Wink Labs just announced that their home automation hub, the Wink Hub, is “transitioning to a $4.99 monthly subscription, starting on May 13, 2020.” Should you fail to pay the fiver every month, you will lose access to their app, voice control, and automations, which is everything it does as far as we can tell.
This is an especially bitter pill to swallow for Hub users, because the device was just that — a hub. It speaks Bluetooth, Z-Wave, ZigBee, WiFi, Kidde, and a couple other specific device protocols, interfaces with Amazon’s Alexa, has a handy Android master panel app, and had a nice “robot” system that made the automation side of “home automation” simple for normal people. In short, with its low one-time purchase price, compatibility with many devices, nice phone app, and multiple radios, it was a great centerpiece for a home-automation setup.
“Nice home automation system you’ve got there. Would be a shame if anything happened to it.”
The Services Shakedown
In a sea of home automation devices that are paying for themselves by spying on you in order to allow multinational monopolies to exert yet more control over your buying and browsing habits or subvert warrant requirements by providing video surveillance to police departments, the Wink Hub was a refreshingly privacy-preserving device that was sold for a one-time price.
But it looks like they fell into what we’ve called the IoT Trap: the company offers a device that relies on cloud services to back it up, misjudges the expense in keeping the back-end infrastructure running, and either shutters the device or has to try to get money some other way to keep the lights on and the investors satisfied.
The writing has been on the wall for Wink for a while now. This particularly prescient writeup summed it up nicely: “The company doesn’t offer any monthly subscription for the service its hubs depend on…if a company is dependent on hardware sales for revenue and that hardware does not exist, then what future can that company have?” Well, now you have a monthly subscription fee. Will that pay the bills going forward?
Ask Hackaday: Alternatives?
I’m sure that Wink Hub owners are feeling a little blue, if not bitter, right now. After all, getting a $5 per month bill for something that you thought was free, attached with a threat to just turn it off if you don’t cough up, seems tantamount to extortion. That’s because there are no alternative services, you must either pay Wink to use their service or your hardware will cease to function. (I assume that their lawyers have done their research and that it’s allowed within the terms and conditions — it only feels extortionary.)
Compare this with the experience of Nest/Google/Alphabet’s past hub, the Revolv. Sold with a “lifetime guarantee” of service, the lifetime in question was that of the product line rather than the owner, and that was significantly shorter than many expected. When faced with a PR nightmare, and under pressure of an investigation by the FTC (PDF), Alphabet/Google generously refunded folks their money. But the Revolv was no longer the central hub that original customers had wanted.
Smaller companies may not have deep pockets for similar largesse. Sonos tried to get users to scrap their speakers voluntarily. TCP shut down all of their IoT devices a few years back, rendering a ton of lower-priced connected lightbulbs useless. Best Buy’s IoT offerings went the way of the dodo six months ago.
Wink Labs is probably also in this category, and their options may not include refunding everyone, so this means they must charge more money or call it quits. At least this lets the Hub owners decide if it’s worth it to them to continue to pay for the device in perpetuity. But if you look at the cost of owning one of these for the next ten years, you’re talking something north of $400, discounted with today’s low interest rates.
Trapped!
Which brings us back to the IoT Trap. If Wink knew upfront that they would need the equivalent of a $5 monthly fee to support the hub, they should have originally priced it at $450 or more instead of $70. (And we thought the $300 Revolv was expensive!) We wonder how the same marketing team could come up with these essentially incomparable figures only a few years apart.
So here comes the “Ask Hackaday”. What should Wink Labs do here? They probably can’t refund folks, so they’re left with bricking all the devices or charging a monthly fee. The first choice is tantamount to just calling it quits, and the latter feels extortionary. Those of you who own their hubs, which path would you rather they follow?
How could they have ended up in this position in the first place? And what other firms stuck in IoT traps do you see out there? What’s going to be the next Wink hub?
This being Hackaday, we honestly hope for a third option: Wink open-sources the software, allowing tech-savvy customers to host their own local servers. If it’s truly the software and hosting costs that are killing them, and they’re profitable on the hardware as a standalone, this would be a win for everyone. We won’t hold our breath, but if too few people take them up on their service contract, or the FTC steps in on behalf of consumers here, this current chapter won’t be the end of the Wink saga.
Thanks to [Hagan Walker] for pointing us to the Wink story!
I can’t say I’m surprised. When Nest came out with a “connected” thermostat that was dependent on a WiFi connection and Other People’s Computers(tm) to operate, I wasn’t interested. Same for Ring, et al.
Let’s rewind a bit back to an appliance I owned 2 of: Replay TV.
When you bought a Replay TV 2000 you got a lifetime subscription to the programming info. The early Replay TV had a modem and it would call home to get listing data. The later Replay 4500 had Ethernet and could use your Internet connection to do this (as well as connect several peer Replays together in the home).
The Replay 4500 I had was one of the first models released with a subscription for the guide data, you could pay monthly or a one-time purchase of $200 for the lifetime of the device. Lifetime ended up being tenuous as Replay TV went bankrupt and their assets were sold on several times.
Wink would have been wise to follow the Replay TV model, grandfather existing customers into the “lifetime” service, then use recurring revenues from new customers to pay for the upkeep of Other People’s Computers(tm).
OH LOOK, CLOUD SERVICES GOING OUT OF BUSINESS OR LATER CHARGING PEOPLE MONEY OR NOT HONORING THEIR LIFETIME PROMISES ALL CONTINUES TO HAPPEN. WHO WOULD EVER HAVE THOUGHT THAT THIS WAS A POSSIBILITY?
Also seeing this with online software “cloud only” packages and other junk that literally disables working hardware due to companies going out of business. All it serves to do is prevent meaningful progress while trying to further screw the sometimes unwitting customer.
It’s the complete antithesis of open source and is a horrible approach to take in almost every case and arguably time and time again a horrible business decision to take on top of that.
Yes, OK. Now what are we going to do about it?
We could start by getting tech journalist to stop gushing and praising this type of business model every time some company come out with gadget that requires an always on internet connection. This entire situation is more to blame on the journalists that flood CES every year and talk about how amazing a internet connected refrigerator is.
Amen!!!!!!
I know we all have our favorite axes to grind, but you’re actually trying to put the blame on journalists? That IoT firms seem to consistently misestimate their cost structure?
You might as well blame the lawyers, or the umpire, or your mother in law…
Well said!
