Belkin ending support for older Wemo products | Hacker News Reader

Belkin ending support for older Wemo products

belkin.com

93 pointsapparent75 comments

Wow, this affects all but 4 Wemo products. Crazy! This is a great way to ensure that no one trusts the Belkin/Wemo brands anymore.

The most reliable Wemo devices I have are the older ones. The oldest, which I got off Craigslist, is rock solid. The newer outlet plugs are super flakey.

I was already unlikely to buy new Wemo products, given their trajectory. But now that they're abandoning their stable products, I'm definitely never buying Wemo again.

Anyone have suggestions for good wall plugs and light switches?

Look at the original Honeywell Z-Wave products, still going strong and still doesn't need an internet connection to work.

They will work with any Z-Wave/Zigbee compatible controller.

I'll stick with my X10 system...

Pick something you like, then look up its Home Assistant integration. If it says "Its IoT class is Local Push.", then you're all good. Lutron Caseta is my gold standard. After that, ZWave or Zigbee.

EDIT: From what I can tell, the Belkin integration _is_ local push, so I don't know what the big deal is. Does removing "support" just mean no more updates? If they aren't already, hook those guys up to Home Assistant and you can keep them until you die.

I think it means you won't be able to control them using the WeMo app or use any features via WeMo cloud services. It does say:

"Wemo products configured for use with Apple HomeKit prior to January 31, 2026 will continue to function via HomeKit in the absence of Wemo cloud services and the Wemo app."

So I'd assume that if you can use something without WeMo services or apps, you'll continue to be able to.

billfor says below that they work when internet is unavailable, so maybe they'll continue to work that way?

That’s how my Best Buy switches (that they gave me gift cards equal to the purchase price of) work - HomeKit continues to work fine, but I’m not sure I’d be able to reconfigure it.

I’d assume so.

That statement does imply products can no longer be configured for use with HomeKit after the servers shut down. Hope you don't need to reset any of your devices!

I did wonder about that. Do you have to use WeMo services as part of configuring for Homekit so that basically any device not set up already is useless?

I was weary but took the leap and installed Lutron Caseta switches throughout my house. They are absolutely fantastic.

I have had Caseta in mine for over 5 years.. only one failed out of like 16.

This is honestly very interesting. I bought a condo that was pre-equipped with Casetas and due to the crappy app I thought “these are budget and terrible” but actually reflecting now they all work tremendously and I haven’t had a problem with any. Goes to show how much an ignorant person (me) can’t recognize function over form

Unfortunately “local push/access” is not a guarantee. When Awair decided to end support for their older devices they decided to also shutdown the local device API.

Oh, odd. I have one of the first Awair devices and its local API is still going strong. Maybe because I haven’t accessed it through their app in about 5 years? They must have lost track of it. I’ll block its internet access entirely to make sure they can’t break it. The beauty of a local API. :D

My Meross plugs have been very stable. Have not had to reset them, reconfigure them, rehome them, re-pair them, etc.

For light switches I tell everyone I know it is either LUTRON or nothing. Your time is not worth messing with anything else, and Lutron has been in the game long enough they have demonstrated commitment to their products. Finally, my Lutron system has been one of the few IoT devices I have ever had that was truly "set and forget". Once the Lutron switches and shades were bound, they have been perfect ever since for me.

Company culture? lutron is used to selling a 20 year lifespan home switch. belkin expects their devices to stay around ~5 years.

From the way Belkin treats their routers, they don't seem to expect their devices to make it to the end of the month they're brought out in. I've received better software support from noname aliexpress devices than I've received from Belkin.

According to Amazon my first Lutron switch was delivered on August 18th, 2018 and after testing it (and the bridge) out for a couple of months I switched every single switch in my house to Lutron.

I currently have 20 Lutron devices, including switches, dimmers, lamp dimmers, and pico remotes.

Between August 18th, 2018 and today the number of times my Lutron devices have failed to operate as intended can be counted on zero fingers.

I wish their product line was more expansive but if focus is what it takes to ensure the level of reliability they have, then ok.

I also have the shades, I love the entire Lutron system.

Sonoff S31 - cheap and have been rock solid for me. I use them with Home Assistant.

> We understand this change may disrupt your routines, and we sincerely apologize for the inconvenience.

Empty apology, for an act that legislators, regulators, and the market shouldn't tolerate.

> If your Wemo product is still under warranty on or after January 31, 2026, you may be eligible for a partial refund. Refund requests will not be processed before that date.

Someone needs to lead a class-action suit, with a settlement that is many multiples of whatever the company thought they gained/saved. Get it publicized for well-deserved brand damage. Make it hurt enough that shareholders are angry at the CEO and board.

