I'd be worried that once an app got a foothold into the product, the company would be unable to resist the urge to spread the app's tentacles across the entire vehicle, adding connectivity and telemetry and DRM, integrating it into the other car's systems, adding remote-this and wireless-that, and then inevitably the product would end up just like the turd cars we have today.
Not only the backend, but what happens 40 years in the future, when our phones don't run the app anymore, or we're all on phones that are totally unlike the phones of today, or if we don't even have phones or apps? I would expect the car to still work after that long, and making it dependent on a technology that is specific to a particular decade risks premature obsolescence.
I'll be ecstatic if my iron filter lasts 40 years.
I saw in another post that a person said there's a difference between "device dependent" and "device augmented" that really resonated with me.
There's diminishing returns on everything, and just throwing your hands up on any subject as bad/good might be a disservice.
If I live through an era where phones are no longer a thing and APKs are a thing of the past.. then I either...
A. Don't use the iron filter like that anymore. (manual programing now) B. Get a new iron filter. (ewwwww) C. Keep a legacy-device for the purposes of programming the iron filter. (doesn't need any internet connection or subscriptions)
(C) would be my most liked solution.
> ...what happens 40 years in the future, when our phones don't run the app anymore...
40 years? How about, like, 3 to 5 years? Remember when Apple decided to kill all 32-bit iOS apps for new hardware? I have an old iPod and iPhone 4S with "landlocked" software I enjoy using but can't anymore because Apple.
Phone manufacturers have shown they don't give a damn about allowing old software to function. Physical devices tied to software is a terrible idea.
Relying on a mobile app is relying mobile operating system compatibility over the years and is just asking for combinatorial methods of obsolescence via OS/app/library breaking changes, plus if your old phone breaks, etc. Open sourced mobile app with open sourced back end might be somewhat acceptable but otherwise it's just asking to be bricked as soon as one of the companies involved goes under as we have seen time and again just in past couple of years.
Not an app, but an API. And an app on top of that, if desired.
Also there are evergreen interfaces, so to say. An RS232 / RS485 connector that serves 115kbps 8N1 serial interface and runs a VT220-based TUI should still be serviceable 40 years from now (VT220 was released 42 years ago). A now-modern web-based GUI also has a great chance to be serviceable 40 years from now.
I have a iron filter that works via app. All configs can be done with button presses on the valve but in a much more tedious process/workflow.
It connects via bluetooth and not WiFi. If the company goes belly up, I'd just need the APK and an android phone to continue using the app to configure the valve and see/download water usage data.
Fast forward 20 years when I can't install the APK on android v79, I'd need an older phone to run the APK.. but that seems to be pulling hairs.
Apps would be great, it's how you handle the backend to it that's the gotcha.
I also have a water softener with an app that no longer works that had it's backend shut down. It can still be configured via the valve head button presses, but none of the "smart" usage data is available. As an example of good design, this is a perfect dichotomy of one company doing it well and one company doing it un-well[sic].