Whatever “misestimating” is to blame, this is hardly the first dozenth time this has happened. It’s a common problem you’ve even given your own name to in this article!
The financial details of this don’t matter. Although that said, how contrivedly naive do you have to be to realise that companies need money, and that $70 != $infinity?
But they don’t matter. Any “journalist” who knows anything about this field, and to be writing about that field, you’d assume they do, should know that not only is this a thing that can very obviously go wrong, but that already has, dozens of times. Anyone reviewing some new gadget really ought to point out this huge potential downside, and particularly if there’s no apparent way for the company to pay it’s bills for the expected lifetime of the gadget.
HAD is OK, you understand this well enough. But consumer tech writers (I’m really having to work not to put a dozen “so-called” quote marks in this) aren’t doing a good job of warning the public. And they ought to. Way back in the day, journalists actually served the public, although back in the day the Antarctic was warm and balmy, but you shouldn’t book a holiday there.
I’d love to be a tech journalist. Imagine being paid to copy and paste press releases!
Why limit yourself just to tech? Although your background will come in handy in pleasing the algorithms who all media is written for now.
Good point.
By a UART dongle and root it, at least thats what im doing.
Will this run all wink compatible hardware?
Would you Please expand on this idea.
Everyone boycott Wink! I certainly am. And I’ll gladly spread the good word.
Years ago, my wife and I had a Sirius (Pre-XM) radios in our cars.
After an initial “honeymoon”, where we frequently used the radios, they fell into disuse.
Trying to cancel our subscriptions was like pulling hen’s teeth. Long wait times on hold, getting the runaround to different departments, and eventually (once after 45 minutes on hold) “loosing” the connection.
They continued to charge my credit card each month for the subscription and automatically renewing the subscription each year… Until the credit card expired. THEN, they were calling me! Asking for a valid CC number.
(maybe I should have strung them out like they did me).
One call to the CC company would have fixed that…
There ought to be a “chargeback with maximum prejudice” option for companies like that, that credited you an extra $20 for dicking you around.
I wish I had thought of that..
I noticed that in my limited interactions with them, that they being “the” satellite radio service, they have an attitude like “the” phone company, i.e. you owe them a living and should feel privileged that they deign to speak to you at all. We got a free trial with the new car and were laughing at them when they tried to “sell” us service, I quoted sell, because it was more a case of “Make us believe we were obligated to pay and as a favor they’d leave it on.”
I did try a few channels in the first week, to see what the fuss was… annnnnd liked it so much I didn’t even bother for the last 3 weeks of the trial. It has no flow, it’s like they downloaded the music collections of several people with radically different tastes (Loosely fitting the channel’s genre) and put it on shuffle. It sounds like an uncurated mess. 90% of the time I put the radio on while driving I want the station with most relevant local traffic reports, weather I might be headed into etc. That’s still going to be on FM. If I want to listen to some opinionated idiot rant for a long drive, that’s on AM.
“We got a free trial with the new car and were laughing at them when they tried to “sell” us service, I quoted sell, because it was more a case of “Make us believe we were obligated to pay and as a favor they’d leave it on.””
That sounds suspiciously like the interaction my father had with them about 10 years ago. They called him six months after he bought his car and told him that since they forgot to deactivate his subscription after the free trial period, he now owed them back payments… Which they would waive if he signed a two year agreement right now. He simply replied to the person on the phone that he was a retired corporate litigation attorney with nothing better to do than take them up on their lawsuit threat, and never heard from them again.
So much this.
I bought a car last summer that apparently included a trial Sirius XM subscription. I didn’t know that until I started getting ‘ emails. No big deal; ignore em.
AND THEN MY RADIO STARTED SWITCHING OVER TO THEM FROM WHAT I WAS ALREADY LISTENING TO!
(Insert your favorite expletives here)
I’ve gone from indifferent to them, to actively telling everyone I can not to use them.
My wife has it in her car and everytime it’s promotion period is about to expire she calls up to “cancel”. She usually gets through in about 5-10min and the operator always seems to fall over themselves to keep her as a customer so they apply whatever promotional package my wife says will keep her subscribed. She told me once she pays on average about $3-$4 a month. Been practicing that same tactic on them for 4 years without fail.
I’ve been doing something similar to my ISP for as long as I’ve had internet. Call them up right after the promotion expires, tell them I’m considering switching to some other ISP and they always offer to put me on the current yearly promotion to keep me using their service. Been paying $20-30 under the normal $70 they charge for over a decade.
This used to be the game with AOL, 20 years back– “free 56k forever” in spite of the (ever-increasing) free trial duration. Just try to cancel before they start charging the CC or checking account you signed up with. It’s kinda obvious now that they just wanted more eyeballs for kinda obvious reasons.
One last thing of note, if you’re going to try something like this, go out of your way to be nice and pleasant with whoever you get on the line. I’ve had many a tech support person/operators tell me that they generally don’t cave that easily but because I was so nice to work with they wanted to make sure I had the same experience in return by giving me whatever promotional offer they could to reduce my bill. It both pays and saves to be a decent human.
the easy way dealing with such companies: call CC company and ask for a new card because yours was stolen/lost. One week later, the problem is solved and those companies will never hire a lawyer for that.
ReplayTV was life changing for me. It was amazing how seamlessly they worked when networked or just in general. It’s not generally understood that ReplayTV predated TiVo and was the originator of the DVR as we know it.
Yep. If you buy into non-local IoT, you deserve this. Don’t let companies pull this on you. Everything in your house doesn’t need an internet connection. The solution? Get up and use a light switch 🙄
‘Deserve’ is a bit strong.
While it’s ultimately the consumers’ responsibility to be informed, the truth is the vast majority of these consumers not only aren’t informed, they don’t know they aren’t informed.
This is always the problem with mass market products in a new space. Most of the users are technologically naive; they have neither the formal education nor experience to be informed of the risks IoT presents, and they don’t realize there’s a gap.
Because who would tell them?
I bought the Wink 2, and specifically checked that it would function locally without an internet connection. Their support page *still* says it does. Yet their extortion…errr, notice of subscription says if you don’t subscribe it won’t work locally.
Now, I understand they are in a bind-no hardware sales to support the server side of things, a subscription model sadly is how the lights stay on-and I am in the demographic that can afford to pay.I have 1 light-and I have gotten used to using it as my alarm clock. But! 7 days notice? And lying about the local operation? That sticks in my craw, and I will find another option.