I agree this sucks, but what would the cause of action be? They're offering partial refunds for items purchased within the warranty period (though you have to have proof of purchase, which of course practically no one will have).

I don't think companies can be sued for discontinuing cloud services that are required for older devices to keep working, otherwise Google would have surely been sued over their Nest security system shutdown, among other things. Maybe a shareholder derivative lawsuit could make sense, but Belkin isn't publicly traded, and presumably whoever owns them is OK with this decision (and wouldn't want a costly lawsuit).

The fact that the refund is only partial seems pretty unreasonable to me. If they are reneging on their side of the deal then they should return the money that they took.

Yeah, especially for devices that have to be installed by an electrician, which can double the cost. I think I paid around $50 for dimmable switches and $50 for each to be installed. Fortunately most of them work with HomeKit, but one of those suckers is only controllable via the Wemo app...

When a product like this just dies overnight, there should be a law stating that the source code needs to be made public. No ifs ands or buts. If you don't want to release the source code, then don't release the product or keep a modicum of support available. To brick products is unconscionable.

I think there needs to be a mechanism in place. When you release a product, you should be required to put the source code in escrow, so it can be released at EoL.

Good luck getting it fron the country that ruled national injuctions are onerous but the ftc can be blocked from requiring click to cancel. You're going to have your devices bricked and you're going to have to live with it.

It's wild to see how for no reason other than marketing, we're seeing devices with a static feature set being bricked. I remember older Nest thermostats being similarly crippled or bricked.

By "static feature set" I mean to draw a contrast between these plugs and light switches, and say, a device that has to have a web browser or which needs to access an external API that may need to change to reflect changing external factors. Literally nothing needs 'updating' about a simple relay. It turns on and it turns off. Same now as in 2010.

Smart home tech really should be fully cost-free to keep working indefinitely. All these "cloud" and "mobile app" integrations that 100% of them have are what makes these EOLs happen, since the company needs to maintain servers speaking a certain API version, or push firmware updates out for every device ever made, and keep updating a mobile "app" just to keep it running.

The solution to this has been with us for over a decade: the Zigbee and Z-wave model. The devices speak a standard interface and talk to the owner's choice of standard hub forever. (If Sonoff goes out of business tomorrow my Sonoff devices will all still work perfectly fine.)

Belkin could have shipped Zigbee-compatible devices even when their first WeMo device came out, but they thought it was more profitable to make proprietary stuff.

In the email they mention that they're focusing on other parts of the company. My guess is that their lousy quality for recent devices (IME) is causing lots of returns and lots of customer service demands. That costs them money. If the new devices worked better, they wouldn't have these ancillary costs, which are probably the true reason for shutting this all down.

Sounds completely believable. Belkin has undergone so many changes. For anyone else who didn't remember, I just looked it up. They're a subsidiary of Foxconn now, as is Linksys, who Belkin bought from Cisco. Now both are sister companies under the Foxconn umbrella.

> It's wild to see how for no reason other than marketing, we're seeing devices with a static feature set being bricked

I work in the smart home space. Your sentiment is reasonable and understandable but, respectfully, you’re wrong.

First, all these devices will continue to work locally. They’re just killing the server, which has continual real costs. They’re not being bricked, they’re just regressing to their static features.

Second, ZWave and Zigbee were never the future. It’s very hard to convince consumers they need a hub. My firms market research indicates that most consumers think it’s a cash grab. And the return rate of Zigbee products because they “don’t connect to WiFi” is scary high.

Finally, you need an app and cloud of some sort for controlling the device. Consumers expect that app will control the device and it will work anywhere in the world, because that’s why they bought it. That’s just the basic market research on the majority of consumers of smart home products. Users need a way to control the device (or at least set it up), and that means an app, which means continuous updates as app stores change. Global control means cloud relays and server costs.

Matter (the protocol) was inevitable. Consumers already have an IP network, and the big tech companies can afford to build and distribute apps and relay servers built into their phones and cloud offerings. This minimizes the risks of device manufacturers and maximizes compatibility.

> First, all these devices will continue to work locally. They’re just killing the server, which has continual real costs. They’re not being bricked, they’re just regressing to their static features.

What do you mean by "static feature"? Like that I can still press a button on my wall and it will toggle the status of my lights? Or that I can use my Wemo app when I am at home (but not elsewhere) to control my lights?

I have noticed that even when my wifi goes out, my programmed Wemos will still turn on X minutes before sunset and off X minutes after sunrise. Do you think this sort of operation requires the server (either always, or every once and a while) to be available?

