Ah, that reminds me of my Westinghouse Radio Hub, an Internet connected radio, cd, weather, etc. I was excited to buy it on sale, then packed it in a closet and forgot all about it.
A long time later, I accidentally stumbled upon this puppy right there in the closet. Took it out of the box, and to my first disappointment, it wasn't wireless. You had to plug an Ethernet cable.
Hit the power button and the screen lights up with the word, initializing.
When you have waited long enough, the screen changes to Network error. I tried plugging it in directly on the modem to no avail. I tried to debug it by analyzing the network, the domain it pointed to was no longer resolving.
No internet feature for me I guess. At least I still had a Radio and CD player....
Let's tune in to the radio. Automatic station search. Apparently, that requires the internet. So after you wait long enough, Network Error.
Well let's go with CDs then. When you insert a CD, it looks for the album info online. Oh you have no connection? Network error it is.
Left it on my desk as decoration for a couple years before I threw it in the garbage.
Some houses we "grew" up with had phone sockets in every room, the modern equivalent would be an Ethernet socket.
And we should also work towards making that a standard, too (i.e. like wall sockets are standard for power, Ethernet sockets should be for data). I'd agree with the notion that "smart" (I'd prefer to say home automation) devices are the way to go, but highly disagree with everyone "Oh just key-in your Wi-Fi password, and good to go!".
Air is a shared medium with inherent collision problems. Avoid it at costs that slightly exceed convenience.
EDIT: Also, home automation. In most people's lives, one's "home" is one of the most static things in life. If anything, your home would be worth spending an additional thought or two on proper planning, to avoid headaches that'll haunt you later.
I know just about everyone in my circle, in tech, filled the walls with smurf tube when finishing basements and any other wall they could get away with.
Not in the houses I've been looking at. Out of dozens I looked sty built within the last 10 years, only 1-2 had it. And the current owners weren't using it.
AFAICT, They're built with the presumption of wifi.
That's weird. It's not expensive to wire houses, and you can't get gigabit fibre speeds with ethernet. You certainly couldn't get anywhere near it 5-10 years ago, but gigabit fibre was available then.
You can use "fish tape" to run wires in walls, mostly non-destructively. Many fish tapes glow in the dark so you can drill 2 small holes, do what you need to do, spackle, and paint.
You probably still wouldn't do any of this in an apartment of course, but I figured I'd throw out the pro tip.
Vertically, sure, but not really horizontally. Aside from just going through a wall, I'd expect most ethernet runs to be more horizontal than vertical, since you're often running it between rooms?
If you have forced air heat or central air conditioning, you probably have a reasonably direct route from your basement/attic to any given room. Just use plenum rated cable, and it's generally pretty straightforward.
Cutting in the box and making up either end of the cable generally takes longer than running the drop.
I rented an AirBnB filled with a few thousand dollars worth of Sonos speakers. Some issues...
1. I was traveling with an older Android phone (Android 4.4.4). I couldn't install the latest version of the app. I tried installing an older version, but it simply said I needed to upgrade. So, I couldn't play any music from the phone. I checked online and a number of people were complaining they used their older phones and tablets in the house as simply a Sonos remote, and they now had to upgrade all of those devices after support was dropped.
2. I installed the Windows app on my laptop to play music. This worked, but then I noticed the subwoofer wasn't plugged in (likely so people didn't annoy the neighbors). I thought I would briefly test it out to get an idea if Sonos would be a good future purchase for myself. Well, I plugged it in and I needed to do a setup process from the app. Well, apparently that can't be done from the Windows app, and I couldn't install the app on my phone, so the subwoofer just sat there as decoration.
3. My partner came home and she had a newer phone, so I had her install the Sonos app so we could setup the subwoofer. She installed it, I went to the app to configure the subwoofer, and it said it couldn't be done unless I updated the speakers to a newer software version. Well, this wasn't my apartment and the owner lived there when it wasn't rented, so I figured it wasn't my place to update their speakers.
