Hacker News Reader

Comment by nomel | Hacker News Reader

I think this shows more of a fundamental flaw in their update mechanism, than anything.

I don't think a botched update is a big deal. It happens, and should be expected, in a sane design. The fact that the customer noticed is a big deal.

There are many implementations that could be used for an "auto rollback" feature. They either failed to implement that in a sane way, or they were goobers, and assumed things would always be rosy.

View on HN

I would be pretty pissed if I went out to my garage to head to work one morning and found that a damn software update bricked my car overnight. This shouldn't even be a thing, why does a car need regular software updates to keep functioning?

It doesn’t. People and these tech companies are tools. And do it largely in search of ways to take more of your money. It’s not a favor.

If ICE tech was the hot new thing in cars, things like spark plugs would have a chip so that it would fire n times then break, but don't worry, there's a subscription for new ones and they will be automatically ordered when the car says so. If the credit card on file expires, your spark plugs won't work anymore, even if you just replaced them.

There's never been any car that 100% will work in the morning when you go to the garage. It's all tradeoffs.

Now we can add 'bad software update' to the list of things that can go wrong with cars. We didn't use to have that.

At the same time, we're losing tons of mechanical problems that use to go wrong with cars. The amount of time lost to car malfunctions is way down over the past couple decades, even with slight regressions like this one.

It doesn't need regular updates to keep functioning. It offers regular updates as they add new features. For instance, in this update a new feature was added to allow for proximity locking at home but disable proximity unlocking. That would lessen the number of times the car would lock and unlock accidentally as you walk in and out of the garage. No one was forced to install the update.

Cars 20 years ago, even most of them 10 years ago, never got any updates unless they got recalled. Nothing broke, nothing got hacked, and most are probably still working fine.

What happened to cars today? I refuse to believe that it's solely because these are electric cars, as if the way the car stores and uses energy dictates that it must be part of the internet of things.

Edit: there were electric cars over 100 years ago. I bet they never got software updates.

As software eats the world, it becomes more and more apparent to the non-developers of the world that software engineering is not, and never has been, a real engineering discipline.

Tech Support: "Oh your garage door is bricked after last nights update? Yeah, apparently the [totally uncredentialed] contractor that wrote that update is only 3 weeks out of coding bootcamp and was just copying and pasting from ChatGPT. Lmao"

Cars 20 years ago didn't have realtime traffic on big touchscreens that you can use to look up your destination and plan out a route that also lets you schedule fueling/charging stops, oh and also stream humanity's entire library of recorded music, books, and podcasts. It's a tradeoff that the vast majority of people want.

All that stuff should be done via smartphones and the screen in the car should be a dumb display for it.

Requiring a cell phone to replicate features a car should have just makes more problems IMO.

But almost every smartphone already has those features, they work fairly well, and it makes more sense for a phone to have those features than a car. (Because sometimes it's useful to know where traffic is when you're not in a car, and stream things while walking around.) It doesn't make much sense for each car manufacturer to replicate everything if they can figure out how to connect a phone to a car, and borrow the phone's features.

> Nothing broke, nothing got hacked

This needs to be heavily qualified or else it is outright false.

The Tesla update is slow probably for this reason. It is probably verifying that it can rollback at any point of failure.

I would naively expect it to just do A/B updates, which unless I'm forgetting something shouldn't incur a speed penalty? (Other than that the update doesn't get applied until restart)

I believe one of the reasons it is slow is because it is also updating the firmware on any number of connected ECUs over the CAN bus. This typically means the image has to be sent over a 500kbit/s bus so there is a limit to how long it has to take.

Reply on news.ycombinator.com