Business majors need to study some of these things to understand consumer behavior, because way too many of them are dooming their companies to spiral down in flames… What loses the customers in many cases is not the cost, but the marketing BS like they think people aren’t paying attention.
*this*. I did exactly the same thing before I purchased my OG wink hub, and the documentation also stated that it was local service.
Not going to reward a company for such blatant misrepresentation.
If the documentation is not an outright lie, then the only thing the subscription is buying me is the ability to access remotely, which is not required or desirable for my use case. If I’m using the service 100% locally, then I dont need their service at all.
I’ve made it far enough to discover an ssh server running, but never felt a desire to figure out more. Now? Definitely worth poking at it. I had hoped maybe somebody else had gone down this road first, but it seems unlikely…
Here’s the the thing, people don’t get. Wink did not offer a lifetime free. It simply states No Contract No Subscription Fee. No where does it state FOREVER. If you read the terms of service it does state that Wink can make changes at any time.
“No Contract No Subscription Fee” That’s why I bought it, that’s why I will discontinue it when they start charging. Disappointing.
Perhaps they could release their code and let people run a Wink server on a Raspberry Pi.
I think the spec for communicating with the hubs should be opened up. This way, you could subscribe to Wink out of convenience (and for their smart features) or a motivated open source community could stand up an alternate project.
There are several open source options HomeAssistant and Mozilla WebThings.
will these open source ones work with Kiddie ? KN-COSM-B-RF Smoke / CO Alarm
Never mind ‘perhaps’ – this should be a legal requirement for anyone/business that makes an offer of ‘free’ services then fails to uphold the deal.
In the same way as there are court cases for ‘right to repair’ there should be court cases against companies that deceive by ‘intent’ (or not). Simply put – they screwed up so they should forfeit.
The open source option is best. Failing that, I can’t think of a reason that they couldn’t allow local control only. I plan on contacting the FTC/state AG. It was sold to me as a device with no subscription. I understand their bind, but it is their own doing and giving customers a week to pay up “or else” is unacceptable.
Simple fix. Allow local control from the app to the hub. Also open up local API. The HUE hub had this from day one.
I completely agree, this is the approach I would expect as well. The Wink was fantastic because it could keep the entire interaction local if onsite. I use homebridge to connect to HomeKit anyhow so all I’m looking for is a local hub for z-wave, zigbee, RF, 3rd party devices, maybe some automation but HomeKit can do that.
Problem is the parent company is broke so they are making a money grab. I hope they release the source code and don’t let it go down with the ship.
I’m not terribly familiar with the devices, but exactly what features require any sort of network connection outside the local network? Does it allow you to remotely access connected things, for when you are out-and-about, and realize that you’ve left your lights on or something?
It seems like there should be some sort of middle ground. If you want to utilize the features just on your local network, the device keeps working without a subscription. If you need remote access, then you pay the $5/month.
I know a lot of internet service providers don’t like it (or allow it), but it may also be possible to expose the functionality on a non-standard port.
If the device is hosting the content, and the only thing needed is some sort of lookup to access it (e.g.: [serial number].hub.wink.com. The users hit that domain for their specific device, and when the device is turned on, it calls home to say “hey, resolve my serial number to this IP address I’m reporting from”. I’d think that a setup like that would take a minimal amount of server resources to keep running, and in the mean time, hope that enough people continue to buy the more premium service (or hardware) to keep things running smoothly.
1 year free service. If the user has already had it a year, one month and then fees start. But a year should be included with the device cost. So if I bought mine last month I have 11 months free left.
I bought the local Home Depot out of their smart outlets when they switched brands and dumped the old ones for $4 apiece. Fortunately I wasn’t interested in their IoT functionality; they’re ESP8266 based and thus hackable. But before I reflashed the first one I listened in as it went about its original business. It reported back to a server in China every minute or so whether anything was happening or not. All the timer functionality was dependent on the server because there is no battery backup or clock chip. It will keep the time if it loses internet as long as it has power, but after a power cycle it doesn’t know the time until it phones home. For most of these devices the server is why you don’t need to arrange a fixed IP and port routing on your home network; your phone goes to the server, which knows how to get to the device because the device also went to the server. It also sends every switch event to the server, and my wifi credentials (in plain text over the serial debug link at least). I felt no guilt at all about installing NodeMCU on it.
Now THAT’S concerning. What was the brand name on the outlets (just ‘Home Depot’ or something else)?
On a scale of 1 to 10 of IoT misdemeanors it’s only about a 2.
Fine … I’ll only worry about 1 out of 5 workdays.
They were branded as “ECO Plugs” but there was a bit of a hullabaloo on the Parallax forums as we tried to figure out who really made them. Consensus among us was that all such devices seem to operate similarly no matter who made them though.
A large majority of the 8266 based smart devices use the Tuya backend
Alexa isn’t connected to your smart device – Your device is connected to the manufacture’s API and Alexa just messages that manufacture’s API. So the devices is constantly connected to that manufactures service – basically a constant web connection – otherwise if you wanted Alexa to turn on your lights you’d have to go through the whole pair process every time you want to do (command) something. That connection costs the manufacturer money – thus why they are now trying to get a subscription fee.
Do you hear that Mr Anderson?
https://i.pinimg.com/236x/34/ba/e0/34bae0eeecc8c32cd83b01e5eb7bd616–agent-smith-the-matrix.jpg
Personally, I wish that the IoT companies would shift to a different model. I purchase hardware and software to meet my needs. If that hardware or software is totally dependent on an internet provided service, I really have to need it bad, or I don’t buy it. The only thing I have currently is an Amazon Echo (and it was a gift). All my wifi controllable cameras, light switches and such are local controllable and with a free service. If the service died this afternoon, I still have full control (maybe without Alexa’s help, but She might disappear, too). I believe, that if a business goes under (or wants to go to a Service as a Service model), that they be required to release their software, or refund (totally) the purchase price. If, I then choose to keep it, I have a choice similar to when I bought it, not be extorted.
Serves anyone right for buying into something like this that depends on a “cloud”, er, I mean someone elses server somewhere. Any such design is stupid. Most of the time MQTT is stupid. Get rid of the middleman. Just say no.
“MQTT is stupid”.
Ironically, the extreme ease of setting up the whole backend software stack, from broker, to databasing, to program logic (Node RED, etc), to front-end GUI goodness that it all enables is the best reason to love MQTT, IMO. Because of MQTT as the common transport layer, a whole metric ton of cool infrastructure works together, and is easily installed on a Raspberry Pi (etc) in your own living room.
