Whenever you do anything on the web your browser is talking to a whole stack of machine specific stuff, graphics drivers etc. The browser is just an abstraction layer on top of this.

I don't see how this would necessarily be worse than the current status quo. Big content providers who demand DRM are just using things like Silverlight instead which don't provide any official support for any of those OSes so they can't access that content anyway (without some wine hacks).

At least this way it is possible to produce a Linux variant that does support these things without having to ask MS for a silverlight port.

> At least this way it is possible to produce a Linux variant that does support these things without having to ask MS for a silverlight port.

Instead we'll have to ask Netflix to port their EME to Linux. Not going to happen either.

These abstraction layers solve problems. Propietary DRM binary blobs try to solve a problem that is impossible to solve: how do you show content to someone and simultaneously hide it?

So the only way out is to make it inconceivably expensive for an attacker, which leads the way to obfuscation, and eventually, as we have seen, invasive technology like kernel modules and other nonsense. They keep on adding abstraction layers, but the other way around: going ever deeper.

I personally see no value in providing an API to someone trying to solve the halting problem. Especially not if the next step is having them pressure us into installing their distributed halting problem solver.

Well I agree that it largely seems like a waste of time, but I don't get how it necessarily makes anything worse.

If you don't want the kernel modules etc you can still use an OS that doesn't package them or allows disabling them. You will still be in the position you are now.