I am trying to figure out if I should get new devices on Black Friday or if it's likely enough these will still work fine for my purposes come next Jan.

> market research indicates that most consumers think it’s a cash grab

This I 100% believe and it's so sad. I'm clearly suffering from the fact that smart home enthusiasts, while being the only ones who can even build an even remotely useful smarthome setup, are a minority, and by the numbers, probably 10x the devices are purchased by very nontechnical users who say things like "I just paid the wi-fi bill" and who would buy and return devices that need hubs. And yeah, we all need those people to make a product 'mainstream successful.'

Among enthusiasts of course, we all know a hub basically gives you (the user) control, rather than being reliant on some cloud server that will be turned off (like this article illustrates). (And they're usually like $30 so I'm astounded at how cheap people are. I've had Hue and Lutron hubs for literally 8 years and I suspect they'll last 10 more.) The problem of NAT which I didn't remember to talk about, created a lot of this, as obviously some kind of cloud thing needs to be involved for out-of-home control, but yeah, I think you and I agree that we're better off letting Apple and Google and Amazon be the cloud relays for things that connect to some device in our home that can serve as a hub.

The thing that's the most irritating is that without standards and hubs, you either have to go all in on one company's ecosystem, and none of them are the best at everything, or you have to use 12 different apps: this one for lights, this one for smoke alarms, this one for audio, this one for cameras, this one for smart speakers, this one for that one switch that you bought because they're the only one who offered a really specific type, etc.

14 years is not long for things built into a home...

Imho there should be some kind of electrical standard of how these things are controlled (specified in voltages, relays, ... and not changing over the years nor between companies), and then have something that can be interchangeable that can be upgraded "with the times" to control all this.

Some of these items were sold much more recently than that! Also, I paid a few hundred bucks to have the light switches installed in my home, so the wasted cost is perhaps double what I paid Belkin for the switches.

Imho you're kinda describing how Zigbee and Z-wave work. The devices have various profiles like "Light bulb" or "button" or "temp sensor" and you can pair them with a Hub device. You could replace the hub with a new one, for instance, if say, Ethernet or TCP/IP became outdated. And new profiles are invented over time to allow new classes of devices.

Matter is a new standard that is incredibly slowly rolling out, the problem with it is that it isn't the 'full picture' because it only specifies a protocol but not the physical medium. So you can have matter devices over TCP/IP that have to be connected to Wi-Fi, or matter devices over Thread radio (more akin to Zigbee). Wi-Fi will come with its own set of headaches since you have to have some dumb app or whatever to use, to get the correct credentials onto each individual IoT device.

Thread is an IPv6 network stack. So if you're using Matter over Thread, you're using TCP/IP (or more specifically, UDP over IP).

TCP/IP and its associated network and transport protocols are not going to be outdated in our lifetimes.

Thanks for the correction. My main point was that Matter over a normal Ethernet network is imho a different animal than Matter over Thread.

Not that I've yet found any decent devices that don't cost a fortune that do anything with Thread, other than apparently a half dozen things I own are Thread Border Routers (hubs). I have all these hubs and nothing for them to speak to.

Matter over ethernet, as designed, should work exactly like matter over wifi and matter over thread. The only exception is the Bluetooth LE pairing component, the rest is a mix of IPv6, mDNS, and other pretty much standard protocols for networking and discovery.

Matter and Thread seem to be progressing rather slowly when it comes to hardware support. Zigbee seems to be the reigning champion still, though slowly companies seem to be moving towards Matter now.

The super cheap AliExpress stuff will probably take years to move on. Matter is a western standard backed by western big tech, so unless you're mostly targeting the west you may as well stick to Zigbee for a while.

The biggest advantage to Thread over Zigbee is probably that phones have Thread radios (iPhones and some Androids) so you can directly control light bulbs and such without any gateways or firewalls to care about. Now that esp32s are also gaining Thread support, things should start accelerating in a few product cycles.

> Matter over ethernet, as designed, should work exactly like matter over wifi and matter over thread.

Sorry, I've again failed to explain properly.

My main point: I dislike smarthome things where the individual endpoints are wi-fi, especially small things like switches and lights, because it means there has to be some usually cheesy way to pair it to the network (and also because I don't have Ubiquiti or anything so I don't have a separate IoT network, so it would clutter up my LAN and annoy me). I like Zigbee (and I suppose Zwave is cool, but don't personally have it) and will probably like Thread someday because they don't require me to get my network credentials into a lightbulb in the first place, which if I know manufacturers, will in practice mean some horrible iOS app is mandatory and that app's eventual retirement will someday brick the device.

