Firstly "lose" is passive. Google is ending support.
But most importantly "support" sounds like they just won't be fixing bugs and security problems. What is really happening is that Google is disabling all online functionality.
"Google is shutting down all online functionality for 1st and 2nd gen Nest thermostats in Oct 2025" would be a much clearer headline.
> "lose support" Is such euphemistic phrasing. Firstly "lose" is passive. Google is ending support
Indeed, had they started to support a product they would never say “15th gen Nest thermostats found support”. You know, they were walking through the woods and stumbled on some Google support. Isn’t that a nice coincidence.
I would be happy if they worked locally. It is not often that I need to set the temperature remotely, when I am not in the house. But often I am laying in bed at night and don't want to get up and go downstairs to adjust the temperature. Why can't they work on the local network? That would take no resources from Google
> Why can't they work on the local network? That would take no resources from Google
I'm guessing since the generations being shut down are explicit, I'm assuming there are more generations that aren't being shut down, so the simple answer is that they want to sell more of those devices, basically force people to upgrade.
So indirectly, letting people still use the devices locally, would mean they'll sell less of the newer devices, which probably have better margins, slurp up more data and will eventually be shut down too.
- Encouraging folks to put older and unsupported Nest devices on a separate wireless network that doesn’t have internet access but access to a Home Assistant server for remote monitoring & control.
- Facilitate a way for Home Assistant based HVAC control setups to work with energy company savings programs where appropriate (in areas of the US this is fairly common, but they may only be supporting major brand vendors so far)
The solution is to stop installing advanced touchscreen computer systems where they don't really belong.
I've never seen a Honeywell thermostat fail, get hacked, or "lose support". I dont have to worry about what will happen if there is a weird software update or power outage. There are physical switches that configure the current mode of the system. Only the temperature setting is volatile, but it defaults to reasonable settings depending on the current mode (60F for heat, 80F for cool). Because of this simple design, there is no option to install a battery. There's nothing to remember.
The Nest Thermostats listed here are falling back to "Honeywell local thermostat" functionality.
They will still work for turning the dial locally to control the temp. They just won't be IoT connected.
The 3rd gen Nest thermostat was released in 2015. The 2nd gen and 1st gen were released in 2012 and 2011 respectively. A decade lifespan for a device that gives me metrics on my climate control and lets me tune it from my phone is pretty solid in my book.
> A decade lifespan for a device that gives me metrics on my climate control and lets me tune it from my phone is pretty solid in my book.
That's the attitude I take issue with. The device should work as it did at the time of purchase in perpetuity. If it can't, because of the architecture of the product (i.e. needs third party servers on the Internet) it shouldn't be marketed as being sold, but rather rented.
Acceptance that products we buy will just stop working (or, in the case of this product, losing significant functionality) is only encouraging companies to ratchet down lifetimes and hurt consumers more.
This feels like a distinction without a difference.
If all the shelves at Lowes that sell network-connected IoT thermostats said "10 year rental price" instead of "purchase price", the average non-nerd consumer wouldn't change their shopping behavior. The nerds... maybe they would? but they're the ones who already realize the impact of a device needing cloud connectivity to function.
The reality is that vendors cannot and will not offer "in perpetuity" commitments to run the cloud portion of their gear. For strategic reasons like "they want to deprecate legacy endpoints and functionality on their end" but also unintentional reasons like "the company went out of business".
Everything you buy has a lifespan. Physical objects degrade over time, objects with cloud connectivity rely on external functionality, etc. Picking between them is part of the decision making process: do you buy the cheap furniture that falls apart the first time you move? Or the more expensive furniture that lasts decades?
Should smoke detectors be listed as "rentals" because they degrade and have to be replaced?
> The reality is that vendors cannot and will not offer "in perpetuity" commitments to run the cloud portion of their gear.
Right. That is precisely why nothing that is meant to last should have any kind of dependency on "cloud". It must run locally only, so there is nothing that can be deprecated or dropped.
My mid 90s thermostat will continue to work independently of whether the original company that made it is out of business. Everything needs to be like that.
I buy a smoke alarm with a fixed 10 year life to avoid battery replacement. Others buy models with replaceable batteries, still others install wired alarms.
You should disclose the lifecycle at sale so consumers can make a decision. I’ve been in the computing infrastructure business for many years - I can model out the expected life and cost structure for all of my gear.
