in (they've massively cut down on internal testing on other browsers, but it seems unlikely that they'd manage to break search, and things like google docs probably don't qualify as a monopoly)

Secure boot maybe has some non-evil uses for servers/cloud stuff (if you trust Intel/AMD to not be evil *and* to be competent and don't trust Google/Amazon/MS). For consumer products the purpose is and always has been DRM. The "but it prevents rootkits" is always bullshit, all a virus/ransomware/etc needs to ruin is your user account and you've already lost everything you actually care about. (And for like a family machine that actually has multiple accounts in use, you're relying on windows security to prevent the infection spreading. Good luck with that)

I mean, the upside is that these datacenters (mostly) aren't exporting and hiding their costs. (Electricity generation's externalities are not the worst, but are often relevant) Companies trying to do the power-hungry AI training or running expensive models have to pay for that electricity themselves (and this provides direct incentive to try and find more efficient ways of training/etc). They might pass that cost on to the consumer, but that's in fairly obvious ways. Contrast all the cryptocurrency nonsense, where transaction fees and problems are generally less obvious, and the entire premise requires high power cost.

On top of that the main conclusion from "lots of new + more power-hungry datacenters = 2-6% increase in power usages" is not "this is an environmental crisis". It's "by Amdahl's law this isn't a particularly important place to optimize". It's new, so there might be low-hanging fruit compared to the things that actually cost energy, but when the absolute maximum improvement is single digit percentages, it's not the sort of thing you blame for any energy crisis.

You do not have copyright over the idea of your app and never have. That's not what copyright is for. You have copyright over the code of the implementation, and maybe could, depending on the random-ass decisions of your country's patent office, have a *patent* on the idea of it. Even the shittier patent offices hopefully would not give a patent for "make my phone into a flashlight", and the devs probably didn't try.

When a company acquires another like that it's to get the existing implementation and/or "hire" the people who came up with the idea. Chinese companies are probably being accused of stealing code/hardware designs/etc provided under NDA (or random nonsensical complaints too).

Even pure clones of an idea are not copyright infringement, that's not how it works, and is not how it has worked for about as long as computer games have existed. Atari lost that suit in like the 80s or something.

In general, the ability to maintain access (at least up to some expiration date longer than would otherwise be available) is the entire point of a session cookie - you don't need to log in repeatedly so long as you are still "there", similar to a screen saver password / lock on idle. And there is no possible way for anything to be stored in a relevantly encrypted form barring a setup where you do not actually control your computer (like iOS) combined with the web browser not being the compromised component in the first place. The web browser needs access to the unencrypted session cookie, and so anything that has access to what the web browser does will have access to the cookie. On a normal PC that's anything unsandboxed being run by the same user or a user with admin access. Things like denying debugger permissions by default only means that they have to wait for a browser restart or pretend to be legitimate for a UAC prompt / fake a sudo prompt / etc.

I mean, wayland's a pile of shit that doesn't implement a bunch of basic desktop features that even X11 implemented (screenshots) or had as standard extensions (all the NETWM stuff), and of course all the desktop environments immediately implemented incompatible versions. Then the wayland shills immediately started selling "lack of screenshot support" as a feature.

Yeah, it's nice in theory just keeping a normal keyboard+mouse. The problem is that the pxel density of current VR headsets (much less AR headsets) is just not good enough for this. You don't need any special laptop to see this, just whatever VR Oculus headset and the desktop view (steamvr/steamvr home I vaguely recall also has a similar desktop view for any non-Meta headset). At a typical virtual distance and size that you'd use for a laptop or desktop it's completely unreadable because it has a small fraction of the pixel density, so you have to blow up the size to be uncomfortably large and/or increase the font size a ton.

The GPL (and other open source licenses) gives you rights you wouldn't otherwise have - to redistribute the source code; if you aren't following the license you don't have that right (and doing so without that right is infringement). The analogy here would be a EULA, which is taking away rights you would have by default. In order for this to be infringement they'd need copyright on the URLs, or handwave that it's all equivalent because computers. (As opposed to merely "you aren't supposed to access this URL this way" CFAA nonsense)

It can prevent things like rootkits if it prevents things like custom linux installs, or requires manual intervention for every upgrade (please no), or lets you install your own keys (they never do that). If you think any of the options other than the last one are a good trade, gtfo, and the last one isn't what is happening.

The most obvious direct comparison is how C++11's move semantics are just laughably awful - in C++ moved-from objects still have destructors run and can still be accessed freely. (Also the default is still implicit copy rather than move, but C++ was always going to be stuck with that because of backcompat)

Sounds like just "authenticated to the ksmdb server", assuming you're hosting some SMB filesystems for remote use on your machine. So they _should_ have access to (some amount of) those filesystems, but not run code.

You do realize that said tax loopholes are probably at least half "no this totally wasn't profit" and would apply to said "flat" tax? "Make things worse for poor people so they get angry" is not exactly a good way to help.

That's already been done somewhat, and that's been enough that basically nothing uses actual popups or browser-made modal dialogs anymore. All the dialogs you see nowadays are just built in html+js that you can't easily detect outside of typical adblock sorts of heuristics.