[ryukafalz]

I mean, only if control of that private certificate is required to run code on the device. I'm sad to live in a world where allowing people to run the software of their own choice on their own computers is considered a "huge no-no."

[sn_master]

sigh..

Its not about "their own computer". Its about consumer "devices", including ones where control should not be given to the consumer for whatever reason. e.g. electric cars or medical devices.

Think of what would happen if anyone was allowed to run their own AutoPilot software on public roads.

Stallman is a radical and he doesn't pay attention to real business needs of large corporations, which is why they all avoid GPLv3 like the plague.

[gopiandcode]

Requiring that a company provides access to the consumer doesn't preclude the existence of regulations ensuring safe usage.

If a user/team of developers were able to develop their own AI system which could operate to whatever safety constraints are set out by the law, on what grounds would you deny them from running their cars using the AI?

Similarly, we can all execute whatever code we want on our laptops, which also allows us the potential to cause harm by attacking other systems, but doing so would still be illegal.

[javagram]

> Think of what would happen if anyone was allowed to run their own AutoPilot software on public roads.

currently, they are? https://comma.ai/

This is more a matter for regulation than for copyright, I think. Even if the car's ECU and AI code were under GPLv3, governements could prohibit using non-certified versions of the software on public roads and levy hefty penalties and jail time when drivers are caught using wildcat software...

[ryukafalz]

>Its not about "their own computer". Its about consumer "devices", including ones where control should not be given to the consumer for whatever reason. e.g. electric cars or medical devices.

Why should the owners of these devices not be allowed to control them? What makes them special?

>Think of what would happen if anyone was allowed to run their own AutoPilot software on public roads.

Sure, there are plenty of ways to modify the software in a car in reckless ways that endanger those around you. But it's possible to do a lot of things that endanger those around you: you can drink and drive, drive while tired, install headlights that blind other drivers, etc. We have laws to discourage that, and they probably need updating for our modern world of software-driven cars, but at a certain point we rely on the social norm of not recklessly endangering yourself and others. (And if nothing else, the owner of the car is liable if the modifications make it unsafe.)

This is not to say I think people should be able to do whatever they want on public roads; they're not yours after all, there are plenty of restrictions you have to abide by to use them, and I wouldn't mind having a software review requirement. But the car is yours: you bought it, you should be able to do whatever the heck you want with it otherwise.

[Dylan16807]

> Think of what would happen if anyone was allowed to run their own AutoPilot software on public roads.

Nothing would change. You're responsible for what the car does.

[umvi]

> I'm sad to live in a world where allowing people to run the software of their own choice on their own computers is considered a "huge no-no."

I think it's a little more nuanced than that. Should I be able to run whatever software I want in my printer? Absolutely.

How about the software in my satellite internet radio transceiver (one rogue transceiver could take down the entire beam/jam the satellite)?

How about the software in the micro controllers in my car? Should I be allowed to modify that code too (if I mess up people could die)?

[Dylan16807]

> How about the software in the micro controllers in my car? Should I be allowed to modify that code too (if I mess up people could die)?

You're allowed to modify any other part of the car. Why should software be different?

> How about the software in my satellite internet radio transceiver (one rogue transceiver could take down the entire beam/jam the satellite)?

There can be some flexibility for the part of the code actually doing the transceiving.

Everything on top should be replaceable.

And radio is really a special case, because it's a shared commons that requires a lot of expertise and where one bad transmitter can cause problems for miles or more.

[NullPrefix]

>How about the software in my satellite internet radio transceiver (one rogue transceiver could take down the entire beam/jam the satellite)?

This sounds like a failure on the satellite part. Can you do satellite comms with SDR?

>How about the software in the micro controllers in my car? Should I be allowed to modify that code too (if I mess up people could die)?

What if people die because you cannot modify that code?

[umvi]

The problem is physics. You have a bunch of people in fairly close geographic area transmitting to the same point in space on the same frequency band. One bad actor can interfere with everybody else's tx.

> What if people die because you cannot modify that code?

I think it's far better to trust unmodified safety critical code that's been tested than to trust Joe Blow's precariously modified braking code that allows his Honda Civic to drift around corners better.

[wbl]

That's what laws are for.