Your MQTT data doesn’t need to leave your LAN to be extraordinarily useful. Heck, you can even use it for interprocess messaging on localhost. :)
Yes MQTT while not perfect certainly simplifies a lot of issues for my automation/monitoring system. The server is locally hosted currently on a Pi. If my internet goes down everything keeps working nicely. Any data I feel I need to access remotely I post elsewhere and all is well in my world.
Anything that doesn’t NEED outside assistance should never connect to the outside world.
+1 as usual for Mr. Williams! 😁
Having had a number of Wink devices (hub, obviously, z-wave door sensors, alarm, motion detectors, one multi-function wall switch, and the litany of compatible other devices) for over 5 years now I am not shocked. The service has been iffy at times and the connectivity to 3rd party devices (Nest, Hue, MyQ, and so on) have been flaky and needing a bit more maintenance than I’d throw at a techie buddy. The robots and “Look Out” feature is the only thing that makes me consider paying the $5. Look Out acts as a pretty decent security system when configured properly and has all the interconnections to make it as robust as other services that charge more than twice the asking price.
That said, I picked Wink out of laziness. I had already spent enough time with a handful of X10 descendant products and wanted something easy and Wink did that. I’m not thinking the DIY open source method is the only way to go. I’m sure if I can find an IFTTT compatible system I can duplicate the functions I care about for little one-time and no recurring costs.
The fact that their hubs are costing so much to operate that they need $5/mo from every user says that they’ve probably implemented their services incorrectly. They should really only be acting as private DNS to provide the public IPs of associated with a given account or as a socket relay only if absolutely necessary. Ubiquiti’s Unifi does this well as does Plex. If Wink had kept their hubs up to date with support for new hardware (which they haven’t despite claims within the last year that they’ve improved support for various IoT devices), then I’d be happy to pay them $5/mo. They’ve already demonstrated that they’re not doing that and that all they’re really providing is a remote access service which I can do without (and haven’t been using much at all anyway).
The hardware in the Wink Hub 2 is actually pretty nice from a connectivity standpoint and I would absolutely love to have the ability to write my own updates to it. There are amazing things done with the original Wink Hub since it was rootable. As it is, I’ve locked my WH2 off from the internet and verified local control still functions, but I can’t rely on that based on how the announcement was worded. Luckily for me, I had already planned to transition off of the WH2 due to the aforementioned lack of useful updates and I have hardware in hand that will go into place this weekend to enable my Home-Assistant system to take direct control of any devices I haven’t already migrated off of the WH2.
I guess I should have tossed in some links.
Wink Hub (1) is already hacked wide open as you said:
https://hackaday.com/2014/08/09/defcon-22-hack-all-the-things/
http://arahuman.blogspot.com/2014/11/how-to-root-your-wink-hub-step-by-step.html
Wink Hub 2 is not yet. If they start showing up cheap on eBay, I would place my bet that it gets rooted eventually. Secrets don’t hide forever, and suddently charging for a previously free service might just accelerate the process.
I don’t have a Wink anything. Does them shutting down the app prevent you from using the device standalone?
They have since patched that vuln. You have to go in over serial and open it up now. I foolishly updated the firmware on mine and have to do it the hard way now.
But “go in over serial” is an option? That’s not so bad.
Yup, I’d say $5 a month is way too much, it’s costing them, but shouldn’t be costing them NEAR that much. I mean isn’t that like a minimum AWS instance, or a digital ocean droplet? One of those should be capable of handling 20-50 people’s hubs. $20 annual is the top end of fair, $10 if you actually want ppl to believe you’re doing it as a service and not a profit center.
Still, they need staff, premises, tech support, etc. That’s probably more of the cost than server cycles. Probably most of it, actually. Just running a business costs money.
I’m not even sure if $5/mo is too much for a service of it were good – but I’ve been using my Wink Hub 2 for two years and while it works, it doesnt work well. I actually thought Wink had stopped maintaining their products a while back.
It can take nearly 20-seconds to turn on a light sometimes after pressing a button on my Wink Relay or using the App on my phone when I standing at the bottom of my steps. My understanding is that even when I’m on my LAN, commands take a trip around the Internet before they reach the Hub.
I went with Wink because their Hub had Z-Wave and Lutron radios built in (Samsung at the time might not have even had Z-wave, but I dont remember).
If not paying a subscription fee meant that I’d only lose control away from my home or Alexa integration (despite Alexa having become our main way of interacting with our lights) that’s a choice I can make based on cost. But the other “smart” features such as timers or macros (Robots as Wink calls them) – like I have three outside light switches integrated (via the Wink Robot) with a fourth indoor switch so we dont have to go out in the rain to turn them on and off. I’m not happy about having to start paying monthly to keep features like that.
On Principle, I will probably migrate to an RPi-based system or might switch to SmartThings (both would require my getting a Lutron Hub for my two lutron dimmers – but still cheaper than paying Wink forever.
I would love to be able to hack the Hub 2 so I can just use the transceivers and run the software on a Pi.
This makes me feel better about being slow to adopt anything IoT. Right now the only IoT device I have is an ESP32 doing a clock
Isn’t there some system that can be run on a small Linux board and that is flexible and capable enough to be interesting for a big chunk of these “not so tech savvy” users? Node RED may be a candidate.
There may be a market for such a small Linux board, in an enclosure and pre-installed, complete with a starter set of remote switches, and runs a web server to connect to.
There are about 10+ open source home automation suites, but can any be simply ordered as a complete beginner set?
I have never looked too deep into this home automation stuff. External servers will never be an option for me.
I once fell very shortly in this trap.
I bought a squeezebox once. I struggled with it for an hour or so, but it refused to do anything without first creating an account with some server who knows where.
I brought the thing back to the store a day later, and now I have mpd running on a small Linux board.
A lot of people swear by NodeRED. Other candidates include the various front-end suites like OpenHAB, Home Assistant, or Domoticz if you’re feeling a little geeky. I think you can get most of these as full setups on a burn-it-and-boot-it Raspberry Pi image. The Beaglebones come with NodeRED out of the box.
MPD is the shizz. It’s been my stereo since I got my first good USB audio interface back in 2001-2003ish. I even streamed audio (icecast) off my MPD server to myself and a friend at work for a few years. So flexible, so awesome.