In the end, I never got a chance to test out that subwoofer, and I lost interest in Sonos because I felt like for the investment, the rug could be pulled out from under me at any time. I hooked up some of my 15 year "dumb" Klipsch speakers in my new apartment instead. Didn't need a new mobile phone to set them up, didn't need software updates, just plugged them into the wall. It sure felt easy. I can't help but feel like we lost our way a bit in recent years.
"Sonos's “recycle mode” intentionally bricks devices so they can't be reused"
Fuck this kind of masquerading profit hunger as environmental responsibility.
Just buy some wired, maybe simple Bluetooth speakers for cheaper (eg. Sennheiser, Audio-Technika or JBL) and enjoy their fine sound.
"Recycle mode" is intended for devices that people are trading up to newer devices. In return for enabling it, you get a 30% discount on any newer hardware, the idea being that you either ship the item back to them, or take it to an e-waste recycler.
Nobody is forced to do this. You're more than welcome to sell/give/etc your speakers to someone else.
The thing is that e waste recyclers generally try to re sell working electronics and have come out hard against Sonos's recycle mode.
There's a reason why the saying is "reduce, reuse, recycle" in that order.
They also don't verify that you recycled it, practically it only makes sure that the otherwise working speaker is removed from the market so that Sonos isn't competing with itself. I'd imagine that a lot of these just end up in the garbage.
Sorry for the late reply, I agree with the others.
If this truly was an environmental initiative, Sonos would rather take a proof of receipt of you giving your previous speaker to an e-waste recycler or even giving away to some place like Goodwill where it'll get reused.
They could just make sure that you were nice and gave away your device for free and/or charity purposes, instead of bricking it forever.
Controlling your home solely from your phone is overrated. I installed a smart home system and one of the first things I did was buy a cheap tablet that could control everything. I always have my phone on me for work but I don't want to pull it out everytime I need to do something. And when the in-laws are over they need to be able to do stuff. So does my 5 year old that obviously doesn't have a smart phone.
I have a Denon receiver. It has some nice connected features, like Spotify Connect and Airplay, but it’s also a completely functional dumb receiver that will work with equipment, input and output, from decades past and future.
I buy my own speakers and plug them into it. It also supports some Denon wireless speaker standard that lets it hook up to other Denon receivers and/or any speakers with the special adapters attached. But the crucial thing is that at the end of the day, all that fancy pants crap can stop working and you still have a bunch of bog standard speakers and receivers that can be hooked up any other way.
Now I’ll concede that maybe I’m ignorant or misinformed or a grumpy stubborn old man, but I’m not sure why anyone technically competent would buy a Sonos or Bose system with their crazy lock-in. It wasn’t so long ago that Sonos made you use Spotify through their awful reimplemented UI. It seems they have relented on that front and now allow use of Spotify Connect, but why would you want to be at someone else’s whim like that? Bose uses non-standard wiring and electrical specs to make sure that you can’t use their speakers with standard receivers. You either have to trash the speakers or keep buying Bose receivers when you want new features (4K, 8k, Atmos, HDR, etc.) Even when money is no object, I feel like a sucker when I buy artificially locked-in products like that. Maybe others don’t feel the same way?
For the not-so-tech-savvy, I concede that home audio gear is still a little difficult to set up. The common practice of making your own banana plugs is ridiculous and pre-made audio cables should be the norm instead. The receiver UIs leave a lot to be desired, even when they are ultimately good enough once configured. Shopping for speakers is like choosing fad diets: you have to sift through outrageous amounts of bullshit to find any useful advice.
I’d like to understand some more of the sociology and economics of why Denon and Yamaha and such can’t get their act together to produce something as user friendly as Sonos but that degrades nicely into standard A/V equipment when desired. If anyone with industry insight sees this comment, please fill me in.
I plugged a $15 Bluetooth receiver into my 11 year old stereo receiver and it works pretty well. The 30-pin iPod connector built into it is not seeing much use these days however.
