Hacker News Reader

Comment by micheljansen | Hacker News Reader

It’s all fun and games until Apple wants 30% of your revenue. I feel Apple has seriously harmed the trust of potential partners too much for them to bet the farm on deep integration.

View on HN

If I were a car company I would try to focus on making good cars.

I'd leave the software to others, making sure to support both CarPlay and Android Auto.

There's a lot of software in a car besides the entertainment system, you can't leave that to carplay or android. To make good cars, car companies need to be able to write good software.

nope nopr nope.

Car manufacturers should engineer good cars.

Leave the good software engineering to good software engineering firms.

That is what GP is saying. There are two types of software in the car. The entertainment system and the software that drives the engines (ECU), lights, etc, etc. The car company should be good at software engineering to ensure that all software directly related with driving is done properly. And then leave the faster paced entertainment software to other suppliers.

Auto pilot, adaptive cruise control, engine start stop controller, brakes, gas, etc, etc. Modern cars can’t drive without software. 100m+ lines of code in a car, 70-100 cpus according to this article: https://medium.com/next-level-german-engineering/porsche-fut...

Exactly! That is the software I want the car engineers to focus on. I do expect most have their firmware provided by Bosch/Siemens or whichever supplier. But Tesla shows how much improvement there is to gain when designing cars from the ground up instead of from parts lists of a vendor.

At the same time the Polestar 2 is an example of a modern EV built using a legacy mindset, and it is great and spends way less time at the mechanic than my colleagues Model 3.

What do you mean with legacy mindset? The polestar 2 also contains many million lines of code, besides the software running carplay/android car.

Yep. If it were possible to buy a car with no native infotainment and just a dock slot for a phone and dumb screen for CarPlay/Android Auto, I would. Car infotainment turns outdated, unsupported, and near-useless at breakneck speed and is generally buggy lowest-bidder junk to begin with, so why not just skip all of that, save some cash on the BOM which can be passed on to the buyer, and just let me plug in my phone which will never have these problems?

Right. There is a lot of innovation possible in controlling all the little embedded systems.

But I think that "car platform" should be clearly separated from the "userspace".

My sense is that why the legacy manufacturers are failing, at software, they have coupled the "userspace" to the "car platform". Updating UI in the center console is the same thing to them as updating the embedded code that controls the windshield wipers.

That being said, I admit you cannot fully leave the "userspace" to apple or google, because you, for example, probably have some custom kind of sunroof that doesn't fit in whatever generic APIs CarPlay and Android Auto offer for Sunroofs.

Additionally, if you go all in on Android Auto, you have no leverage when Google tries to extract more money.

At the same time, competing with Google or Apple on UX and continuity is a losing battle. When I enter the car, I want my car to continue playing the podcast I was listening to. No car manufacturer can do this, only Carplay/Android Auto can.

I guess the answer is to have good support for Android Auto/ CarPlay. But also have your own version as an insurance against being held hostage.

Tesla seems to do just fine without CarPlay or android car.

0 growth at the moment.

That is one way to spin it. At the same time Tesla model Y is also the best selling car model of 2023 worldwide, so outselling all models that come with CarPlay available…

I would wager that there were more cars with CarPlay and/or Android Auto sold in 2023 than Models Y, or any Tesla for that matter.

Sure, and i also prefer carplay, rather than something provided by the car manufacturer.

But Tesla shows that there are car companies that can build satisfactory entertainment software, which is not surprising as writing good software is something every car manufacturer need to do, considering how much software is used in a car.

It’s superior to carplay…

Depends on how loosely you define superior. It doesn't have my podcast app. It cannot send iMessages. I won't buy another Tesla if they don't add CarPlay support, because there are plenty of good EV alternatives that do now.

I agree with the previous poster. If there’s one thing Silicon Valley has taught us, it’s the danger of allowing a tech firm to control the customer interface for your business: whether that’s taxi scheduling, food delivery, search, or automotive UX. Everything will go beautifully until lock-in is achieved, at which point the business switches to revenue maximization. I don’t know exactly what that will look like in the automotive world, but I suspect customers will pay for it in the end.

Most car manufacturers (with the notable exception being Tesla) think like this. A lot of the in-car software comes from third parties like TomTom. It makes sense until what happened to phones happens to cars and the hardware becomes increasingly a commodity and software the factor defining the user experience.

GM clearly does not.

They were in the news pretty prominently recently trumpeting their intention not to support Android Auto or CarPlay, but instead to roll their own system that they intend to monetize heavily.

Except they've gone away from standard sizes and replaceable head units in order to push people to new cars, so the software integration has become a driver of car sales and a big part of the "good car" experience. You can't assume you'll replace the stereo head later if you don't like it, so software features now influence the car industry.

Case in point, I have owned mostly Chevies through my life and have been mostly happy with them. But I won't buy the new ones because they've done away with plugging the phone in (or will in the future) and made their own. It's a cash grab to get our data and force subscriptions on us and I'm not buying it even if it's a good car.

The whole thing is absurd. We have to get a whole different car if we don't like a screen, but that's how it is now.

Almost certain when EV charging apps get embedded in carplay.

On a flip side - having one app instead of 10 would be nice…

Don't they (Apple) want a monthly fee from each install, causing the car manufacturers having to pass the cost on to their customers?

No, Apple doesn't charge for the existence of CarPlay in a particular car. The manufacturers used to for a while, probably to reclaim their perceived implementation costs and because the industry has worked that way for a long time, but that's not as common anymore, most (besides certain manufacturers) ship it as a default now.

Right - the problem is car companies ceeding that 30% to Apple, when they could be rent-seeking instead.

Isn't it rent seeking in both cases?

I worked for a car company in Infotainment for several years, and while I wasn't privy to any conversations like this, I can say that they (my company at least) are desperately brainstorming ways to come up with anything to make money. This is why we've seen all those crazy ideas like BMW trying to make seat warmers a subscription. They're trying to find some way to maintain a better cash inflow than selling a product.

I also spent time working with car makers while at Visa. There was a LOT of wishful thinking by carmakers that they could magic-up their own digital app stores (and pay for gas, for parking for drive-thru) all through their own proprietary infotainment platforms. There were these hopelessly unscaleable expectations that brands (McDonalds, Chevron etc.) would want to build and maintain custom apps just for, say, this model year's new 3-series BMW, and then do the same for every other car maker and model type. It was very much like the shortlived mindset of carriers trying to control downloadable mobile apps, prior to the iphone/appstore.

Rent-seeking by apple and rent-seeking by car makers are both legitimate concerns!

I guess the sentiment for the rest of us punters and hoopleheads is, if we have to be rent-sook by one of them, at least CarPlay is likely to have a less-unpleasant user experience?

If there are ten rent-seeking car manufacturers, you can conceivably buy your next car from a different company. If there are 1-2 phone OS makers (both coincidentally charging identical fees and duplicating the same business model) you’ve got many fewer options.

Depending on the specifics, 30% for a deep Apple partnership sounds like an amazing deal

Reply on news.ycombinator.com