I’d be a fan of a law that companies who drop support of their product would have to release code that lets 3rd parties or users themselves offer alternative support. If you want to fully abandon a product opensource it. If you’re a big company that doesn’t want to do that release a feature for users to self host before you cut ties. I know it’s not a simple thing to do in the current world but if laws mandated it then tech would have no choice but to adapt.
Effective [some future date], in order to sell any device connected to the Internet (or Bluetooth, or whatever), you must register your entire codebase and all internal documentation with the FTC, and keep it updated, along with any signing keys to lock bootloaders. The day you abandon support, if you haven’t provided everything required for end users to take complete control of their device, your code base and any other IP enters the public domain, and the FTC uses their discretion on release of keys.
It would take new laws, and you’d have to be careful with language and structure to prevent abuse of “third party” code and abuse of corporate structure to try to prevent old devices from being usable, but you could do it.
This sounds like a security nightmare though. A central repository of all code and keys is a gold mine for exploitation. Don’t get me wrong, I would really want this to work, but if it was compromised it could he catastrophic.
I do think there should be regulations in place that are clearly and easily enforceable by the FTC though. I’d love to see companies be hit with fines and/or compulsory refunds if they stop supporting devices and don’t provide some path forward for customers to keep using the device. That doesn’t solve for startups that go out of business, but it would at least cover the tech giants who are doing this garbage.
This is essentially the premise of Stop Killing Games but in a different world.
This is a commendable goal; though it would still rely on good faith that a lot of these companies won’t have.
They’d rather screw the users anyway, sell the IP and let it rot within the maws of some holding company.
We’ll need some clauses that the tech cannot go inactive as it trades hands as well.
Further, some teeth will be needed toward feature deactivation, as there’s nothing stopping a company from yanking features and packaging it up as efficiencies made or product evolution.
Just because a product went defunct does not mean the entire code base is obsolete to the company.
Suppose I release software that makes a profit for a while, then falls off and starts costing me money, obviously time to retire that thing. However, a ton of code in that original product was a stepping stone for newer projects. I now have two choices.
A) Drop support and give world+dog my code, giving everyone a look into my existing products.
B) Keep losing money on the old project and make up for it by overcharging for my latest work.
That’s a lose-lose proposition.
Your self-hosting solution sounds mighty fair!
I’ve been in software for more than 20 years now. I’ve done some pretty innovative things from time to time. There is nothing I have ever done or seen in any proprietary code base at any company I’ve ever worked at that isn’t at every other company. The only unique thing at any company is how all the puzzle pieces get connected. It’s pure ego to think that any idea you have in that now open source project is unique or what’s giving you any competitive advantage in your other projects.
If there’s nothing unique or special in any company’s code base, then why does SAP software suck so much more than anything else?
I’d like to see a requirement that products and devices which have been deemed by their manufacturer to be end of sale/support/repair/life are required to be unlocked, with technical schematics and repair documentation made freely available, upon request of the owner.
I support it at face value, but as someone who knows little about what this exactly means/involves - what are the risks for misuse by others if everything becomes publicly available and stops making developments?
A big problem is things tied unnecessarily to an internet service. We need to educate people that there may be alternatives and we need our purchasing decisions to support that. For example, most home automation stuff should NOT require or use any internet.
The article calls it “software tethering”. If any support commitments encourage manufacturers to stop that, we’ll all be better off. Let’s start with requiring users be clearly notified of software tethering, so they know what they’re buying
this is so infuriating
I have a Hatch, which i have programed for my babies bed time, just play white noise sound at this time, turn off this sound in the morning, play bird song when hes supposed to wake up
I specifically got the older model, since the newer model has a lot of these basic features locked behind a paywall, while the old one they are just free.
Went traveling, and without a wifi connection it wont even do this. Apparently making an alarm clock remember its settings without a wifi connection is too much work.