I've got this tiny device that plugs into those 30-pin connectors to make them Bluetooth, really nice solution to utilise redundant tech (originally bought it for a 2015 Maserati that couldn't play music through its Bluetooth sigh talk about shitty consumer experiences).
But yeah, I'm a massive believer in keeping things un-integrated. Need a feature, but a component and plug it in
>I was traveling with an older Android phone (Android 4.4.4). I couldn't install the latest version of the app. I tried installing an older version, but it simply said I needed to upgrade. So, I couldn't play any music from the phone. I checked online and a number of people were complaining they used their older phones and tablets in the house as simply a Sonos remote, and they now had to upgrade all of those devices after support was dropped.
Reminds me of a similar frustration with OS X. In December 2017 I was trying to use bluetooth headphones with a MacBook that was still on Mavericks or Yosemite (10.9 or 10.10, latest version at the time 10.13), so yeah, a little behind the times.
But when I tried to connect, it said I had to upgrade the OS to use bluetooth with those headphones. What? I'd connected to my phone with it with no problems, and the protocols for the connection are like 10 years old, and almost certainly existed in that version of the OS.
I have a system from Sonos. Their app to set-up the wireless speakers adaptive sound feature didn’t (at the time - not sure now) work with the latest iPhone. I dug an iPhone SE out of a drawer to set it up.
The semi-future is that some of the Sonos gear supports Airplay 2 (and possibly some Google format, but I wouldn’t know.)
I only ever use them with Airplay 2, and they are firewalled off from the Internet. I also don’t have the Sonos app installed on any of my devices, so it can’t brick my speakers and insist on my signing a new TOS to get an update I didn’t ask for to provide features I didn’t ask for.
I fell right out of love with Sonos the first time they pulled the “your devices will have limited functionality until you agree to our update TOS” stunt, especially because that TOS seemed to be all about agreeing to even more data collection.
Then they stated shipping speakers with microphones, and it was clear which direction they had chosen as a company. They aren’t a hardware company. They’re now a surveillance capitalism company that uses hardware to collect data.
I bought an Apple HomePod, foolishly thinking it'd have some hardware inputs as well. Obviously that's not the Apple thing to do these days. But it has Airplay 2, so I figured there'd still be a way to interface with it easily. My macbook worked fine, but there was a 2 second delay for every input unless you used Siri.
Funnily enough, my Windows computer streams to the HomePod in real time, so now it's a pretty overpowered gaming speaker. But at least that's all it does; there's no bespoke Apple software that has to be installed and updated. And it's fairly dumb in terms of what you can manage with it (none of this stuff like yelling your shopping list at it.) Before I found that particular software, I was about ready to return the thing and switch to a Sonos... until I remembered what they were doing with legacy devices.
It would be nice to know if there's a way to build anything custom for it at some point though, since it's just a massive speaker with iOS inside it.
If you ask me, there should be regulations about providing a reasonable fallback when selling physical products that depend on some internet service to work. If the service is shut down, the device should be made functional without it (or with a custom thing you can host yourself).
Well, when I first set my speakers up, they used the Sonos app and my phone’s microphone for setup.
I don’t believe the speakers I own have microphones (although there is probably some way to hack the speaker into detecting sound, after all... It has a coil, a magnet, and a diaphragm that moves...)
And I’m now extremely wary about buying devices with cameras and microphones, especially for my home.
"...only FOSS survives"
But yeah on of the many learnings of the long now foundation is that for things (wind Mills, businesses, etc.) To survive over a thousand years they need a community to maintain them.
Netscape died and yet we still have Firefox.
Editor have come and gone and Emacs is still here, and I'm sure it will be here in 10 years.
Not to worry: once there is sufficient regulation in a given product category, said regulations may eventually mandate that products refuse to work without server connections and kill switches. You know, for consumer safety. Or to combat wrong information. Or whatever it is that decade.
I acknowledge that this is a huge problem, but I can't help but feel like this is a massive opportunity for a competitor to charge a premium for a saner, more consumer friendly business model. So it might take a couple years but I'm not convinced yet that regulation is necessary - especially considering said regulation in the long run can have unintended side effects and chances are because of lobbying and regulatory capture it won't actually benefit the consumer much, if at all. Legislation is much more difficult to overturn than shitty market forces if consumers are savy and willing to vote with their wallets.