For that reason, I am rooting for Thread to hurry up and actually make an impact on the market, and not really interested in Matter-over-Wi-fi because in that sense, moving from Zigbee (or Caseta, my other major thing) to Matter-over-Wi-fi would be a big step backward in that way.

I've installed a lot of Wemo (old) devices in my house and they worked well. Tried to buy some new ones (simple socket switch) and they pretty much didn't work. Extremely frustrating. I spent a lot of time futzing with it and swore a lot about closed-systems.

Yesterday was getting tempted by Zigbee again and then this arrived at my email. This is the last straw and decided to go full steam ahead [0]. btw the dongle is 25% off on amazon atm (prime day).

[0] It's this one, not tried it yet (fingers crossed it will work): https://www.amazon.com/dp/B09KXTCMSC

I also had lots of issues with the newer Wemo products, even when using them without the Wemo app (over HomeKit). I spent tons of time on the phone with their support team, manually resetting the devices time and time again. They work maybe half the time.

I use that dongle with Zigbee2MQTT, everything works perfectly and you can pry this setup from my cold, dead hands.

The Home Assistant-branded ZBT-1 [0] is $30ish and, if your goal is Home Assistant, you're probably better off going that route especially now that IKEA is going all in on Thread/Matter (which the ZBT-1 supports in addition to Zigbee.)

I do have that particular Sonoff dongle - as well as a ZBT-1 - and I can vouch that the dongle works as well as the Hue hub it replaced, controlling a couple dozen lights.

(Obligatory prime-day-is-not-much-of-an-actual-discount so don't worry if you miss out on this sale - https://camelcamelcamel.com/product/B09KXTCMSC)

[0] https://www.home-assistant.io/connectzbt1/

The docs for ZBT-1 seems to indicate that it can only do one of ZigBee or Thread at a time (that is, you must choose one of the two firmwares to flash), right?

It looks like there's third-party work to make Sonoff ZBDongle-E (which is _not_ the linked device) do both at once, but that's not officially supported…

I use this exact Sonoff dongle! I use it with ZHA and I've been very happy with it. I have it paired with a Hue bulb, several cheapo Aliexpress scene switches, and (my recommended switch) a MOES ZigBee scene switch.

Only tip I have for you is that everyone suggests to use a USB extension cable with these to get it away from your PC or whatever device. My dongle is sticky-tacked to my office wall.

A warning on Zigbee - Ikea just announced that they're moving their smarthome products to Matter over Thread. It's a win for interoperability, but they've been one of the main reliable and affordable vendors for Zigbee buttons/remotes, door sensors, smart outlets, lightbulbs (way cheaper than Hue), etc. There will still be others, but the ecosystem may not be the best long term investment anymore.

Stop killing ~games~ [1] functioning products with forced proprietary server connection

[1] https://news.ycombinator.com/item?id=44478083

I hate cloud-required hardware and software because it always dies at the whims of the manufacturer.

Pre Android, Motorola Mobility's phones had a huge fleet of CentOS boxes that were required to run a large fraction of the social app functionality on their phones.

The only games/apps/hardware I want to use are those that can be run completely offline with no dial-home DRM or server backend.

If they do in fact continue supporting HomeKit, this isn’t the worst way to EOL the product. I’d love to see them all updated to support matter before being put out to pasture, though I’m sure Belkin feels that’s difficult or impossible for some reason.

> continue supporting HomeKit

As far as I understand, Apple insisted that HomeKit devices be manageable locally with cloud connection optional. Your device can still do cloud connection, but Apple devices will only talk to it locally though a HomePod or an AppleTV or some iDevice that you leave on your home network.

So it's not really how they are choosing to continue to support these products, and more of a byproduct of how HomeKit works.

Makes sense. The protocol is certainly local only. And I guess once added, there’s no need to rip HomeKit support out. My iHome devices have been similarly abandoned but still work just fine.

I guess I’m conditioned to expect this stuff to die one way or another when something like this happens. I shall cautiously reframe my expectations…

Only caveat there is setup. IIRC a lot of devices have to be added to HomeKit via their own apps. Would be frustrating to never be able to move your device to a new “home”.

Save the pairing keys you get for those devices. You shouldn't need to re-request them. They are unique per device and don't change (unless if you flash the device firmware or its EEPROM or they fuck it up on an update for some reason, but they shouldn't).

Generally the only reason you'd need an app to pair a HomeKit device is if the SRP key isn't printed on the box, and as far as I know that was limited to some early devices that were shipping while HomeKit was in beta (I know LIFX lights were in that camp for example. After Apple announced HomeKit, they said that all existing lightbulbs will become HomeKit compatible through an OTA upgrade. But of course you had to use the app to get the HomeKit code)

HomeKit is a local protocol. The Apple device connects directly to the outlet or lightbulb it's trying to pair over bluetooth or through local mDNS.