Deviating from your argument, but 10y battery life is a EU regulation. In some EU countries, you cannot rent out a living space without smoke alarm that has a 10y battery in it, even if it’s wired in. If you’re looking for a high quality smoke alarm (ability to trigger on different types of fires and certified for EU) I failed to find one that came with a replacement battery, and only a few that are wired in, all of which will have a 10y lasting battery.
The more expensive units do connect with internet/vendor and can do smart stuff like call fire department etc, you’re right that there’s a choice. But, the choice is very limited.
I’m not sure that an average consumer is even aware that some smoke alarm sensors degrade with time, or not trigger at all types of fires, there is probably some logic in forced unit replacement every decade.
That’s interesting. There is a certain logic to a regular maintenance cycle for high impact items.
It may be an example of an unintended consequence - to be fair to the Nest people, when you’re releasing a new product, you may not be thinking of end of life. Today we should be stamping a support date on them - Google is pretty good about this with Chromebooks, for example.
> Everything you buy has a lifespan. Physical objects degrade over time, objects with cloud connectivity rely on external functionality, etc. Picking between them is part of the decision making process...
Exactly. Consumers have no way to know, right now, if something they "buy" is truly a purchase or a rental. They can't make informed decisions without that knowledge. A technical person can tell the difference, but the average consumer cannot. That's why the distinction needs to be explicit.
this is where the FTC should step in and say "FOSS it" so at least consumers can have a choice and companies compete. Google could argue that it costs some nonzero amount (which is infinitesimal, I know), therefore they can't offer it perpetuity, but they could be required to unlock it so a 3rd party can compete and offer you the same service or let FOSS users control it.
A decade? Wow. If 1.5 years is a computer generation, the. These are more than 12 years or 8 generations behind. They should warn people about this. 12 car years = 8 computer years = 200 human years. I hope they still keep track of time.
The thermostat should be a remote display and a sensor. The real brains should happen on a controller near the heating system, that can be swapped in and out as time goes on.
As is the thermostat is a set of brains that can be swapped in and out as time goes on. I just swapped two thermostats in two homes a few weeks back and it took <20 minutes each with no prior knowledge, even though in both cases I had to swap the backplate.
What is the benefit you see of moving the swappable brains closer to the heating system? Doesn't that actually make the whole system more complex because now you have two separate devices where there is currently one?
The thermostat is the controller is most simple HVAC systems. The evaporator and air handler will have some "brains" re-compressor cycling and potentially fan speed, but the thermostat is calling for heat/cool.
You mean condenser instead of evaporator, the evaporator is inside the furnace and has no controls as it’s just a coil. The outdoor a/c unit contains the condenser and compressor and has the contactor (and controller if it’s an EC motor) for the compressor motor. The contactor control coil is 24v and the contact is closed when a call for cold is received from the thermostat.
Newer A/C units have EC motors with variable speed motor controllers and newer furnaces will have an EC motor blower fan and lots of factory control points wired up to a controller, mostly for the burner (and related gas/burner safeties).
Still, even in newer furnaces and a/c units, the thermostat is the main controller as like you said, the calls for heat/cool are the only input to the system.
Commercial HVAC controls for air handlers, boilers, and chillers have a lot more control points, an AHU can have 40 or more control inputs and outputs tied to it. Discharge and return air temp, multiple differential pressure sensors, static pressure, outdoor air temp, outdoor air pressure, duct avg temp, etc. These systems are much more complex than any home HVAC setup and usually have multiple controllers networked back to a master controller which orchestrates everything.
Did you read my last point about "a way for Home Assistant based HVAC control setups to work with energy company savings programs where appropriate" ?
Unless you're living off-grid, I think there's still some benefit to having one of the largest energy loads (building HVAC) coordinate a bit with utilities to ensure there aren't blackouts and such. Even better if there's an approach that works with open-source solutions like Home Assistant.
Where I live staring at the grid load and available energy is a thing during heatwaves and other inclement weather, and the utility generally issues "energy rush hour" limits to smart thermostats to help ensure grid reliability. The utility also incentivizes this I think by providing a small discount on the utility bill.
Why do they keep releasing these products if they don’t plan to support the users? Users are left frustrated, they are producing ewaste and ultimately hurting the Google brand.