There was a good point raised about this very idea on an earlier thread, which resonated with me based on other things I've read.
Basically, small and medium sized family farms, (i.e. the hearty, resourceful folk who want to fix their own damn tractors, thank you very much (e.g. my father and grandfather)), are disappearing.[0] They are being replaced by large, corporate "factory" farms, who have little interest in doing repairs on tractors that go beyond basic maintenance. Rather than building a staff of what are becoming increasingly sophisticated (i.e. expensive) technicians, along with the necessary facilities that meet all safety regulations, etc., a simple "tractor service" is preferred.
So, the customer base of the "consumer friendly business model" may not be large enough to support such an endeavor. The times, they are a-changin'...
I don't see any reason to be optimistic about consumers guiding the market. It seems like everything points to cheap, ad-filled crud winning out over premium "ethical" alternatives all the time. Apple might be a counter example, but I think many would debate that.
I think the problem is you get crushed by the incumbents before you have a chance to take off. John Deere has lots of infrastructure (manufacturing, service) already in place. In a different market, see Google trying to break into the ISP space.
They had plenty of expertise but an extremely hard time securing utility pole right of way, IIRC, to the point of being sued my incumbents to stop them providing service.
I test drove a Tesla. That thing is constantly calling home and it's only a matter of time before it starts snitching on drivers who break the speed limit, run red lights, or park incorrectly.
There's a story someone wrote in a comment or otherwise posted here about Tesla telemetry identifying an employee who crossed one of the SF bay area bridges at 130mph
Legislation requiring them to do so. If they have the capability, eventually law enforcement will start asking for the data. If Tesla don't provide the data, you get a lot of headlines about how police are unable to solve hit-and-run cases etc. because they're not getting data that's just there for the taking. Cue legislation.
The difference is that with FOSS anyone in the world with the skills and the disposition can help maintain a thing. With closed systems you are always at the mercy of a single bean counter somewhere.
And why do these things need support? If you build them with open standards and protocols they will continue to work until they break, no support required
Even if Sonos went belly up you’d still have airplay and tv/line in, along with direct control from stuff like Spotify (until Spotify inevitably change their protocols). But I don’t see airplay ever quitting until the hardware does
Does AirPlay not require software updates or certificates? There are a lot of things which are fairly robust but still fail when things like certificates or NTP servers disappear.
No, my first generation airport express from 2009 still works as an AirPlay device from the latest iOS. Sure you can’t configure it from latest macOS, but airplay still works. I can’t recall when the last software update was, but I haven’t done any updates for at least 3 years - it’s unplugged most of the time.
To Apple's credit, they have released fairly recent firmware for all their airport devices, even adding Airplay 2 support to (either 2nd or 3rd gen?) Airport Express' that had the hardware to support it, despite the devices not being mfged anymore (for many years) and the entire product line discontinued. Fwiw, I just used a 1st gen Airport (>$20) as a wifi to ethernet bridge to get a GoldenTee machine online for a client. They also work great as a wireless airplay receiver for older stereo equipment using the 3.5mm output from the Airport!
Huh, I saw this comment 5 days ago, and noted the name "westinghouse". Today I was researching an "ACME.eu" "cloud" camera, discovered it contacts ubia.cn, which is specified in a listing on the FCC website by Southern Telecom for a "Sharper Image" camera that is my exact model just with a different brand name.
Apparently, Southern Telecom has bought some American brands and sells spyware "cloud" devices to end-users.
I still haven't found a way how to watch the camera stream without the app. It's possible, but it's a UDP protocol that I do not understand. It works with or without Internet, so I'm sure I could watch it without giving it Internet access...
I was extremely skeptical of this, so I googled it. Seems like Tesla owners have had no problem starting their cars without an internet connection, though obviously the internet dependant features don't work.[0] This doesn't seem comparable to a CD player that won't play CDs without an internet connection, it's more like a CD player that gracefully degrades to not showing the extra info it would get from the internet while still playing the local media.
