Live data from Hacker News

Internet-connected cars fail privacy and security tests conducted by Mozilla

gizmodo.com

221–230 of 660 posts

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#221
Note that this is all based on this source material. https://foundation.mozilla.org/en/privacynotincluded/categor...

I haven't dug too much into the methodology, but it seems like it's done based on privacy policies rather than actually looking into the car telemetry traffic. It's also written in a very caaual and sensationalist "omagad" tone that doesn't serve well the seriousness of the topic or findings

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#222
post #212

Earlier quoted context omitted.

The touchscreens are really a dealbreaker on new cars. They’re not just prone to quick obsolescence, they’re dangerous.

Is this a US thing? I recently bought a reasonably high end Skoda (which I think would be made in the same factory as VW, SEAT etc). It has a big touch screen in the centre console, but that's really only used for the radio / media / phone calls etc. This is all non essential stuff and so I can deal with it being on a screen. Everything relating to driving is an old skool analogue control. A couple of things I don't…

For me controlling the radio using a touch screen is an issue - i cant take my eyes off the road, and using driving wheel buttons that require fiddling is not great either. I need buttons and dials that leverage muscle memory.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#223
A lot of people are making comments about defeating the telemetry, but it's not just about selling ads this time. This is about actual surveillance and (eventually) control, and the governments of the world will make this a requirement, and make circumvention of it as illegal as copyright infringement. In the US, the government will throw up it's hands and say it's not us; it's the "free" market, and this must be what people want, while all the automakers collude to do it, collect the data, and either let the NSA have it, or don't resist when they tap into it. And some people will think this is a good thing, because then we could throw literally everyone who was at J6 in prison, but then a conservative government gets thrust into power, and now they can go after everyone who was present at a BLM protest that turned violent (but I repeat myself). Whatever power "we" let "them" have will eventually be used against "us."

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#224
post #17

Earlier quoted context omitted.

It’s an eSim. Not a physical SIM card afaik.

Do you have a source for this?

Google. I’ve searched and searched. I have a 2018 Truck from a different manufacturer and I completely went down the rabbit hole of attempting to remove it. It’s not possible. It’s an assumption that most manufacturers are following the same logic due to economy of scale.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#225
post #197

Earlier quoted context omitted.

I think they put in the touchscreens because it makes internationalization easier. No physical labels and OTA updates. Personally, I don’t think digital internationalization is good for the driver since it appeals to the lowest common denominator of interface while adding another expensive component to replace in case of damage.

I wouldn't even go that far: It's less wires to run. Yea; my cars interface has had 3 radical redesigns since I got it (which I mostly appreciate), but think of how much easier to build a Model 3 is when it foregoes 95% of a cars normal physical buttons for a single screen.

It shouldn't be expensive to make an I2C smart-switch that allows you to place as many buttons as you want on a single set of 3 wires (or 4 if you want a simpler circuit).

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#226
post #6

I think it will be a good differentiation factor in a few years, that a brand comes up with an offline car, i.e. a car that you just refuel/charge and drive - no telemetry/connected features involved. It may be a niche thing in future, but certainly something that would be appealing to me as a consumer.

Open source Car!!

The rally fighter and other local motors vehicles are open source iirc, but I agree this should be the future.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#227

The more I read about these things, the more I think I'll be driving my 23 year old Toyota 4Runner until the end of my life

Except when you live in a city where they start to limit and ban transit of older cars, to force people transitioning into lower emission models, or public transit.

Like in Spain (through rules ultimately coming from Europe) there is a class of vehicles which are gradually being kicked out (banned from crossing certain very ample boundaries around the city): gasoline cars made before 2001, and diesel powered cars made before 2006.

For example, your 23 year old Toyota 4Runner would be deemed too polluting (or noisy, or both) to drive near the city center and auxiliary accesses of Madrid, and starting from 2025 it will be outright banned from driving on any part of the city, with a circle area of ~23 Km (14 miles) diameter from the center.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#228

Of all the horrible things go on with privacy these days, this is the one I hate the most. I’m a “car guy”, but not the sort that obsesses over old cars (although I do love 60s and 80s cars). I like new tech, I like the advances in engineering we’ve made in new vehicles, I like EVs even. Nonetheless I’ve been in the market for a new car for months and haven’t bought because it’s hard to find any cars that meet my req…

Kia driver here. There is a screen in my car, but also actual knobs and steering wheel controls for everything I might need (audio controls, climate controls) and when I use the screen to project my phone, almost all functions there can be activated through voice control.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#229
post #189

Earlier quoted context omitted.

I assume it's a cost thing. Slapping a tablet everywhere and letting the code monkeys figure it out is probably cheaper than making various knobs and buttons. I dread the day I have to get a new car. Even Subaru, who are usually behind the curve, have gone all touch screen.

The feds have mandated a screen (for backup cameras). Then the makers try to minimize costs by having the screen do everything. I’d like to say I’d pay more for real buttons, but I’d never buy a new car.

It's probably true that once you (more or less) need a screen for a decent backup camera and most people like a screen for at least GPS, it must be pretty tempting to at least think about what physical buttons can reasonably be eliminated given that the touchscreen is a given. And I do think a lot of designs go too far.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#230
post #197

Earlier quoted context omitted.

I wouldn't even go that far: It's less wires to run. Yea; my cars interface has had 3 radical redesigns since I got it (which I mostly appreciate), but think of how much easier to build a Model 3 is when it foregoes 95% of a cars normal physical buttons for a single screen.

It shouldn't be expensive to make an I2C smart-switch that allows you to place as many buttons as you want on a single set of 3 wires (or 4 if you want a simpler circuit).

https://en.m.wikipedia.org/wiki/CAN_bus

Modern cars do it with two wires.

Post reply on HN