In October of last year my mom came home from the ICU, now unable to get out of bed. I replaced all the bulbs in the house with smart bulbs and put the fans on a little smart plug thing. It made me really like the idea of home smart home features, but I’m not techy. They’re just Alexa enabled for her to use with the fire stick, and I use google home on my phone for em.
Can you offer any advice for ones that don’t require internet? Every time our power goes out (any time there’s a storm), I have to go around and reset them while they flash at me like the worst night club
but I’m not techy
That’s the entire problem. Automation products depending on the cloud do have fewer requirements, are simpler to get started with, even if they’re overall poorer choices. They’re a fundamental fact that something needs to control your devices: cloud devices means the company takes care of that for you.
Taking the cloud out of the loop means you need to manage some sort of automation hub. There are many choices but I don’t know which may be approachable for non-techies. The new Matter/Thread standard tries to solve this while making everything work together, but adoption has been slow so you may not be able to use it for much yet. Your choice of automation hub drives your choices.
Maybe some of what I do is applicable, but I AM techie, I love to tinker, and I own my home, so maybe not. I try for local control where possible plus make choices that are additive: things still need to work normally.
- Prefer smart switches, not bulbs, so they still work normally. They also act as repeaters for local Zigbee or Z-Wave networks
- Z-Wave and Zigbee are local low-powered mesh networks commonly used in home automation. “Thread” is a newcomer but not widely used yet. Think of them as a network between Bluetooth and WiFi, but designed as a mesh to better cover your home. These are local networks only, there is no cloud connection (directly). They use a lot less power than WiFi, which makes battery devices more practical, and different frequencies for less interference. You should expect to pick one.
- Home Assistant (HA) on a Raspberry Pi. This was always the techie choice as opensource software that supports everything but in the last two years they have made it a lot more approachable, and Nabu Casa sells it installed in custom hardware. Among other things, it lets me use all of Z-Wave, Zigbee, and Thread devices, as well as Ethernet devices and gives a lot of automation power. When you’re starting out, you should stick to one for simplicity
- I do use Alexa, which requires internet access. I choose to subscribe to Nabu Casa to manage that in HA, since it also funds further HA development but this is not required and there are no ads or data collection
- My latest automation was for my teen staying up late all summer: at midnight, dim the family room lights and announce over Alexa that he should go to bed. He’s back at college now but I’m sure I’ll want it next time he’s home
Personally, instead of smart bulbs, I’d use smart switches for automating lighting. There’s no need for every bulb to be individually controlled and carry all of the overhead involved in that. On that note, I’d also love to see DC circuits that can take LED bulbs without needing a transformer for each bulb (which tends to be what causes it to fail IIRC).
Just tried looking at the state of the smart switch market and fuck Samsung for naming their app for transferring files from phone to PC “smart switch”. Especially because there’s plenty of ways to do that already that don’t require a shitty Samsung app.
Excluding Samsung from the search, I’d suggest not looking for products directly but finding enthusiast communities that are building their own smart homes. There is more to it than just getting devices that don’t rely on some specific company’s web services. You’ll need to also setup a controller/server, connect all of the devices to that, and then figure out how you want to interact with it (eg via phone, scheduling, voice commands, etc). I haven’t done this myself, but I’m guessing all of these are solved problems, but doubt that anyone would call setting it all up easy.
Bulbs v switches is actually why it took me so long to get them to begin in. Having bulbs, inherently disposable things, full of tech seems… Less than ideal. But where I live, the electrical is rough. Like, we have 1 circuit, the light circuit. Everything else is messed up and turned off at the box, including all the 220s, except for the water heater. Run everything off power strips and have to turn everything off to cook. Lol. I don’t want to mess with the electrical in this house. I want to move. Lol
There are several competing protocol: zigbee, zwave, and the newest is matter. They just require a hub to be controlled.
I use all zigbee smart switches, and a conbee II usb stick. I have a old laptop running homeassistant os, and it has a one click addon for conbee II. I think amazon echo also have builtin zigbee controller.
If the power goes out, all the switch automatically connects back to the hub when the power is back. Obviously, this setup is not affected by google, amazon, even the entire internet goes down. As long as my router is functional, I can control everything using my phone, when I am at home.