Elliot – have you looked at WebThings by Mozilla yet? I’d be interested to hear your detailed take on it.
https://iot.mozilla.org/
No, not yet.
Honestly, having just posted what I did, it might be a nice topic for a multi-part series: kicking the tires of the different OS home-automation suites.
I’ll put it on the list…
Look into Hubitat
Squeezebox server still functions! The web portal for it still works at mysqueezebox.com. And so does my Squeezebox! I still use it to connect to a few internet radio stations all over the world. And for its day, it handles many formats including some lossless ones. The company that built it envisioned Open Source software. SlimServer was the music server if you wanted your own bare-metal server. Booting up Sqeezebox, the beautiful VFD asks if you want to connect to the cloud or your own server, before it times out and uses your last selection. See https://en.wikipedia.org/wiki/Logitech_Media_Server for independent operation. After acquiring Slim Devices, Logitech had their own spin for a while, then discontinued it. But they did not shut down the cloud server (yet) and for this I am thankful. It has now been 20 years? If they do, having your own LMS is still possible… if necessary with Alexa, et al.
Making your “smart” home dependent upon some external backend is a terrible idea. It always was and always will be. Avoid using devices like that, especially for core functionality. These companies have short lifespans or suddenly realize they are not making (enough) money and then they do stupid things because of it.
Wink has had a long lifespan and I give them credit for holding on this long and also for giving existing customers and option to stay on. If Vera never had to charge for remote access then Wink shouldn’t, but I think it is an acceptable last resort. Sure releasing the source and opening it up would be nice too, but the unrooted Wink was sold more toward people who didn’t want to tinker. So this is an easy option for those people. The same thing happened with Chumby, they gave people the option to keep access for $3 per month.
I find it very difficult to believe it costs anywhere near $5 a month.
How many wink devices do you think a single Amazon EC2 A1 instance could power? Is it more than 4? Because you can get a A1 server for less than $20 a month.
And I think a more realistic estimate is several thousand winks per A1 instance.
The thing receives small packets from the voice server (Amazon Echo or Google Home) and sends small packets to the house. We’re talking less than 1MB per user per day.
I don’t think this has anything to do with covering their server-side costs.
$5 to cover server-side costs + to continue paying investors (such as Will.i.am.) of this company registered in Delaware.
I’m running a integration service for Alexa on the cloud and my costs, currently covered by amazon are 2 cents a user
It’s not what it cost for them, it’s what they think they should earn per subscribed customers. It’s a basic return on capital spent.
The solution is pretty simple, just purchase a locally managed system like HomeSeer. The system runs totally standalone, you can access it remotely if you open a port on your router. It is open so you can write your own extensions and the API is open sourced. If HomeSeer goes away, your system continues to function. There are few other systems like Hubitat and Vera that are also locally managed, so you have choices.
This is the best answer for the current hub owners who don’t appreciate being “winked”. Find a different hub that can communicate with the protocols supported by your devices, being sure to select a hub that doesn’t depend upon an external cloud (unless you want one.) HomeSeer and Vera are mature commercial options, and won’t disappoint you. OpenHAB, Domoticz, and HomeAssistant are open source solutions.
(Full disclosure: I’ve run Vera hubs for over 10 years; other than growing pains I have been very happy with them.)
I have one hubitat hub. all is local. it manages caseta lamps (local), zigbee and z-wave devices (local). added node-red for additional connectivity (external devices, all local). Works fine, No cloud. Just peace. And no fees
I have read somewhere about a possible solution: the cloud services for such devices should be standardized and modularized. Then the user could pick and choose between various providers, who would not be interested in the devices themselves, they would just run servers for hire. Even if the hardware company goes under, the service works as long as there is a provider and a user willing to pay for it. It should also improve security: if a provider was sloppy with it, the users would just switch to another one.
I have a simple rule for IOT devices: If they can’t work without going outside my LAN I don’t get them. If I ever decide to enable remote access to anything it will be using software that I control. Period.
When Wink was introduced, the market was littered with cloud-managed hubs. There was Wink, Revolve, IRIS, Staples Connect and SmartThings. The ONLY one left after this will be SmartThings… and for how long? The lesson is pretty clear. Dump the cloud and move to a local hub like HomeSeer, Vera or Hubitat.
Insteon lighting and the ISY-994i series controller have been around long before all of these and will probably be around long after. Their stuff works with or without a cloud connection which is why Im still using it today. You can use their cloud connected Hub if you want simplicity. Even that is okay.
This has class action lawsuit written all over it, Wink is finished.
Their only way out is to release the software needed for people to set up their own open source server as open source.
Why anyone buys into these connected devices that rely on a company providing a service in order for it to work is a mystery to me. I refuse to buy anything that won’t work on at least a basic level without a service.
So no cell phone, no ISP, no television?
The middle ground solution that nobody is going to like to hear, but that would work:
These peckerwoods are no doubt hosting their server-side on somebody ELSE’s cloud service, with some kind of elastic-swell-to meet-demand option that is now, no doubt, eating their asses.
SO, just build a “standard” server image on whichever linux platform they are no doubt using, for whichever cloud service they use, prolly AWS, and make a micro-image version available of it for no charge.
They continue to sell hardware, make money, and include of free service, fee per month after that. If you don’t wanna pay them for the back end, host it yourself on an AWS micro image (and park other shite up there as well).
Oh, and agree to update their packages for the free image going forward.
That solution would give them certain liabilities, admittedly probably not big ones, for no gain at all. Besides all that, they’d be giving up $60 a year! It’s much better to be a company that sells nothing, and just leases something ethereal like information or, even better, a “license”. That’s how software companies are all going. In this case they just needed to create a market for their services first by selling the hardware that needs them.
In fact it might be this was the plan all along. Would you rather pay a high initial price for a product, to pay for it’s ongoing support? Or pay a monthly subscription? But wouldn’t you rather pay NEITHER!? Right, and that’s how it went.
If $5 a month is anything like what it’s actually costing them to provide the service, the products must have been meant as a loss-leader to start with. The purchase price surely couldn’t have covered too long keeping the company going.
If it were my own personal choice I wouldn’t bother with AWS for running all this. It’s not a lot of data, just a few voice commands a few times a day per customer. You could do it on in-house servers with a rack or two surely enough to support as many customers as they have. Much more control, then, you can run your own raw code without meeting any of Amazon’s specs, and it’s likely to be cheapest.