Nothing will compare to ads in terms of margins. It makes no sense for Google to work on anything else, because eventually some VP will see compare the margins across business lines, and the decision is obvious. Chop everything that does not sell ads.
The 1st and 2nd gen Nest thermostats weren't released by Google, they were released by Nest prior to the acquisition. And they've been supported for over a decade. Their replacement (the 3rd gen nest) came out in 2015.
If what you want is a basic thermostat, there are any number of models you can buy. They're physical objects without network connectivity and they last basically forever.
But if you buy a Nest, you're buying a network-connected device. A decade replacement cycle for network-connected computing devices is not crazy.
> A decade replacement cycle for network-connected computing devices is not crazy.
Sure it is, and it’s why we desperately need consumer protection laws mandating lifetime security updates for IOT devices.
My grandparents had a 100 year old home - some of their light switches were well over 30 years old. Why should I be forced to replace a smart light switch before its useful life is over? Because some company wanted built in obsolescence so they could sell more product?
Edit: and to be clear, I’m potentially fine if the vendor takes the approach of open sourcing all firmware as a way to allow lifetime updates if they need to get out of the product for some reason and can’t maintain it themselves. “Reasonable accommodation”?
Hypotheticals are useful, but here we have facts we can discuss. A smart device released in 2010 was already using the REST api and HTTPs over WiFi. A brand new smart device will communicate exactly the same way.
So it is just Google’s decision to deprecate the device. And even if some of these protocols were deprecated, Google could just release a hub to keep the legacy devices online.
If this was the 90s or early 2000s I might agree more. But these days I really can't. I have network devices over a decade old, they still work great and are useful, and my computer itself that I still play modern games on is also over a decade old. Sure there is newer faster stuff, but I don't need or really have much of any use for that. A thermostat isn't gaining anything useful with faster networking or processing or any of that because what it already had was already way beyond any use case.
This is why I will most likely never own an EV, and most likely never own a car newer than 2015 or so (depending on model).
The entire world offerings are moving to this type of model under the part guise part reality of tech dependencies that just don't exist in a stand alone product.
Stand alone products are disappearing more and more every year that passes. It is certainly no accident.
> A decade replacement cycle for network-connected computing devices is not crazy.
The crazy part is that consumers are allowing the idiotic product lifecycles from the IT industry destroy products that, otherwise, had long and predictable lifecycles. It's shocking how the IT industry is successfully conditioning people to think it's normal to throw away otherwise working gear because "upgrades". The industry should be held to account before they take any more consumer goods hostage.
Local boring thermostats are still for sale. So are boring locks (even local-only electronic locks).
It feels like your actual beef here is that the majority of consumers are interested in buying cloud-connected gear that degrades when the vendor moves on, but you're framing their choice as "crazy".
> It feels like your actual beef here is that the majority of consumers are interested in buying cloud-connected gear that degrades when the vendor moves on, but you're framing their choice as "crazy".
My concern is that consumers don't know what they're buying. They don't know what "cloud connected" means. I don't care if people but this stuff, I care that they know what the bargain is that they're making.
> But if you buy a Nest, you're buying a network-connected device. A decade replacement cycle for network-connected computing devices is not crazy.
It should be, though. Okay, I get spare parts not being available after 10 years, particularly the "smart" chips simply don't get produced any more. But firmware and API support? How fucking hard can it be for a company raking in 90 fucking billion dollars in a single quarter of a year to keep a dozen people on payroll to keep the old servers, APIs and build tooling for the firmware up to date?!
Absolutely will be in business. The structural advantage and trajectory is tremendous. It is treated like an enterprise organization and ran very different from traditional Google business. Products are supported and customers have significant influence in the business. (Disclaimer: I work in the Cloud @ Google).
Nest have done a really good job of poisoning the well of consumer IoT. If a well funded startup acquired by a tech giant isn't going to do this properly then what chance does anyone else have? Absolutely disgraceful.
Fundamentally the idea end point devices should be connected to the Internet is wrong, and this means the idea the manufacturer should run any associated cloud service or be responsible for "security updates" is also wrong.
The only long term solution to this is that "smart devices" not be given direct access to the Internet, where the security problems are, and live instead in their own strictly segregated networks controlled by something conceptually like Home Assistant but where it also acts as a broker between all local devices, with a whitelist only policy for networking access. Until that becomes the default mode for deploying these things the consumer IoT space will continue to fail to meet expectations.
