I was hoping that at least my Home Assistant integration would keep working. That sucks terribly. Lesson of the day: Avoid any IoT device that you can't use without an external service.

I've had both and am much happier with Ecobee.

Does Ecobee require an app and account to be activated, or is it something that HomeKit just detects on the LAN?

I was really disappointed with the lack of analytics / historical data.

Thankfully the open source beetstat makes ecobee a lot more useful, with full history and graphs for heat/cool runtime, aux heat, indoor/outdoor humdity, etc

Does operate locally retain any smart features or does it just essentially become a dumb thermostat?

depreciate by deprecation

Deprecate*

But yes, this is why I'm staying away from Gemini. Seems like an amazing product! But no way am I putting my AI eggs in the Google basket.

I wonder what year they’ll sunset Gmail. Nothing is forever, surely some PM in 2033 will decide that email is over and it’s time to push everyone to Gemini Messenger.

The day they finally deprecate Google Ads will be a good day.

Wave bye-bye!

https://www.youtube.com/watch?v=4Z4RKRLaSug

When I bought my first Nest in 2012, Nest Labs was not yet part of Google (that acquisition came in 2014.)

So for the owners of Gen 1 and 2 thermostats, your rule of thumb isn’t helpful like it is for post 2014 purchases.

That's what my Nest was saying as it flew out the 3rd floor window.

Worse than most thermostats that aren't made by Google though.

Which says something about how bad modern appliances are.

my car is from 1999 and I can still keep it working. I replaced a perfectly working furnace from 1973 a couple years ago (nearly double the efficiency). I still use my 1939 tractor and can get parts for it. I have plenty of other appliances older than that - many modern ones don't work as well

Most thermostats last for decades and decades. It is simple equipment.

You’ve got to use the one specific brand and type of battery they insist on. It’s some energizer lithium thing I’ll have to check for you and comment later. Anything else fails very fast, the correct ones last for years.

That parallel universe exists its just on aliexpress.com not amazon.com. Cheap hardware and open protocols as promised.

The key usecases for me are:

Adjusting the thermostat (which is downstairs) from bed.

At the airport - oh shit did I turn off the AC for the two weeks we'll be away? Ok I just did.

I used to have a Honeywell wi-fi thermostat. It looked like any other thermostat you've ever seen, except you could connect it to a home hub. It was nice because you could exactly what you're describing, but you could also do it in the app.

What made it worth it was being able to turn off the air or heat when you weren't home automatically. Now all or the "AI training" garbage? Yeah, forget that. I used to work in an office with a nest and it was torture if you showed up too early if stayed a little too late.

I really appreciate mine. My big use case is that we go away for a week or two over winter sometimes and turn the house down to 55F. The radiators take quite a while to heat the house back up from that temperature, so I turn the temperature back up remotely the morning before we fly back.

That said, I'm quite annoyed that Google is nuking my perfectly functional thermostat, and I will be buying an Ecobee to replace it, and integrating it with home assistant.

Everyone always says this stuff, but man these things are such garbage to use: terrible user interface, LCD screen with random blinking elements to tell you that's being edited, response rate slower than a ping to Mars. Modern app-connected thermostats are so much better.

I have the same thing and to be honest, if I had to replace a $200 thermostat every 2 years I would gladly do it. In fact, this whole thing has made me go and research which thermostat will fit where I live.

I like being able to adjust HVAC based on temperature sensors that are not co-located with the input keypad.

Being able to use the temp reading in a specific room is choice.

Well beyond a basic dumb thermostat I like that my EcoBee can use several wireless temperature sensors. During the day I have it set to only use the downstairs sensors and at night it only uses the upstairs sensors.

Can a $20 Honeywell thermostat do that with wireless sensors? If it can, I will get one.

The issue that the Nest solved for me was figuring out how long it took for the hot water to get to the radiators, and what the bounce looked like after it got there. It'll stop calling for heat before it reaches the final temp because it will still keep going up.

It doesn't need to be cloud-connected to do so, but that's not a feature I'm aware a $20 Honeywell has.

When I think of Google, killing products is one of the first things that come to mind.

Or just open up a little JSON server on the thermostat.

> Would be nice if google released a self hosted server

These mofos are too greedy to do this.

They are so shady about this stuff

I have a Honeywell t6 that I got when they installed a new unit - Honeywell INSISTS that you create an account and download the app to connect it to your home network

Thankfully this is bullshit and you can connect it directly from the thermostat to HomeKit - you will not find a single piece of documentation on this though and will be told it’s not possible

The real kicker is that there is a notification to register your device that you can’t get rid of unless you register your device

You can only snooze it for a couple weeks at a time

How I’d love to have one on one conversations with the evil people who approve this type of crap

Honeywell's z-wave thermostat basically does all of the shiny shit you'd want out of a connected thermostat while making basically impossible to lock the user out because Honeywell decided it didn't like the product anymore. Why people have to keep relearning this with IoT devices baffles me, and that Z-wave or maybe Zigbee isn't what's insisted upon.

EcoBee is happy to work without WiFi.

I bought a $20 Zigbee thermostat from AliExpress and it has been fantastic. It turns on when it's cold, and off when it's hot. Anything else, I can do with software, because it's just Zigbee.

My Insteon thermostat is a great dumb thermostat that I can also send commands to over a serial connection to a powerline/RF modem. (Very similar to Z-Wave's RF, though proprietary.)