Products with hard dependencies that CAN be bricked should be required to have a consumer warning label stating such, like the warning labels required on other products like cigarettes.

That's a standard lifecycle for "smart home" appliances as I've been observing for some time now. Tbh building a home that risks requiring full replacement of e.g. sockets every 3 years sounds at least unreasonable

I have ZWave devices that are now 20 years old which work like the day I bought them. Stick with a standard, local solution and you don’t have to deal with these kinds of messes.

Only local solutions. Also local solutions may also fadeout, depending on security measures. But people strive for cloud, turn-on-my-washing-machine-from-work solutions which are doomed to last 3-8 years max

This is why I'm glad HomeKit is the basis for Matter. Apple knew this crap would happen, and required HomeKit-certified devices to function "cloud-less".

HomeKit was not the basis for Matter, although obviously some HomeKit concepts (and I suppose actual HomeKit code) were contributed by Apple. The dominant input technology to Matter was actually Nest Weave. This can been seen by comparing the code bases.

Weave was designed from the start (circa 2013) with local control in mind. However features based on local control were a tough sell inside Google after the acquisition (as one might imagine). Today, many of the pivotal people in the Matter effort are ex-Nesters, and they, along with their (equally passionate) Apple counterparts have been strong champions of local control.

I wish they weren’t still selling Wemo as they announce this

I ended up getting several Tasmota based devices from https://www.athom.tech/. Run your own RabbitMQ like service to control them, and no internet needed. They're super cheap, open-source, and flashable with your own firmware if you want.

I stopped using Wemo app because it requires precise location to function, absurd. I have some cheap Belkin switches that had not seen a firmware update for years, should replace them with better products anyway.

I will never buy another Belin product. If they screw us once they will screw us again! Teaching these companies a lesson by never buying their products again is the only option consumers have in this day and age.

Anyone know if the older Wemo products (the switches) work any other way? My understanding was that they required cloud. I believe my home bridge Wemo plugin uses the cloud to talk to the devices, but maybe I’m wrong.

I have wondered about this also. I noticed that some of my older plugs seem to keep on chugging even when internet is out. I am guessing they have some internal clock that they fall back on when the internet is not available. However, this would only work until the first power outage happens, at which point they would no longer know what time it is, and there would be no way to reconnect to them via the Wemo app to sync up their internal clocks.

They don't require the cloud if all you want to do is turn them on or off. You can send SOAP requests to them, or use OpenHAB, HomeAssistant, etc... to control what you have, but longer term it might be a good idea to invest in Tasmota or ESPhome switches if you have technical knowledge. Paying up for Lutron might be a more long term solution if you want hands off, though they don't have the device coverage that Wemo had.

https://www.openhab.org/addons/bindings/wemo/

https://www.home-assistant.io/integrations/wemo/

It seems pretty unclear what is happening with the app. The article says

> After this date, several Wemo products will no longer be controllable through the Wemo app.

That makes it sound like some Wemo products will still be controllable through the Wemo app. But when you read the rest of the article, there are zero details that indicate this, or even imply it. This bit seems to contradict that notion entirely

> The Wemo app will no longer be updated after January 31, 2026, and can manually be removed from your smart device.

It would be great if these devices could still be controlled/configured locally via the app, but I'm guessing that won't be the case. If they froze the functionality in time, I wouldn't be as unhappy. But if they're basically turning their products into bricks, I'm pretty not thrilled.

At least with the older devices, all the app does is a upnp broadcast to find the wemos, and then issue SOAP requests to turn them on/off. The app will work without an internet connection - you can verify this by disabling your internet and checking. If you need automation or firmware updates, or possibly with some of their newer devices, then you need the internet connection. So just keep the app around (back it up or download from f-droid if on android).

Older Wemo switches can work locally with Home Assistant using the wemo integration, which communicates via UPnP without cloud dependency. You can also flash some models with alternative firmware like Tasmota if you're comfortable with that.

Any good youtube videos you'd suggest for how to do this? I haven't taken the leap to using HA, but have heard good things and considered it. If it would keep my old Wemos alive, I'd do it!

By "older" they mean some that are less than two years old.

I had to use a different title since the original was confusing. I considered using the title "nearly all", but didn't want to make it seem like I was editorializing. But honestly, that might be the more accurate description.

This is why any piece of automation that comes into my home is required to function without internet or an app, in its own private vlan.

Reply on news.ycombinator.com