Few years ago I had a Nest thermostat and Google did an update. It bricked my thermostat. I had to go buy a newer model from Home Depot that was unplanned. I called their help desk and they said no way that the device was bricked but I was able to prove that an update was pushed out that likely was the culprit.
Then, years later there was the push to move people to the Google Home app from the Nest app. I still have both. But Nest was better before Google bought them. I had better customer service when I called them for help wiring my thermostat.
This week, Hydro-Québec, the nationalized company that provide electricity to residents of the province, has announced a major investment program to reduce electricity usage by using smart thermostats. I'm extremely worried about the life expectancy of those smart thermostats in the long run and whether it's a good use of public funds. I'm also not super thrilled of the amount of functional regular thermostats that will end up in landfills because of that initiative.
That doesn't mention the key detail of the Hydro Quebec smart thermostats: they keep remote control of the thermostat.
The whole idea is they can turn down your thermostat and sell the electricity to the US instead, or it was. Somehow they also encourage the adoption of electric cars, while also claiming we should save electricity. It's a very confused message.
In a similar vein, my 'smart' thermostat solution for my oil furnace has been to splice in a relay board that's connected to an esp32 running esphome through home assistant. On that board, and many others around the house, I have many 'virtual thermometers running again off esp32's using dht11's and esphome.
This configuration has allowed me to control the single-zone furnace using any or a combination of 'virtual' thermometers to turn the relay board on/off depending on conditions. I left the real old honeywell thermostat on the wall as a fail-safe situation in case the system comes crashing down, to prevent the house from freezing up in the winter -- Just set at a low minimum.
The downside of this is I either have to directly adjust the temp via phone interface / tablet interface / or google-assistant interface, but when that is done, it is all done locally. Simply toggling a simple relay board that just connects the 2-wire thermostat feed like a typical thermostat does. It has worked great!
In addition, one of the heat pump units (MrCool!) in the house was able to use a cloud-free dongle to bring it's controls into local-only control as well via home assistant, as well. This can also be controlled via phone / tablet / google home interface (I subscribe, I feel home assistant is worth supporting).
All of the local only devices are able to be accessed from afar using tailscale into the local network, with subnet routing.
When you have something like home assistant controlling it all, it's quite easy to set up situations like 'when the outside temp is less than 15(f)', turn the heat-pump off, and use the oil furnace instead, etc.
It seems the commercial control boards that do things of this nature cost a heck of a lot more, and effectively do very similar things.
It's a good time to start really considering a more local-based solution if you have the time to get the initial configurations all sorted!
Hey!
Reach out if you are looking for a
mecury/bi metalic tilt switch thermostat, door knob, or m3chan8cal light switch, if you get locked out of your cold ,dark, house.
"lose support" Is such euphemistic phrasing.
Firstly "lose" is passive. Google is ending support.
But most importantly "support" sounds like they just won't be fixing bugs and security problems. What is really happening is that Google is disabling all online functionality.
"Google is shutting down all online functionality for 1st and 2nd gen Nest thermostats in Oct 2025" would be a much clearer headline.
> "lose support" Is such euphemistic phrasing. Firstly "lose" is passive. Google is ending support
Indeed, had they started to support a product they would never say “15th gen Nest thermostats found support”. You know, they were walking through the woods and stumbled on some Google support. Isn’t that a nice coincidence.
I would be happy if they worked locally. It is not often that I need to set the temperature remotely, when I am not in the house. But often I am laying in bed at night and don't want to get up and go downstairs to adjust the temperature. Why can't they work on the local network? That would take no resources from Google
> Why can't they work on the local network? That would take no resources from Google
I'm guessing since the generations being shut down are explicit, I'm assuming there are more generations that aren't being shut down, so the simple answer is that they want to sell more of those devices, basically force people to upgrade.
So indirectly, letting people still use the devices locally, would mean they'll sell less of the newer devices, which probably have better margins, slurp up more data and will eventually be shut down too.
Ahh. Marketing and profit. Bless you Mr Crotchett.
