Or just support a local API with a published spec. Matter is supposed to be that standard, but adoption is going slowly. Pretty sure zero people using the original reverse engineered cloud API for the Haier stuff in HA want to be using it in preference to a local API.
This seems naïve: "Yes, it's making money, but it isn't making enough money" is the phrase on the tombstone of a lot of goods, services, and come down to it, countries through history. Being a customer is no guarantee you get to keep being a customer.
> and drive manufacturers to make locally-controllable devices to appeal to price-sensitive consumers.
"Special Deal: Comes With Free Six-Month Subscription With Purchase!" because six months is how long they intend to support the device, after which they turn off the service and brick all of them, and of course you can't opt out of the special deal because they don't get any of your data from a device that isn't slaved to their servers.
I'm jaded enough to know it's not a panacea. It's further along the spectrum from "This is costing us money..." at least.
Collective action would end this problem overnight, just like it did here (although I think they are insane for buying cloud appliances in the first place).
What boggles my mind is how the average person seems so accepting of this. I get this pervasive sense that because "it's computers" or "computer adjacent" it's a given that you'll have to accept loss of functionality as devices age.
I guess it needs to get a lot worse before it gets better.
Needs to isn't the write term... Going to get a lot worse is going to happen. Getting better, well the issue is in doubt.
They don't know much about how those things work. They do know if they're losing functionality even if they don't know how they work.
My observation is that people seem endlessly patient about losing functionality in computer adjacent devices-- far beyond what they would tolerate with other devices. I simply don't understand why this is the case.
For virtually every consumer good there is some ability for independent repair. I think people would be up in arms if their car stopped working and the manufacturer said "Yeah, you just need to buy a new one". We have consumer protection laws in the US to prevent this very thing.
With software connected to "smart" devices there are no significant protections. I don't think people even know to be up in arms about a device that doesn't work because it's "old" and the native app stops working or the Internet hosted servers went away. Worse, they just landfill the device and go buy the newest one.
I assume it has something to do with the subsidized nature of smart phones and relatively short time people stick with a given phone. The "wow" factor of new technology probably dazzles people into forgetting about long-term usability too.
Hopefully when the "throwaway" mentality of smart phones really starts to cost people (as more durable good type "smart" devices embrace this model) we'll see some public backlash.
The relevance to this situation is that PetNet made a pretty good pet feeder about 5 years ago. It dispensed well, had a good app, and mostly just worked. Then, during 2020, the company just disappeared. The app stopped working, and eventually, the feeders stopped feeding. I tossed mine into the ever-growing box of "I'll get to it later", and forgot about it for a few years.
Getting it up and running wasn't tremendously difficult, and now it works almost as well as it did when its new, working significantly better in one critical area: Instead of having to sign into some other random app and fiddle with it there, I can just press a button in HomeAssistant and my cat gets a bunch of food. Maybe later I'll wire in things like the hopper sensors, scales, and other features that the device had, but for now a 1 second pulse of power works well enough
There's not really any valid reason PetNet couldn't have just released the firmware flashing keys on the way out. The device likely runs on a microcontroller very similar to ESP32/ESP8622, and so had those keys been available, I could have used the "native" hardware.
I suspect the overwhelming majority of these feeders wound up in landfills, as very few people, even the technically competent, are willing to take it apart, solder in a new control board, and get said board up and running. Which is an absolute shame.
[1]: https://pdx.su/blog/2024-01-19-fixing-a-broken-smart-cat-fee...