[frenchie4111]

I think their point / the general FOSS argument is that those 1000 of man-years would be turned into 10000 man-years if these things were open sourced. A similar security concern could be waived at things like openssl, but it seems pretty inarguable that openssl is a net-positive for security.

[svorakang]

I'm all for open access to the code for the sake of safety. On the other hand, I'm completely against hobbyists accidently bypassing a safety mechanism.

Open access, but secure access to software download could make sense, at least for commodity parts.

When it comes to features with competitive advantage, though, I don't see that OEMs or its suppliers have anything to gain.

[jacquesm]

> On the other hand, I'm completely against hobbyists accidently bypassing a safety mechanism.

Accidentally.

Besides that: it should be fairly obvious that hobbyists are not going to 'accidentally bypass a safety mechanism', they can cut their brake lines as well and they don't generally do this. What you'd see is that the aftermarket would finally be able to produce stuff without dealers in between and people with the 'right' kind of tooling (authorized by the manufacturer) to get your replacement to be recognized by the firmware. Because of course absolutely none of this would ever be used to protect the bottom line. Right?

Also: if anything open sourcing this stuff would likely result in more rather than less safe vehicles, maybe at the expense of a couple of embarrassments. Because I have absolutely no illusion about the people working on these systems professionally to be somehow magically better than the ones that work on them for themselves, after all, they have a pretty big stake in the outcome.

Imagine that, working on your car in a safety related way... replacing brakes, steering housing components, linkages, suspension components tires and so on is all at least - if not more - risky than working on software.

FWIW one of those 'safety features' tried to kill me twice and caused me to let go of my recent car and switch to a 1997 issue vehicle that has behaved quite predictable compared to that modern one. Whose 'safety features' could not be disabled.

[lm28469]

> Besides that: it should be fairly obvious that hobbyists are not going to 'accidentally bypass a safety mechanism', they can cut their brake lines as well and they don't generally do this.

I can already picture the YouTube videos on "how to gain 15% hp" explaining you how to "hack" your car with a 1s "it will severely reduce your engine life expectancy" message at the end. Thousands of people would run this patch without thinking twice

Also how would you pass the controls most countries do every other year on cars ? I don't expect people checking my brake pads to know how to review the random piece of code I deployed to my car

[jacquesm]

Well, that's sort of the point: this is already possible, so in that sense nothing would change. Changing the mapping (essentially the amount of fuel injected based on a bunch of parameters) is regularly done by 'tuners' (between quotes because they don't really tune anything, they mostly burn more fuel for questionable gains).

But that's really not what I would care about. I'd like to read that code to figure out what the failure modes are and what might impact my safety in a negative way.

[mikrotikker]

I can go out to my late 80s truck truck right now, undo the lock nut on the fuel screw and wind that sucker up. It'll be a completely different truck. Will I half my fuel range? Sure. Will it be fun? Sure will be. If I care about the engine I'll attach an EGT gauge to ensure I don't melt the alloy head.