Anyway… you think customers are going to learn from this? No, I don’t either. It’s only the hundredth time this has happened to this same market, maybe next time monkeys really will fly out my arse, completely free of charge forever.
Why do you use North as a substitute for more? It’s just a direction not a quantity, you might as well use left or behind. I don’t know where you are from, but here, the North is the place of cheaper prices, lower wages, colder and less populated, although it is better but when you’re talking costs, North means cheaper which makes your use of the word nonsense, or should I say South?
Funny words, right?
I think it’s because it’s “up” on a map. Of course, that’s a northern-hemispherism. It would completely confuse a Kiwi.
I could have written “upwards” huh?
If I were Wink I would implement some P2P distributed connection scheme into the hubs firmware. That way I, as the Wink, would charge my clients $2 monthly effectively for nothing. Nobody would bitch for $2 monthly.
If you were Wink and you did those things you would be sitting in a courtroom, paying a squadron of lawyers $2000 an hour to get you out of the mess you made and keep you out of prison for fraud.
If the folks at Wink have any integrity, sell or even give the company to one that can provide the service. I am sure several of the larger companies would like to absorb their customer base. They may be able to add the Wink devices to their own network of products. Though, of course, even a big company can leave consumers with a bunch of useless devices.
I suspect it is time for “smart home” industry to get regulated like a utility. Perhaps require them to provide as much of the functionality as possible without the server. And clearly label on the box, “warning the following features could be discontinued at any time…”.
I really want to upgrade my X-10 system, but will not make the investment until there is a good solution to avoid the pitfalls of the cloud-based service.
I totally agree!!! I’ve been doing this since the BSR-X10 days. I have a room full of “junk” hubs etc. I have a Smarthings piece of crap, a Sylvania Osram hub and two downlights that don’t connect to anything, a Philips Hue, and the list goes on. Most of them are tied into somebodies controller service so when in the case of Sylvania they go out of the busines, the money I paid for the products goes down the tubes.
What these companies fail to acknowlege is that their code comes from DIY (Hackers) in the opensource community. They are making a profit of the work of others. The cell phone data networks are a direct result of the HAM communities work. Their code is basically the old TNOS. The same for the whole PC world. That all started as a result of Hackers taking the old 8088’s and creating something useful. IBM just rode on their coat tails.
One of the biggest offenders out there is Verizon. They take all kinds of opensource code and use it in their networks and NEVER give anything back to the open source community.
So paying someone else for a service that they basically stole from the open source community just gets the hair on the back of my neck to stand up.
” That all started as a result of Hackers taking the old 8088’s and creating something useful. IBM just rode on their coat tails.”
While the idea of personal computing, owes a lot to a grassroots homebrew movement, the IBM PC was practically an Intel application note made flesh.
PC-DOS was basically a hacker/fan port of CP/M from the 8080 to the 8088 which Microsoft bought for $50K and slapped a different name on.
The lousy bastards, paying the sum agreed.
I’m sorry but open-source licenses do not contain any sort of provision requiring users to give back, in fact they explicitly encourage people to do just as Verizon has done, to freely use the software within the limits of the license.
I do find it rather fascinating when people complain about free software accomplishing exactly what it was supposed to do, establishing itself in markets beyond the hobbyists.
What you are saying is: “yes it is free software and no you don’t have to pay for it but really you do have to pay for it even though it says you don’t.”
And how much have you “given back” for what you’ve taken? If you’ve gotten $10,000 of value from the linux kernel over the years then you are the freeloader because you haven’t sent your check for $10,000 to Linus.
Why not a clause to release the software (or at least the communications protocol) to the public through the license, if the support ends, or goes to a pay model only model? I don’t see a problem with companies using FOSS in their product, and there are many ways to contribute back.
How do know the origins of the code in question? How do you know that Verizon wrote it and they didn’t buy it from someone else, in which case Verizon has no rights to release the source?
“there are many ways to contribute back”
https://verizon.github.io/
wow, thanks captain obvious
And yes I have. Have you? I actually know how much work is involved. I write my own software and yes release it to the open source community. I pay for using someone else’s work. Oh and by the way if you haven’t yet read it you should actually read the software licensing such as MIT. It states that if you use licensed software you must publish what parts are opensource that you use. Verizon NEVER does that.
wow so now the MIT license applies to the linux kernel?
Oops. Old age must be setting in. 8088 should actually read 8080. They were after all 8 bit processors. Cheers.
Lazy Wink owner here, honestly up until this point I haven’t had any significant problems. I completely agree that they should open source it, which is something I believe I suggested in the email I replied to them with when they notified me (nestled among a few other choice words). They also picked a really bad price. Honestly, for $10 a year, I would pay up. I would be mad, but I also wouldn’t have to reconfigure my fairly extensive IoT infrastructure. The effects of this are already being seen; just a few hours after this announcement went wide, the price of the Hubitat on Amazon jumped $30. Alas that is where I will go, I will sacrifice a simple interface for local control and a few hours of my time. I would hope for a class action, but let’s be real that will make some lawyers very rich and a bunch of people will be $5 richer.
Yeah. You’re not going to get enough out of a class action to buy you dinner. Sigh.
My guess is that they thought “if we charge anything at all, 50% of our userbase will drop out” and then jacked up the price until they could survive with 1/2 of their users. Which of course, raises the number who would drop out, so they recalculate. Eventually they reach a fixed point and it’s probably expensive enough that many folks (like you) will drop, but the few staying in will keep them afloat.
That is, if they didn’t just make up a number and wing it.
Anyway, my point is that due to the dynamic alluded to above, the price is gonna be high.
When I worked at an TV & Appliance store (decades ago) we sold TV stands for $35, even though we “could have” sold it for $20 or so. When I asked the boss why we charged so much for it, he said, “If the price was lower, customers would just ask us to “throw it in” with the TV”.
Once we delivered a new TV to a customer, and was picking up the trade-in.
The Horizontal Hold on the TV was off. I was ready to adjust it, when the boss waved me off.
He later told me, he had done something like that once before, and the customer canceled the sale, because his old TV was working okay.
It does takes guts to admit to criminal wrongdoing in public!
Statute of Limitations
I don’t understand. Why ever would they want to open-source it? In the case of bankruptcy, I would agree. But open-source it, only because you don’t want to pay a monthly subscription??? Are you living on another planet?
