After 16 years, Ecobee is shutting down support for the original smart thermostat
After 16 years, Ecobee is shutting down support for the original smart thermostat
Thermostats are easy to change out. So this isn't a huge deal. But I don't love the idea that tech isn't built to be self-hosted or maintained in any meaningful way. If you're not shipping an open source version of your software when you close up, you're an asshole.
Yeah, self hosting isn't for most lay people if it's just a GitHub repo. But GitHub repos quickly become adopted by nerds like me who build tooling around it that eventually let lay people self host software with the click of a button.
Yeah, self hosting isn't for most lay people if it's just a GitHub repo...
If ecobee put their backend code on GitHub, I bet it would be self hostable with docker within a week.
It is also nice that these just degrade to regular thermostats. It isn't like they are completely stopping working. It would be nice if you could swap out the API, or they keep the API running longer (how much work can maintaining it be?). But this sounds like a pretty graceful degradation.
It would be nice to have these speak some common Zigbee protocol or similar. But this isn't the worst behaviour I have seen from companies.
Are you aware of a decent number of mainstream products that didn't go full asshole? I agree with you absolutely, but I feel like the majority of connected products pull this same shit.
Yeah the majority do it and I think it's bad.
This is pretty much what happened with HomeAssistant. Tying all the integrations together in one platform.
It's now at the stage of "copy these files to a pi/buy this box we make"
The overall aim is to integrate most open things, and find ways to work with/around more closed off products.
16 years? That's like 8 separate Google project lifetimes.
That’s 591 Mooches haha
They just killed my nest cameras, but the thermostat is still supported. I was planning on replacing it with an ecobee this year just because API access is kind of a pain but this is giving me some second thoughts.
The newer Ecobee's can run entirely locally through their homekit integration. I tie mine into home assistant and use it that way. I would never have bought the device if that wasn't available.
If this old version doesn't have that available then I'm assuming people purchased it knowing that it was reliant on cloud services. It would be nice if they offered customers options besides just letting the device turn into e-waste but you can understand why they don't want to burn development hours on a device that's a decade and a half old.
Just another reason to never buy devices that can't function without a cloud service.
I don't understand the mindset of people who buy these things in the first place. Occasionally there's an article like, "guy's entire house suddenly inoperable after Amazon ban," people just don't think that will happen to them? It is local control on a standardized protocol or nothing for me.
Today yes, not 16 years ago.
Probably because this came out 16 years ago, before HomeAssistant even existed, it will still maintain wifi support for checking and controlling with your phone even after they cut off the cloud connection, and all their new products do have a local API and can still be used with HomeAssistant or whatever other local home automation server you have.
Yeah. The server and software should be open source and API available. That way we actually own the system and don't have to just toss it out if someone goes out of business.
Fucking ulock for example suddenly wants me to create an account and sign in to their website to use my front door lock! What the fuck is that! We need consumer protections for this sort of shit. I didn't sign up to giving away when I come in and when I go out of my house! WTF to the max!
16 years old? That thermostat has sure had a run, must have been designed pretty well to last this long without some electronic failure.
Assuming it's cloud connected, anyone aware whether it got updates for the newer versions of TLS and root certificates? As an example I'm aware quite a lot of android and similar devices from that era have expired certificates now, and outdated/vulnerable SSL libraries...
Edit: Edit example
For a thermostat that's built into a house, 16 years doesn't seem long enough, tbh. A 'dumb' thermostat can easily be in use for 30+ years before anyone would even consider replacing it.
But yeah, as you said, if it's connected to the internet you have to worry about software patches, certificates, etc.
Yeah, the old dumb ones in my house have been there for 50 years.
Yeah the thermostat that was in my home until ~2010 was added in the 20s lol. 16 years isn't long at all.
16 years old? That thermostat has sure had a run
I have game consoles that are more than twice that old and still play reliably. Apple really skewed our idea of lifespans for electronics, didn't they? It's a thermostat, they should be designed to install and forget for the next half-century. It's a core part of a house, like the plumbing and breaker box.
Didn’t the pace of change influence our perception more than anything else?
Don’t old computers on old operating systems work as well as they did when support was dropped? Much like your example of consoles?
The rate of software gobbling up newly available resources seems to a big reason people feel the need to move on. But I think that is starting to flatten out as the pace of processor improvements slows.
The bloat on the web is a huge burden on older devices too. Especially for your average person.