Did you try using keyless start from your phone? I’ve relied on this before when I forgot my key but I’d worry about lack of service making it ompossible to start my car.
It is not a lie. The idea is that Tesla would go out of business and a vulnerability in the car would be found and exploited to kill you in the absence of security updates. Similar to how the article says:
> We're guessing a Charter alarm would still be able to make loud noises when someone breaks into a house, but that doesn't mean it'll work with an alarm-monitoring service.
What lots of assumptions? that's a single inevitability that a self-driving car manufacturer, Tesla or otherwise, will go out of business, rendering the car offline only (the point made in the article) or worse, like the case with Android phones, you'll have plenty of vulnerable cars on the street and some community updates (like the work of LineageOS for instance).
It's the same point made in this computerphile video[1].
To add to this, a good idea would probably be either a physical "allow access" button on the device (think like the clicky button on top of the Philips Hue hub), or a secure random password physically printed on the device that's the default access password.
Think carefully and lock-it down (not as an afterthought), because unsecured IoT can be exploited to cause devastating harm.
Phoning home should be disclosed and disable-able because you're opening up an attack surface and a privacy risk. And, if there's no clock synchronization, it will be impossible to check the validity of X.509 certificates used in https://.
You should support APIPA and Zeroconf.
Bonus points for Bluetooth to allow configuration from a mobile app. The mobile app should also be able to securely download and send a firmware update.
ZigBee is also low-power and mesh oriented, but will need a bridge/gateway to the internet to be connected.
Device pairing like NetVue cameras having QR barcodes and/or Bose headphones with NFC might be Good Ideas™ too.
What info are you using? If it's weather, let me pick a provider. If it's satellite imagery, let me update the domain if necessary manually. If it's something that only you provide, put it on github. If it constantly needs updating, find a place that will re-index it. Doi.org is a suitable solution for a lot of things, is there a case that i am not covering?
IMO, it's because most manufacturers' main goal with "smart" devices isn't making better products; it's designing products that are less expensive to manufacture, but bring in at least as much revenue.
A device that's controlled using the owner's phone (or a web application, etc.) is cheaper to produce in bulk than one that has a built-in interface to interact with.
A security camera system that stores all its data in cloud-hosted storage is also cheaper to produce in bulk, because it doesn't need to include storage hardware of its own. That has the added benefit (to the manufacturer) of tying the customer into ongoing service fees.
In other words, while I agree with you that they're bad design decisions from the consumer perspective, they're good design decisions (at least in the short term, which is all most corporations care about now) from the manufacturer's perspective.
Unfortunately, at least for a lot of manufacturers, the decision process is going to go something like, "Should we make sure this can operate even if that domain goes offline?" "Nah, if that happens, the company will be belly-up, so why should we care?"
(And yes, that is assuming that they'll never accidentally—or intentionally—fail to renew their domain name...)
For now, because it still sounds cool and futuristic, but if the whole audience of HN is already rejecting smart devices wholesale, that's a bellwether the industry might want to consider before the attitude percolates out to public at large.
Or...there are engineers who realize this regardless of hacker news and push better design factors by communicating customer needs and impact on branding. Hacker news is not as influential as we may believe.
Sadly, unless we as engineers, or ideally some broader body promotes some standard of device independence -- customers will eventually reject all such devices regardless of where they reside on the spectrum of cripple-able <---> independent.
For many of these products it was a very deliberate design decision, both in order to lock-in the user and also to be able to collect the data on them.
I literally just did a search for "Westinghouse radio hub" and other variations and this post was the first thing that came up, the rest was all about 1940s Westinghouse electric tubes and things like that. Westinghouse effectivly went out of business effectively in 1995 as they were bought by CBS, when the Gopher protocol had more users than the WWW. This just sounds like such an outdated argument or example it does not make any sense. Sure, network connectivity has disadvantages but refusing all connected products because of oulier examples is extreme paranoia.
