> X11 was a security disaster.

This only matters if you compare properly sandboxed apps, otherwise an app that runs with your uid can still do harm and practically indirectly completely compromise the system..

Are most flatpaks _properly_ sandboxed? Of course not.

And X11 always had a mechanism for isolating clients as well, i.e. trusted and untrusted clients. Nobody used it because it was irrelevant before sandboxing.

> there is always, always more work yet to do

And they always, always forget that it's not about "work", it's about whether a particular person will be able to contribute work that someone is willing to pay for. It's definitely NOT true that there'll always be more paid work to do that can be done by a particular person.

But this is what you get when these authors are wondering if something is good for "the economy" instead of thinking about actual people.

Yeah, all these "work has always been fine!" writers forget that we've never invented cheap artificial people before.

Since after doing this Google knows the user knows what they're doing (and officially they say they don't want to get in the way), why does this only enable installing unverified apps (still unprivileged), why is the system still insanely locked down? I thought the 24-hour delay solved the "security" problem?

I didn't know a single company could just pay politicians state-by-state to pass a given law - in my country that would be a crime, but it seems in the US this is how the legislation process works :)

So TempleOS is still illegal?

And well, the law represents an intent.. if self-reporting won't work (obviously won't), then the scenario where PCs end up as locked down as smartphones is not far fetched.

Just as open source software represents the intent that all for-profit activities shall be illegal...

No no no, wasm has shitty speed if you want to emulate something (it doesn't even support JIT), the problem is in its architecture (tons of restrictions like no self modifying code, no jumps).. this can't be fixed, we need something real, something like WebKVM.

On the web you can dynamically create new Wasm modules and use JS APIs to load them, though there are ergonomic issues. There are per-module costs and systems like CheerpJ and CheerpX currently do batching of multiple functions into a module to mitigate the per-module costs.

I've created a proposal to add a fine-grained JIT interface: https://github.com/webassembly/jit-interface

It allows generating new code one function at a time and a robust way to control what the new code can access within the generating module.

Wow, thanks for this, this is exactly what v86 was missing! Runs faster than my demo: https://zb3.me/qemu-wasm-test/jspi-noffi/

Even though it has no JIT. Truly magic :)

Nowhere, because there's no such department..

Is there an up-to-date list of their phones which allow bootloader unlocking? Not all of them do..

Motorola? Is there an up-to-date list of devices where they're "so kind" as to allow bootloader unlocking? Because it's a lottery to me..

It's always a hit or miss with Motorola, but this should up your chances:

https://github.com/zenfyrdev/bootloader-unlock-wall-of-shame...