The bloat on the web and in native software (and in non-native software that is just another copy of Chrome posing as native software) comes from our newly available resources allowing for “lazy” development practices that prioritize cross platform development and other factors over writing efficient native software for each platform.
There are a lot of factors involved in the rate of device turnover. I don’t think any one factor is consistently forcing people to upgrade hardware. It’s a collective situation.
I use my desktop computers, for work, for nearly 10 years past their introduction date before replacing them. (Three more to go for my current machine). For my gaming computer, I swap a major part every 5-6 years.
And to reiterate, I think this rate of change is slowing down. At least for raw processing power and how long it is relevant. The rate of change over the course of personal computing has been massive. And it’s just starting to slow down.
Apple really skewed our idea of lifespans for electronics, didn’t they?
Apple's a weird pick for this.
If you're talking desktop/laptop hardware, I had a 2009 MBP running just fine as a personal server until a couple of years ago and would probably still be doing it except the battery turned into a spicy pillow and I wanted more performance anyway. And I've got a 2016 that's going strong as a daily driver for personal projects.
If you're talking phones, that's even weirder. It's pretty well known that Android users change phones more frequently. Which makes sense, cuz Android phones tend to get stuck on old major versions and stop getting security patches.
For instance if you got an iPhone 5s in 2013, running iOS 7, you could still be using that today on iOS 12, which received security patches as recently as 2023.
If you got a Galaxy S4 in 2013, you could update from Android 4 to 5, which stopped receiving security patches in 2017.
Ah yes. Apple, the company with the longest support windows for secure patches of any phone/tablet manufacturer, are definitely the ones skewing our ideas on the lifespan of electronics.
The company is offering affected users a 30 percent discount on a new Ecobee thermostat, valid for up to 15 thermostats.
...
While I very strongly agree with your message, I have to say that this is one of the least fitting usernames I've ever seen.
The company should be giving away new ones, but that’s none of my business [Kermit meme]
I wouldn’t be surprised if someone reverse-engineers the protocol and codes up their own replacement backend as a one-file Python script in a weekend.
That truly depends on how secure Ecobee made it… I’ve seen some smart devices that use SSL (https) for all communication and do some sort of certificate authentication, making it virtually impossible to decrypt its communication protocol without a valid private key…
Having said that, it’d be nice if Ecobee took the initiative and opened up these older devices, if they could do so without comprising the security of all their others.
In the last 16 years there's been multiple SSL vulnerabilities, so if someone was motivated enough, they could probably hack it, especially considering they'd have physical access. You could probably even dump out the filesystem and overwrite certificates with your own.
Is the firmware enclosed in a SOC with no way of reading/extracting it? If not, if all else fails, someone will extract it and dissect it with Ghidra or something, extracting whatever encryption keys are needed. If so, and there aren’t any documented side-channel attacks for reading the firmware from this SOC, if firmware updates exist, they too constitute an attack surface. (They probably would be encrypted, but how strongly?)
That's ok if you ask me, considering that they will still continue to function as regular thermostats
I disagree - definitely not OK by me, though likely legal. People bought this because they wanted and paid extra for an internet connected device, and a regular thermostat is not that. I mean, would you be OK if your TV manufacturer disabled the screen and streamed radio stations instead?
The last TV that would've lasted 16 years was probably made 40 years ago
If my TV was 16 years old, and the manufacturer cut off the internet function to it, id be ok with that.
These thermostats still work as thermostats, just without the smart features. Comparing that to turning a TV to a radio is disingenuous. 16 years is a long time, and there are security protocols amongst other things that go obsolete over time and can't be updated at a certain point on legacy devices.
I didn't say it was ideal, but it's ok. And it's definitely better when compared to other companies. Sure, in an ideal world they would have published the source code for their server ensuring that anyone could run their own instance at home. But we don't live in an ideal world.
If you bought one of these because you have a heat pump and want to consider the outside temp, that service is now cut off. Not ideal.
Why do we allow this? Companies that contribute to operate should be severely penalized if they don't A)) continue to support legacy products B)) offer FULL replacement if servers/apps shutdown or C)) open source EVERYTHING is they are going to try to kill a device.
As per the Ecobee notice this only impacts the two original models and they still function as regular thermostats still, they are just not providing any of the smart / cloud features anymore.
Newer units support local homekit control, which can also be paired with open systems like Home Assistant for full local control for automation.
Honestly this makes me feel that not adopting IOT is still a good idea. Yep I am probably leaving some efficiency on the table, but I get more reliability in exchange.