To be pedantic:
- newcomers are Thread and Matter, based on IPv6 and include standardized device profiles that should really help things work together. The standard is widely sponsored, including by Apple and Amazon
- Matter works on your regular ethernet or WiFi network, and can access the internet. This is likely used by things permanently wired to power. For example Apple HomeKit uses this to control devices like light switches or printers
- Thread is a local low powered mesh network, sometimes called Zigbee 2.0 because it’s the same frequency and comes from the Zigbee Alliance. It is very suitable for battery devices like lightbulbs and sensors, and is bridged to Matter for controllers
Philips Hue have the ability to work without internet and it seems like a lot of people like them, though they are kind of expensive. I’ve used Sengled bulbs before and they were fine, not sure how well they work without internet. But I think for you the problem isn’t the bulbs reliance on the internet, they just seem very forgetful. For both Hue and Sengled, when the power comes back on after being out, they just start working again on their own after about a minute. No need to reset anything.
The main non-techy issue even for locally controllable smart things is that the big voice assistants are all entirely internet dependent. So even though Hue bulbs are technically controllable locally if your internet is out, Alexa and Google Home both won’t be able to do anything with them.
At least make it required to not brick at EOS if it’s a device that would otherwise run. Like a laundry machine.
There’s no reason a laundry machine requires an internet connection
- if an internet connection provides additional functionality such as notification, it easier to have the machine work normally without notifications
- there’s no reason a machine requires an internet connection, especially with the release of the Matter/Thread standard to unify home automation local protocols
When I got new machines about five years ago, I briefly considered connected machines. It would be really nice to get notifications on my phone but how can it possibly cost that much and why does the only option depend on a cloud service?
I think a LAN-connected machine would be good, if you could use an app or open-source, well-documented API to control the machine, but there is no reason a washing machine should need to connect to the outer internet. You can VPN into your local network if it’s that necessary to control it away from home.
It’s not only that, the issue is that they release updates that slows down the device, and you get so irritated that you buy a new “faster” device
I’m not an IT expert, so I don’t know how feasible this is, but they should have at least two branches of updates. One that is strictly security and the other that are all the bells and whistles. I would love to stick with an old os if it’s not vulnerable to the outside world.
It doesn’t need security updates if it doesn’t connect to the Internet*, and most of these things really shouldn’t need to anyway.
I just recently returned some smart outlets that would only work with my local Home Assistant if my Internet was up, because it all went through some cloud server. Now I’m using a combination of Zigbee and Z-Wave devices, which all connect to a local hub I have on battery backup. Now everything keeps working even if my power and Internet go out.
*Caveat: Wireless protocols could need updating too, but at least nobody’s going to recruit my Zigbee devices into a botnet. Worst case my neighbor figures out how to turn my lights on and off.
Or they take out 70% of the features it had at launch and make you wonder why you still use the thing anymore.
Apple Mindset. I gave up on the iPhone after v3 bricked within a month of the v4 release.
Android does the same shit, and I hate it.
I’ve had to give up phones that still worked fine because the app updates killed the responsiveness
Android does the same shit, and I hate it.
My OnePlus has been solid for the last five years. Idk about the Pixel, though.
Products that change the service they offer after you purchase it should also be eligible for a full refund (plus a % on top for the hassle). Such as offering a service through the product for free at the time of purchase but then moving it being a monthly subscription paywall later on, or just removing the feature completely.
For the first example, absolutely. If some execs have a meltdown, it could change future services but anyone who was promised Disney+ on their Tesla with no limit on it should get a fair refund. I understand that there’s a slippery slope argument here, and no– the value of Disney+ in a car isn’t 100% the value of it. But it’s BS that a manchild having an Internet meltdown loses people a service they had and “paid for”
Might sound stupid, but perhaps then they shouldn’t be offering services like Disney Plus and instead simply offer a car that lets you download any streaming app you have your own subscription for.