See https://community.home-assistant.io/t/rooting-to-save-the-ne... , particularly the post asking about using the older unofficial APIs with Nest Thermostats on an airgapped network controlled by Home Assistant ( https://community.home-assistant.io/t/rooting-to-save-the-ne... ):
I'm wondering if an approach to saving older Nest thermostats and reducing ewaste might be:
- Using the older ‘unofficial API’ (from around 2014) assuming it still works, and integrate it with Home Assistant (see GitHub - gboudreau/nest-api: Unofficial Nest Learning Thermostat API ( https://github.com/gboudreau/nest-api ) & GitHub - wiredprairie/unofficial_nodejs_nest ( https://github.com/wiredprairie/unofficial_nodejs_nest#readm... ) )
- Encouraging folks to put older and unsupported Nest devices on a separate wireless network that doesn’t have internet access but access to a Home Assistant server for remote monitoring & control.
- Facilitate a way for Home Assistant based HVAC control setups to work with energy company savings programs where appropriate (in areas of the US this is fairly common, but they may only be supporting major brand vendors so far)
> reducing ewaste
The solution is to stop installing advanced touchscreen computer systems where they don't really belong.
I've never seen a Honeywell thermostat fail, get hacked, or "lose support". I dont have to worry about what will happen if there is a weird software update or power outage. There are physical switches that configure the current mode of the system. Only the temperature setting is volatile, but it defaults to reasonable settings depending on the current mode (60F for heat, 80F for cool). Because of this simple design, there is no option to install a battery. There's nothing to remember.
The Nest Thermostats listed here are falling back to "Honeywell local thermostat" functionality.
They will still work for turning the dial locally to control the temp. They just won't be IoT connected.
The 3rd gen Nest thermostat was released in 2015. The 2nd gen and 1st gen were released in 2012 and 2011 respectively. A decade lifespan for a device that gives me metrics on my climate control and lets me tune it from my phone is pretty solid in my book.
> A decade lifespan for a device that gives me metrics on my climate control and lets me tune it from my phone is pretty solid in my book.
That's the attitude I take issue with. The device should work as it did at the time of purchase in perpetuity. If it can't, because of the architecture of the product (i.e. needs third party servers on the Internet) it shouldn't be marketed as being sold, but rather rented.
Acceptance that products we buy will just stop working (or, in the case of this product, losing significant functionality) is only encouraging companies to ratchet down lifetimes and hurt consumers more.
It’s a thermostat. It should work as designed for decades.
This feels like a distinction without a difference.
If all the shelves at Lowes that sell network-connected IoT thermostats said "10 year rental price" instead of "purchase price", the average non-nerd consumer wouldn't change their shopping behavior. The nerds... maybe they would? but they're the ones who already realize the impact of a device needing cloud connectivity to function.
The reality is that vendors cannot and will not offer "in perpetuity" commitments to run the cloud portion of their gear. For strategic reasons like "they want to deprecate legacy endpoints and functionality on their end" but also unintentional reasons like "the company went out of business".
Everything you buy has a lifespan. Physical objects degrade over time, objects with cloud connectivity rely on external functionality, etc. Picking between them is part of the decision making process: do you buy the cheap furniture that falls apart the first time you move? Or the more expensive furniture that lasts decades?
Should smoke detectors be listed as "rentals" because they degrade and have to be replaced?
> The reality is that vendors cannot and will not offer "in perpetuity" commitments to run the cloud portion of their gear.
Right. That is precisely why nothing that is meant to last should have any kind of dependency on "cloud". It must run locally only, so there is nothing that can be deprecated or dropped.
My mid 90s thermostat will continue to work independently of whether the original company that made it is out of business. Everything needs to be like that.
People will often choose to not have a countdown.
I buy a smoke alarm with a fixed 10 year life to avoid battery replacement. Others buy models with replaceable batteries, still others install wired alarms.
You should disclose the lifecycle at sale so consumers can make a decision. I’ve been in the computing infrastructure business for many years - I can model out the expected life and cost structure for all of my gear.
Deviating from your argument, but 10y battery life is a EU regulation. In some EU countries, you cannot rent out a living space without smoke alarm that has a 10y battery in it, even if it’s wired in. If you’re looking for a high quality smoke alarm (ability to trigger on different types of fires and certified for EU) I failed to find one that came with a replacement battery, and only a few that are wired in, all of which will have a 10y lasting battery. The more expensive units do connect with internet/vendor and can do smart stuff like call fire department etc, you’re right that there’s a choice. But, the choice is very limited.