Spelling
Every one of my IT colleagues over about 35 is an absolute luddite. No IoT, no smart appliances, and a hardened firewall for everything that needs access. Location tracking and biometrics disabled on our phones, no cloud services, etc.
Shocker..it's not.
I was told I needed to sell ecobee and nest when I worked in HVAC. I refused because of this and sold everyone white Rodgers because they can't/won't just shut down the thermostat screwing the end user. I saw this bullshit from day one.
Yeah I have always felt the simple programmable thermostat is all I need. I do like the full week versions, but it seems like the 5-2 models are more readily available if you are just looking in the store.
Edit ~~Oh I just realized I forgot a not in my first post ~~ never mind I can't read
IOT can work without any cloud service. I have some things automated at home and everything works locally. To control it remotely I use a VPN.
Yeah I know that is a thing I might even be able to figure it out, but I am a bit of Luddite. By choice though, I have an okay understanding of tech but I don't see the advantage in many cases. I much prefer the reliability and simplicity of legacy tech. Also I am much more likely to be able to fix it myself if needed.
Before sears took a shit, I had to fix my mom's range. It was built some time in the 90s. The manual has a trouble shooting guide. I was able to call the sears help line and buy just the part I need and get it mailed to me. Everything was designed to be fixed and there was legacy company support.
Even with an IOT LAN. Repair of the hardware and often the firmware is not possible. You just have to buy a whole new expensive smart thing. I don't like that.
The article (which nobody here bothered to open) says they'll still function as "dumb" thermostats, so actually it's less of a big deal.
I will never understand the appeal for cloud based home automation.
All the spying aside, 16 years and your thermostat is no longer supported sounds ok until you think about how if you replace everything with cloud based equivalents you are always going to be having to replace things going out of service. I think its funny that you pay more for a device that lasts for lot less (How many thermostats have people seen fail?). I guess the companies are happy?
The appeal is remote and centralized management, easier programming and more features. If that’s not worth it to you to replace your thermostat every 16 years, then nobody is forcing you to get one.
But being able to change the temp from my phone from anywhere is worth it to me, as well as including it with other automations for all my connected devices. The appeal is honestly not hard to see, even if it’s not worth it for you personally.
AA much hate this might be getting, they're offering discounts on a new product, and 16 years is a hell of a lifetime. Imagine having to support software written in c99 maybe even c89, with some homebrew UI full of bugs.
I understand you may be from a field where supporting software from the 70s is required, however someone is probably paying big bucks for that software as well. Replacing the software you work on might cost millions, replacing a thermostat costs 300 usd.
I would love to live in a world where software support lasts 70 years. But consumers don't look at software support, so it's not budgeted in the price, and thus doesn't happen in the consumer space. Getting 16 years in a consumer device is long.
In the field you're working, stability, longevity, and robustenes is probably a requirement, not a nice to have.
I'm in my house right now with a perfectly working thermostat that's 70 years old.
And given the mechanism of action it will continue working in another 70 years.
16 years for hardware used inside of homes is a ridiculously, absurdly, short lifetime. Even for a vehicle that would be pushing the edge of "too short".
That said 16-year-old software is not that old. If it's built using sane language choices it should actually be functioning and modern today.
and 16 years is a hell of a lifetime
Think about it like this: Even if the average home nowadays had only about 10 such devices (I am quite sure the average home has a lot more), that are needed for kitchen appliances, heating, warm water, window shutters, solar panels, etc to function - that means on average about once a year one of the essential functions in the house stops working unless you replace a part. Not because it's broken, but because "SW support is discontinued". Seriously, I want to smash everyones faces for those "early adopters" who think smart homes are great, and of course the companies who put software in every little component.
That's why one uses an industry standard that is brand-independent,operates offline by design and does not require a central component besides a power supply.
Sounds like utopia?
This standard has been available since 1990 in its archaic form, since 2002 in its current form. It is downwards compatible and over 400 companies worldwide are part of the standard. HomeAssistant, ioBroker, openHAB,etc. all support it directly and there are multiple crossover gateways with other standards like DMX, ModBus, Dali,etc. exist. And no, it's components are not more expensive once you look at the TCO.
For fucks sake, people, use KNX.
(PS: There are even a few open-source/DIY components available)
Get a few components (e.g. Actors, a sensor), a gateway (USB is enough for the start,they often go cheap if you buy used,got mine for 10 bucks) ,a power supply (Meanwell is a good idea) and the free version of the ETS programming tool. (The ETS is the only downside of the system - it's expensive especially for larger installs)
For fucks sake, people, use KNX.