“It states that if you use licensed software you must publish what parts are opensource that you use.”
Maybe YOU are the one who needs to read the MIT license, it doesn’t say that. Here is what it REALLY says:
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
It’s not “what you use”, it’s “persons to whom to whom the Software is furnished”, so if you don’t “furnish” the software to anyone you don’t need to release the source.
Says you who is too big a coward than to use this name. Why don’t you just run along and write graffity on the walls like the rest of you anonymous morons. Or does the X just mean kisses.
where does it say that posters have to use their real names? Why don’t you just run along and learn to spell.
“Mr. X”, later unmasked as George Eastman, was the anonymous donor who paid for the MIT main buildings so that name is used with reverence.
Software is furnished to the persons who buy the product that the software is used in. All sorts of companies have been pursued legally over not offering source code, when including open source software in their products. Stuff like set-top boxes. There was a big kickoff over Tivo, the details of which I can’t remember.
My favourite Mr X was Homer Simpson in that one episode.
Why can’t existing, proven, free cloud storage services be leveraged instead of each gadget maker developing and having to support their own cloud infrastructure? Why not just have their gadget require the end user to enter in their credentials for Google Drive/Dropbox/etc., where the gadget can then dump/pull data to/from a folder it creates within that cloud drive? These services are tried and true, they have solid security and data integrity protocols, and most users already have at least one set up.
I understand that continuous firmware & app development costs money. But citing cloud storage of user data as a significant monthly cost is pretty ridiculous these days (especially since we’re talking a relatively small amount of data per user).
This isn’t about storing data, which surely wouldn’t be much for light bulbs. It’s mostly about voice recognition, which needs fair CPU power, but also whatever the app and “automations” are. I suppose in that case the computer provides an interface for a gadget that doesn’t have much display or buttons.
The voice rec would be the only one couldn’t trivially be offloaded. An app doesn’t need to be continually updated, once it actually bloody works. That could do all the rest on the phone. Actually the phone could probably do the voice rec too, but I dunno what Google’s policy is on commercial companies using their service for that. Probably “no”. Or else a “yes” that involves more messing about than just doing it yourself.
And companies just love to have their users tied to them through the Internet all the time. Even when it serves no apparent purpose. I think investors and shareholders just like that shit. Maybe it reassures consumers too, even though it shouldn’t, makes them think somebody cares.
Anyway… it’s not about data storage. Though you can buy CPU power on a “cloud” basis, it’s not in a standardised form, and again involves contracts with corporate behemoths. Generally the same ones we’ve mentioned.
Just sticking a bloody ARM chip in there and having the user train it for a few simple voice commands would remove the need for all this fuckery, but business likes fuckery, it’s another way of extracting money from people.
You’d think the company name would make their intentions clear, but a lot of people bought a car called the “no va” so it seems there really are a lot of suckers out there.
What they should have done is make something that doesn’t need to be connected to the Internet all day just to function. Incidentally that’s also my policy in buying anything. I still have the old, manual, lightswitches and my fingers haven’t fallen off yet.
IBM managed to get Dragon Dictate working on a 486-33. It let you write documents with speech at just-about ordinary speaking speed. And had a full vocabulary of words. Why can’t modern computers do the same? I’m sure they can; in Google and Amazon’s case, they just like making us depend on them. Though only for fairly silly features we don’t need.
Maybe it keeps down on the tech support calls, having it in-house, and I suppose having a giant database of everyone’s voice helps it understand speech better. Still I’d go home-made if I was bored and rich enough to bother with all of that.
Maybe you could set up a Ras Pi type thing, with it’s own Wifi network. Have the hub connect to that. And it spoofs whatever Wink does. Would be nice if Wink could at least give us the protocol if not the actual source code. But since they’re now on the gravy train of $5 a month forever, why the hell would they? Even if they were going out of business, all that nonsense would still have commercial value, and giving it away for nothing might be seen as illegal by whoever supervises these things.
Still, it’s not beyond hackers to reverse-engineer it. I just wonder how many hackers are daft enough to have bought this to start with, particularly since, as you mentioned, they apparently had no plans for supporting the voice-rec software it would perpetually need. Unless they did have plans, and they were “charge $5 a month once there’s plenty of customers on the hook, but don’t tell them at first!”
It is possible to use your wink hub 1 with Openhab as long as you have it rooted. That will be the path I pursue.
I wonder if it will be possible to use local control after this change out of you will be extorted for the $5/month for that feature as well.
Death, taxes, and will.i.am got to get his money…
I must be prescient. About 2 months ago I became fed up with Wink’s constant platform outages and the increasingly regular need to disable and re-able the Wink skilll in the Alexa app to restore voice control. After some research, I purchased a Hubitat smarthub on sale for $99. No sooner had I finished moving all my devices over to Hubitat when I got Wink’s notice about the subscription fee. While there is no doubt that Hubitat is less polished – the mobile app is garbage – the system is light years faster because it allows local control. I can’t remember the last time I saw an update from Wink that added device support or functions. I have applied at least 3 updates to the Hubitat in as many weeks – each adding multiple devices, features and fixes. I’m still learning how to exploit all the features but I couldn’t be happier about switching now.
Definitely going to have to check out Hubitat.
“the cloud” is done entirely wrong, because these companies want to exert control and physically have access to devices… when in fact, the devices should run more like FOH and torrent… where your computing process, as well as everyone elses, is what serves up the actual hardware to make the computations. This would satisfy zero downtime, and zero monthly cost. But humans are greedy and inept. Decentralization is the best and least used idea in almost everything that humans do. Cept blockchain, but even then, they have centralized processing platforms which need to be done away with if you ask me.
as of right now, ‘the cloud’ is just a synonym for running it on a a centralized set of servers. when in reality, clouds are highly decentralized and dependent on no single factor.
Humans are stupid. “the cloud” will never succeed until they learn to write software that is self sustaining across every machine and platform with encryption.
same reason solar farms suck a bag of dicks…. solar should be widely distributed due it’s major downfalls and environmental hazards when put in huge mono-culture installations.
I don’t think that someone who worked for 2 years developing a device and its software, is greedy when he wants to pay his debts that he surmounted in those 2 years…
Always the same. People chopping off the heads of other people who stick out their necks to do something that those same people really, really want, but don’t really, really want to pay for.