I’m not sure that an average consumer is even aware that some smoke alarm sensors degrade with time, or not trigger at all types of fires, there is probably some logic in forced unit replacement every decade.
That’s interesting. There is a certain logic to a regular maintenance cycle for high impact items.
It may be an example of an unintended consequence - to be fair to the Nest people, when you’re releasing a new product, you may not be thinking of end of life. Today we should be stamping a support date on them - Google is pretty good about this with Chromebooks, for example.
The forced upgrade assumes that sensor technology will improve between purchases.
> Everything you buy has a lifespan. Physical objects degrade over time, objects with cloud connectivity rely on external functionality, etc. Picking between them is part of the decision making process...
Exactly. Consumers have no way to know, right now, if something they "buy" is truly a purchase or a rental. They can't make informed decisions without that knowledge. A technical person can tell the difference, but the average consumer cannot. That's why the distinction needs to be explicit.
this is where the FTC should step in and say "FOSS it" so at least consumers can have a choice and companies compete. Google could argue that it costs some nonzero amount (which is infinitesimal, I know), therefore they can't offer it perpetuity, but they could be required to unlock it so a 3rd party can compete and offer you the same service or let FOSS users control it.
A decade? Wow. If 1.5 years is a computer generation, the. These are more than 12 years or 8 generations behind. They should warn people about this. 12 car years = 8 computer years = 200 human years. I hope they still keep track of time.
Or make the device as SIMPLE as possible.
The thermostat should be a remote display and a sensor. The real brains should happen on a controller near the heating system, that can be swapped in and out as time goes on.
As is the thermostat is a set of brains that can be swapped in and out as time goes on. I just swapped two thermostats in two homes a few weeks back and it took <20 minutes each with no prior knowledge, even though in both cases I had to swap the backplate.
What is the benefit you see of moving the swappable brains closer to the heating system? Doesn't that actually make the whole system more complex because now you have two separate devices where there is currently one?
The controller for the HVAC system costs a heck of a lot more than the thermostat.
The thermostat is the controller is most simple HVAC systems. The evaporator and air handler will have some "brains" re-compressor cycling and potentially fan speed, but the thermostat is calling for heat/cool.
You mean condenser instead of evaporator, the evaporator is inside the furnace and has no controls as it’s just a coil. The outdoor a/c unit contains the condenser and compressor and has the contactor (and controller if it’s an EC motor) for the compressor motor. The contactor control coil is 24v and the contact is closed when a call for cold is received from the thermostat.
Newer A/C units have EC motors with variable speed motor controllers and newer furnaces will have an EC motor blower fan and lots of factory control points wired up to a controller, mostly for the burner (and related gas/burner safeties).
Still, even in newer furnaces and a/c units, the thermostat is the main controller as like you said, the calls for heat/cool are the only input to the system.
Commercial HVAC controls for air handlers, boilers, and chillers have a lot more control points, an AHU can have 40 or more control inputs and outputs tied to it. Discharge and return air temp, multiple differential pressure sensors, static pressure, outdoor air temp, outdoor air pressure, duct avg temp, etc. These systems are much more complex than any home HVAC setup and usually have multiple controllers networked back to a master controller which orchestrates everything.
Yeah. Wrong side of the cycle. Good call.
Did you read my last point about "a way for Home Assistant based HVAC control setups to work with energy company savings programs where appropriate" ?
Unless you're living off-grid, I think there's still some benefit to having one of the largest energy loads (building HVAC) coordinate a bit with utilities to ensure there aren't blackouts and such. Even better if there's an approach that works with open-source solutions like Home Assistant.
Where I live staring at the grid load and available energy is a thing during heatwaves and other inclement weather, and the utility generally issues "energy rush hour" limits to smart thermostats to help ensure grid reliability. The utility also incentivizes this I think by providing a small discount on the utility bill.
Why do they keep releasing these products if they don’t plan to support the users? Users are left frustrated, they are producing ewaste and ultimately hurting the Google brand.
Nothing will compare to ads in terms of margins. It makes no sense for Google to work on anything else, because eventually some VP will see compare the margins across business lines, and the decision is obvious. Chop everything that does not sell ads.
The 1st and 2nd gen Nest thermostats weren't released by Google, they were released by Nest prior to the acquisition. And they've been supported for over a decade. Their replacement (the 3rd gen nest) came out in 2015.
Thermostats have an operational lifetime far more than a decade.