I thought you'd say OpenTherm!
Haha, no. But I know someone who was part of the OpenTherm development.
To quote him:
OpenTherm exists because Plumbers don't trust Sparkys and Sparkys don't trust plumbers.
OpenTherm is easy to install but "stupid" as hell and not adaptable to modern needs mostly, especially if you consider modern heating concepts like passively heated houses, heat pumps,etc.
That's different from KNX (or Modbus in that regard) - They are much broader in their appeal. Singule use/walled garden systems are always a bad idea,imho.
And to add another unpopular opinion:
A smart temperature control is the one I never ever need to use. Because then the room always has the temperature I want.
they made it online and dont want to bother actually supporting it...
so we replacing thermostats every decade and a half now?
This is why I'm all in for non-"smart crap", I don't even have inductive heating stove top because they never have basic knobs.
Long clicking on [3] then + + + + + + to boil your f eggs? No thanks.
That is a horrible argument.
Just get a non shit induction stove
Yeah I guess I'll consider it one day when they will be cheap enough, but now even the stupid-UI ones are fairly expensive and the knob ones quite expensive (thanks for all induction knob pists btw 😁).
Neff has a (magnetic/removable*) knob as well. (* Which is a great idea in theory. Unless you have kids. First it's great because they can't start the cooktop on their own. But then you are constantly looking for the knob. A friend nearly got insane...His daughter "accidentally" took the whole fucking thing on a schooltrip to France....HOW? That's why I have the Siemens one with touch.... It's okayish touchwise and it works...)
Wow so cheap too 😅
Long clicking on [3] then + + + + + + to boil your f eggs?
A lot of them have a terrible UI. But that's far from all of them. Enough have sliders. Sometimes one with a pan detection. Sometimes a slider per area.
It's just on and then hold b for boost now. It automatically detects which slot you have placed your pan on and selects that for you. But I get your point...
Induction is great 👍
I just bought one of these! First press on, then wait a second for the pan to be detected, then select the burner to turn on, then + which sets it to 5/10, then press + 5 more times for full power! The one with knobs was like double price...but hey the pan heats up quick while you are cursing at it.
If it detects the pan, why do you need to tell it which burner to use?
I just bought one of these!
There weren't any better options?
Did you try e.g. long pressing buttons, pressing -, or anything?
I had the +/- buttons ages ago (cheap alternative when renting). Never again.
Then one small drop of water (what is water doing in the kitchen??) touches the touch-surface and all goes beep beep beep while your pasta water slowly cools down...
I got the Frigidaire professional 36” induction and it does have knobs 😎
A link to the official notice: https://support.ecobee.com/s/articles/Connectivity-and-Support-for-Legacy-Products
(It was the first link in the article, good job The Verge)
That's why I have KNX.
Most thermostats would fail in that timeframe. Our original Nest thermostat failed this year because the connection that turns on the furnace wore out or became thin. Caused our furnace to click on and off repeatedly and ruined a relay on the furnace’s circuit board. Had to replace the thermostat and the furnace circuit board. Costly repair. Upgrade your thermostat before it wears out.
My thermostat is original to our 1941 house, and has survived a fire. My mom just sold her 1989 house with a digital programmable thermostat - also original to the house -that functions perfectly.
No part of a house should wear out and break after 16 years, except MAYBE carpeting. Building things like shit is no excuse for things being shitty, it's an indictment against it!
They don’t build em like they used to
Our smart thermostat has never been all that useful to me. The main thing is I don't have to walk over to it to change the temp. But that convenience isn't really worth the $150 I paid for it.
I get the idea and it's kinda cool, but not +$100 cool when I can just turn it down/off before I leave.
A lot of those smart thermostats do not support things like OpenTherm. As a result they often either start the heating or they do not heat. There's no modulation.
OpenTherm is from 1990s. Having a thermostat with a presence sensor is also not really anything new. Adding remote functionality would be nice if those "smart" wouldn't be so utterly terrible at being a thermostat. Meaning, the lack of modulation. And no modulation is pricy way to heat your home.
Yeah, but don’t give the power company control over the temperature in your house. That has gone poorly in Texas with people unable to turn on their AC on super hot days because they opted in to some program that they may not have understood.
The built in features of occupancy detection / geo fencing are cool though.