I have been a big proponent of Wink since the first generation hub. I always thought their mobile app was the best designed and easiest to use (it still is). I was VERY happy with the 2G Hub and the expanded radio functionality. However, there have been an increasing number of issues and no meaningful updates or fixes for 12 to 18 months and out of the blue we get a demand for payment? If this had been floated when support was good or were we given time to transition, the backlash would be much less. This is just bad business.
My hope is that this push for a service fee leads some more adventurous folks to implement custom opensource firmware for the Wink 2 hub. That’s beyond me for the time being, so I’ll probably move to a Z-wave USB stick and one of the open source options in the meantime. But that Wink 2 is a beautiful box with lots of radios, and it’d be nice to see what it could do unhindered by Wink Co.
Why would anyone want to hand their trust to a stranger? If you can’t run it locally then don’t run it at all.
They same the same for ‘real’ money (i.e. gold) – if you don’t hold it, you don’t own it.
Root it, change the IP address it uses to contact the Cloud service. Reverse engineer the communications. Write your own Cloud service in Python for that Raspberry PI 1 that’s lying at the bottom of your box of obsolete parts.
Keeps your stuff running, and you’ll finally have that privacy that you thought you had.
And do this now, before they go bankrupt, because it will be a lot harder to reverse engineer the communications if their services go down.
I’m not keen (will not) be forking over USD$ 5 CAD$ 8 to Wink to continue service. I bought the Wink HUB because there WAS NO MONTH TO MONTH fee. I have a stand alone app for my garage device and a stand alone app for my lighting, but I do not have a stand alone app for my Kiddie “KN-COSM-B-RF” Smoke / CO detectors. Does Kiddie have there own protocol? Agree there any other free hubs that support the Kiddie devices? Thx.
I have the same issue, which is the Kidde KN-COSM-B-RF” Smoke / CO detectors. As far as I know, Kidde used their own proprietary protocol for these detectors and Wink was the only licensee to get the specs needed to interface with it.
That’s the only reason I purchased the Wink Hub – to allow those Kidde detectors to report to my smartphone if the house was on fire. Now that Wink has turned to extortion, I’d rather invest in an open source home automation hub that I can host inhouse and totally control myself. (A Raspberry Pi can probably handle all that I actually need, and while I’d love for it to monitor those Kidde detectors, I’ll be okay with switching to some other brand of Smoke/CO detectors that uses a standard protocol.)
As an IT professional with 40 years in the business, I have observed the drift of IT from secure data in data centers only accessible from dumb terminals to cloud-based client/server everything that is a swiss cheese of security holes, remote single points of failure, disappearing hosting services and insecure code running in our browsers. The general public has no idea how inherently insecure computing has become, despite the practically daily litany of well-defended enterprises that get hacked successfully. (If they can’t protect their infrastructure, then we don’t stand a chance – we’re just pretending to ourselves that we can so we can play with our digital toys & not be appropriately paranoid.)
Rock solid home automation for the 21st century needs to be open sourced, in house, and not cloud dependent. It cannot fail because of anything happening in the outside world, outside of a total prolonged power outage, a nuclear war or a zombie apocalypse. Cloud-based home automation services that are “free” are a come-on to build marketshare – the subscription fees will come at some point in the future when you are the most dependent upon your HA Hub and the vendor either gets hungry enough to make their move or sells the company to someone who is. It is a business model that preys upon our short-term thinking & our techno-lust for “solutions”. Our consumer brain knows that, eventually, we’re going to get screwed.
Gee Elloit what a nice article and commentary on the poscast, but thinking it costs $5/mo to provide IoT services??? That’s like half a NetFlix that provides video streaming and pays royalties on content. Handling a gateway is mostly just message passing of a few tens of bytes. For video you just need NAT punch through, so to enable phone to video doorbell or whatever is still just a message.
So how much does such IoT gateway servers cost? How many servers do you need? Suppose you need a message per minute (seems high to me) at 100B/message. So 6KB/hr, ~100KB/day, 30MB/mo. Say you have 1M customers, then the data transfer is 3TB/mo * 2. How many servers are needed? At 20K messages/sec (1M/50) for routing, that’s within the ability of one dedicated server. At $.15/GB transfer fee, then it’s costing $300/mo for data transfer and say $100/mo for 10 servers. Round up to $500/mo per million customers, or $.0005/mo. less than a penny per year per customer.
If they charged $5/year (people might pay that) subscription they would be way ahead.
As you say, they’re really not dealing in that much data. They probably have software support for the phone app and stuff to pay for too. Still…
My guess is that they looked at the other side of the market — how much folks would be willing to pay — and did their work on that side. I bet people in the off-the-shelf home automation bracket are enthusiasts and used to spending money on their stuff. If they keep just more than 20% of the customers they would have by charging $1, they win, and my bet is that the difference between $5 and $1 matters less than 80% of their market.
They’ve got a monopoly on software that can run on the device, and they’ve got an essentially captive audience. They’re going to try to milk ’em.
Everyone needs to refuse the subscription and promote other alternatives. Spread the word to cripple this slap in the face.
I feel like they should have tiered services… so the whole thing isn’t bricked! Like offer some types of devices can connect free then charge for others… or you get so many devices to connect for free then you have to start paying monthly to unlock more. Then people would be less angry because their device isn’t useless at that point. It just does less.
When mine stops working, I’m going to take it to the road and sledge hammer it for YouTube.
My hub just sent a notification that they have decided to pause the subscription…
Mine as well, light was still blue when I left the house this AM
The deleted the original blog post, and the subscription info page now says this:
“Will I be able to access my devices if I don’t subscribe?
If you don’t subscribe you will only have access to local control from the Wink app for switches, lights, and locks that are already connected directly to your hub. You will not be able to control your devices when you are not on your local WiFi network, use Alexa or Google Home or any API devices.”
As someone who doesn’t allow remote access to my home network, and doesn’t trust Alexa or Google home (for basically the same reason), this seems like a reasonable way forward.
On second thought, the “already connected” and “or any API devices” parts are a bit sketchy. Does that mean existing customers can’t add new devices, and can’t control existing devices through anything other than their phone?
As a (former) Wink customer, yes, that’s exactly what it means. No adding new devices, so you better hope those Zigbee bulbs last a LONG time. Because you won’t be able to replace them. It’s really a shame, because the hardware is neat.
I suppose you could pay the $5, add your replacements bulbs, and then stop paying. But $5 every time you want to add a bulb? And why do they require a subscription to pair devices? shouldn’t that be a “local-only” sort of function?