If what you want is a basic thermostat, there are any number of models you can buy. They're physical objects without network connectivity and they last basically forever.
But if you buy a Nest, you're buying a network-connected device. A decade replacement cycle for network-connected computing devices is not crazy.
> A decade replacement cycle for network-connected computing devices is not crazy.
Sure it is, and it’s why we desperately need consumer protection laws mandating lifetime security updates for IOT devices.
My grandparents had a 100 year old home - some of their light switches were well over 30 years old. Why should I be forced to replace a smart light switch before its useful life is over? Because some company wanted built in obsolescence so they could sell more product?
Edit: and to be clear, I’m potentially fine if the vendor takes the approach of open sourcing all firmware as a way to allow lifetime updates if they need to get out of the product for some reason and can’t maintain it themselves. “Reasonable accommodation”?
> A decade replacement cycle for network-connected computing devices is not crazy
What's your reasoning behind that belief? A device being "modern" doesn't automatically mean it has to be user hostile. That is a choice.
Hypotheticals are useful, but here we have facts we can discuss. A smart device released in 2010 was already using the REST api and HTTPs over WiFi. A brand new smart device will communicate exactly the same way. So it is just Google’s decision to deprecate the device. And even if some of these protocols were deprecated, Google could just release a hub to keep the legacy devices online.
If this was the 90s or early 2000s I might agree more. But these days I really can't. I have network devices over a decade old, they still work great and are useful, and my computer itself that I still play modern games on is also over a decade old. Sure there is newer faster stuff, but I don't need or really have much of any use for that. A thermostat isn't gaining anything useful with faster networking or processing or any of that because what it already had was already way beyond any use case.
After a year, CarOS™ 25 has reached end of life and will not be updated.
To maintain security and guarantee continuing safety, CarOS™ will now drive you to your local showroom where you can purchase updated hardware.
This is why I will most likely never own an EV, and most likely never own a car newer than 2015 or so (depending on model).
The entire world offerings are moving to this type of model under the part guise part reality of tech dependencies that just don't exist in a stand alone product.
Stand alone products are disappearing more and more every year that passes. It is certainly no accident.
> A decade replacement cycle for network-connected computing devices is not crazy.
The crazy part is that consumers are allowing the idiotic product lifecycles from the IT industry destroy products that, otherwise, had long and predictable lifecycles. It's shocking how the IT industry is successfully conditioning people to think it's normal to throw away otherwise working gear because "upgrades". The industry should be held to account before they take any more consumer goods hostage.
Local boring thermostats are still for sale. So are boring locks (even local-only electronic locks).
It feels like your actual beef here is that the majority of consumers are interested in buying cloud-connected gear that degrades when the vendor moves on, but you're framing their choice as "crazy".
> It feels like your actual beef here is that the majority of consumers are interested in buying cloud-connected gear that degrades when the vendor moves on, but you're framing their choice as "crazy".
My concern is that consumers don't know what they're buying. They don't know what "cloud connected" means. I don't care if people but this stuff, I care that they know what the bargain is that they're making.
How old is your cell phone?
How old is your toaster?
> But if you buy a Nest, you're buying a network-connected device. A decade replacement cycle for network-connected computing devices is not crazy.
It should be, though. Okay, I get spare parts not being available after 10 years, particularly the "smart" chips simply don't get produced any more. But firmware and API support? How fucking hard can it be for a company raking in 90 fucking billion dollars in a single quarter of a year to keep a dozen people on payroll to keep the old servers, APIs and build tooling for the firmware up to date?!
[1] https://finance.yahoo.com/news/google-stock-rises-after-it-b...
To add that Nest team probably isn’t doing a whole lot. It’s mostly bug fixes it says which is probably just open source dependency management.
My thermostat was installed in 1940. Hopefully Honeywell doesn’t come for it lol.
Margins for cloud are not like ads, but yeah, it exists and is a huge part of Google now :)
I can bet AWS and Azure will be in business in 10 years. Wanna bet on whether GC will be in business (and not sold to pieces by Google) in 10 years?
Absolutely will be in business. The structural advantage and trajectory is tremendous. It is treated like an enterprise organization and ran very different from traditional Google business. Products are supported and customers have significant influence in the business. (Disclaimer: I work in the Cloud @ Google).
> Why do they keep releasing these products if they don’t plan to support the users?