The key is do not buy smart devices with Wi-Fi. There are better products for serious people. Everyone here with a Zigbee or Z-Wave product probably learned that the hard way first. ;)

If you're VC-funded then the valuation is the most important thing. The only way to juice your valuation is to get recurring revenue, because it comes with an 8x to 10x multiple. So you don't want to be in the hardware game, you want to use hardware to get a foothold in someone's home and then get them to pay you a subscription to maintain that hardware.

I think the valuation thing is what drives 90% of this stuff. Whereas an established company like Honeywell is more interested in building products and selling a lot of them, so they're going to charge you 5-10x of the cost of a Nest for the same feature set but with a local-first implementation instead of a cloud-first implementation.

I don't think I would ever buy a hardware product from a company billing themselves as a VC-backed startup.

Also, FWIW the Nest is a perfectly functional thermostat even if you never hook it up to their app. We found the scheduling and learning features to be really annoying so we turned them all off and never connected ours to the cloud.

Plenty of companies are successfull and don't rely on forced cloud. Reolink for example. Plenty of others.

The real difference is that these are not american sv vc backed companies like nest or ring. they are chinese companies set on disrupting those vc backed companies using this local first mindset as the differentiator.

When it's a device intended to be installed permanently in a house that could last 100 years, controlling a device that could last 20-30 years, and the company still exists (and is worth trillions to boot), I think 30 years of support would be a completely reasonable expectation.

I don’t know how these were advertised when originally sold, but I think products like these ought to clearly display a minimum sever lifetime. Something like “Nest connected thermostat, with 10 years of guaranteed cloud connectivity” displayed relatively prominently on the packaging. Otherwise, if the vendor can arbitrarily remotely yank access to a key feature, then the consumer is being sold a false bill of goods.

Yes. Unless Google is going out of business?

liftime of the building it is installed in. if I choose to replace it that is fine but there are still buildings in use 1000 years later, you need to escrow enough to support servers that long just in case.

>Serious question: how long do you expect cloud hardware vendors to support the servers for their products? 10 years? 20? 30?

Serious answer: I do not expect any specific lifetime at all (though legal return period is an obvious floor), but at a bare minimum I do expect (and think should absolutely be mandated by law) that power be intimately tied to responsibility. Ie., it's fine if a hardware vendor decides to retire their cloud services (or OS updates or the like) in 1 year or 10 years or 20 or 30, but IF they cease to support it, THEN they must also remove any technical obstacles to hardware owners pointing it at another service of any kind. So any signing keys required, code, docs/APIs etc. Decide that a given product no longer makes commercial sense for you to produce or support? Sure, fine, it happens. The problem is then ensuring the hardware/software dies with that.

The basic issue is that these places generally want to have their cake and eat it too. They want all the financial power of a monopoly tie-in and feudal rent extraction, but no responsibilities to go with it and the ability to force customers onto new stuff (or nothing). That should be illegal. Honestly, I think any tie-in should be illegal, fully accessible local APIs should be required and any 1st party subscription should earn its place on its merits.

But at a minimum, no one should be able to have it both ways. If they want power over their customers, they should have responsibility proportional to that. And conversely if they go full open source community friendly hackable from day 1 (and are fully upfront about that), I'm fine saying they have very minimal long term responsibility. There can and should be room for many different approaches to the market, but not extractive lock-in.

We’re retiring something like this at work. The devices were manufactured between 2009-2017. They will continue to operate in non-smart mode until some other part of the device breaks. We’ve notified customers and no one seems particularly upset. To a large degree it seems like the fleet’s obsolete and we could’ve pulled the plug a couple years ago. There’s probably not a good answer in a general sense. It really depends on a host of things.

you can buy these sensors off aliexpress cheap

This should be pretty much true for every "connected" device out there. They should all have a mode that works by directly connecting over the local LAN. Why do device manufacturers refuse to support this configuration?

If I want to change the volume of my "smart speaker" from my phone that's also on my LAN, it shouldn't require a round trip to a server on the Internet, or an account with credentials, or any of that nutty stuff.

I’m affected by this, and as pissed at Google about it as anyone, but the headline is overblown. The old Nest devices continue to function as thermostats, and the on-device features like scheduling still work. But I need the cloud-based features (particularly remote control via the app), so I went ahead and paid the upgrade tax.

This is exactly why I’ve started only buying smart devices that work with Home Assistant and don’t rely on cloud services.

> "Smart Cloud" home devices work as expected for about a year and then they fail in new and exciting ways, and then you replace them.

> "Smart Local Control" home devices work as expected until the electronics fail

ftfy.

While I agree with the message, I think honesty is important. The Nest gen 3 was released in 2015[1]. People got a 10-14 years out of these devices.

Also, that posts says the thermostat will still work locally so the failure state of the "smart" device here is that it became a "dumb" device after a decade+.

[1] - https://en.wikipedia.org/wiki/Google_Nest#Nest_Learning_Ther...

Well said.

Smarthome tech like this is just trying to make a quick buck at the expense of a lifelong relationship with a customer.

I shied away from wyze when they killed some features in some wifi cams I had years ago. It used to work like most others where it would detect movement and record a 30 second clip. Somewhere along the way that became a paid subscription only feature. Should have never updated that firmware...

I'm using reolink now for doorbell and will probably stick to them and other such brands going forward. poe of course too since all battery based cams suck in comparison (no live streaming capability or preroll recording before detection, needing regular charging). Wifi units are kind of crappy too compared to poe. Running cable is kind of satisfying in a strange way imo.

But then the customer might not buy a new nest....