Because the money is made by selling the products, not by supporting the users.
Nest have done a really good job of poisoning the well of consumer IoT. If a well funded startup acquired by a tech giant isn't going to do this properly then what chance does anyone else have? Absolutely disgraceful.
Fundamentally the idea end point devices should be connected to the Internet is wrong, and this means the idea the manufacturer should run any associated cloud service or be responsible for "security updates" is also wrong.
The only long term solution to this is that "smart devices" not be given direct access to the Internet, where the security problems are, and live instead in their own strictly segregated networks controlled by something conceptually like Home Assistant but where it also acts as a broker between all local devices, with a whitelist only policy for networking access. Until that becomes the default mode for deploying these things the consumer IoT space will continue to fail to meet expectations.
This model fundamentally makes it harder for devices to upload data to the cloud, which is the primary business model.
Few years ago I had a Nest thermostat and Google did an update. It bricked my thermostat. I had to go buy a newer model from Home Depot that was unplanned. I called their help desk and they said no way that the device was bricked but I was able to prove that an update was pushed out that likely was the culprit.
Then, years later there was the push to move people to the Google Home app from the Nest app. I still have both. But Nest was better before Google bought them. I had better customer service when I called them for help wiring my thermostat.
This week, Hydro-Québec, the nationalized company that provide electricity to residents of the province, has announced a major investment program to reduce electricity usage by using smart thermostats. I'm extremely worried about the life expectancy of those smart thermostats in the long run and whether it's a good use of public funds. I'm also not super thrilled of the amount of functional regular thermostats that will end up in landfills because of that initiative.
[1]https://news.hydroquebec.com/en/press-releases/2172/hydro-qu...
That doesn't mention the key detail of the Hydro Quebec smart thermostats: they keep remote control of the thermostat.
The whole idea is they can turn down your thermostat and sell the electricity to the US instead, or it was. Somehow they also encourage the adoption of electric cars, while also claiming we should save electricity. It's a very confused message.
My kingdom for a thermostat that doesn’t expire after a decade.
I’ve heard great things about Ecobee and they can be controlled locally via home assistant.
In a similar vein, my 'smart' thermostat solution for my oil furnace has been to splice in a relay board that's connected to an esp32 running esphome through home assistant. On that board, and many others around the house, I have many 'virtual thermometers running again off esp32's using dht11's and esphome.
This configuration has allowed me to control the single-zone furnace using any or a combination of 'virtual' thermometers to turn the relay board on/off depending on conditions. I left the real old honeywell thermostat on the wall as a fail-safe situation in case the system comes crashing down, to prevent the house from freezing up in the winter -- Just set at a low minimum.
The downside of this is I either have to directly adjust the temp via phone interface / tablet interface / or google-assistant interface, but when that is done, it is all done locally. Simply toggling a simple relay board that just connects the 2-wire thermostat feed like a typical thermostat does. It has worked great!
In addition, one of the heat pump units (MrCool!) in the house was able to use a cloud-free dongle to bring it's controls into local-only control as well via home assistant, as well. This can also be controlled via phone / tablet / google home interface (I subscribe, I feel home assistant is worth supporting).
All of the local only devices are able to be accessed from afar using tailscale into the local network, with subnet routing.
When you have something like home assistant controlling it all, it's quite easy to set up situations like 'when the outside temp is less than 15(f)', turn the heat-pump off, and use the oil furnace instead, etc.
It seems the commercial control boards that do things of this nature cost a heck of a lot more, and effectively do very similar things.
It's a good time to start really considering a more local-based solution if you have the time to get the initial configurations all sorted!
Check out SAT Smart Autotune Thermostat https://github.com/Alexwijn/SAT
Its heating curve is corrected by a PID. Room temperature precision is within 0.1 Celsius.
Less gas consumption, lower bills and higher comfort :)
Hmm, I think I'll stick with my traditional, non-smart thermostat. It's been... supporting my house for the nearly 40 years now without any issue.
~14 years, which feels like a long time, but really no time at all in terms of equipment.
Looks like they will look great in garbage cans. You would think that people would not fall for this, but they do.
Wait. Oct? They run Windows 10?
Hey! Reach out if you are looking for a mecury/bi metalic tilt switch thermostat, door knob, or m3chan8cal light switch, if you get locked out of your cold ,dark, house.