This is why my smarthome strategy is, that everything must work locally (as in without internet access) - but can be enhanced by connectivity to the internet and cloud-services.
I use https://www.home-assistant.io/ on a Raspberry Pi 4 and I only buy devices that work locally, are either super dumb (think zibgee lightbulb) - or are either already open or can be flashed/hacked to be open.
That depends heavily on where you live. If the most common type of power outage is one that affects a small region, it's quite likely internet connectivity could remain unaffected. I had this experience in northern Sweden, where I had a couple of power interruptions but zero network interruption.
It is not just about power outages, it is also about cloud services closing, having outages etc - as in this post. If you only have devices that will work locally, that becomes much less of a problem.
Eh. Depends on how mission critical it is. I don't mind an appliance needing the internet as long as I can still physically operate it during downtime.
It's interesting to me that this is (effectively) one of the requirements to get Apple's HomeKit sticker: it needs to expose an API that will keep working with HomeKit even if the manufacturer cloud service goes away.
The fact so few smart devices are HomeKit enables says a lot, unfortunately.
It’s this type of garbage that has turned me into a luddite when it comes to smart home type technology. Lack of interoperability, inability to understand easily whats actually going on and massive privacy invasion causes me to look for dumb devices whenever possible - I know how to deal with them, I can have other people fix them when something goes wrong, and there’s just less ongoing maintenance. I’m sure this isn’t a novel idea though.
While the interoperability of IoT devices sure does suck there a projects like Home Assistant[1] which make this less of a pain my advice is stick to locally controlled devices that follow existing standards (ZigBee, Zwave, WiFi) and shit like this can't happen.
There are of course alternatives that have been around for decades like Control 4 but they cost serious money.
My criteria is: all the logic must be contained within my home, the devices must adhere to some open standard, they must be repairable and modifiable to some degree.
The Zigbee protocol the devices use is a standard, but from what I picked up from between the lines of the article is that the firm locked the devices because of liability fears. If someone DIY's their alarm system and then it fails to work during an actual home-invasion, they don't want to be sued.
Surely a reasonable solution to this kind of problem is to change the law if necessary so that businesses aren't on the hook for damages that weren't their fault? I suspect this may be a particularly US-based problem anyway given the excessively litigious environment that seems to persist there.
The idea that we're deliberately bricking useful technology devices because of lawyers is abhorrent. Realistically, their owners are probably then dumping them in many cases rather than sending them for any useful form of recycling, which is hugely wasteful. You'd think with the increasing awareness of issues like environmental protection and right to repair, we'd have been more enlightened by now.
it's pretty functional and thoughtful for the wealthy. instead we need a congress responsive to the long-term common interests of the citizenry and not beholden to the singular interests of the rich.
They should still be responsible for damages if they were their fault, though. Just like you shouldn't lose the warranty on your car because you get the tyres changed at an independent garage.
These standards won't be adopted by the market for at least 10 years, because it's not in anyone's financial interest to do so.
The hub manufacturers all want to lock people into a proprietary garden, even if they're using open protocols, so they can sell accessories and "Certified for X" marks.
The only thing that would really push adoption would be a killer paid service that ran on top of one of the open platforms.
Expect after 10 years everyone will be fed up with sunsetting and interoperability issues that standards-compliance will be a product feature.
But then again it took 20 years + the EU to nudge the market towards USB connector chargers?
Oh there are masses of zigbee enabled components that are perfectly compliant with those constraints. From big labels to Chinese knock offs. Sure the Sonoses of the world have much more visibility and don’t but who searches will find :-)
Indeed - ZigBee is quite good for this. I know from experience that a Philips Hue bridge will gladly pair with and manage non-Philips bulbs (e.g. I have some IKEA ones in addition to Hue), and similarly that Philips bulbs work fine with a non-Philips bridge. You can also block the Hue bridge's access to the Internet and it continues to work fine.
Sonoff devices don't do this out of the box, but with some basic equipment you can flash [open source firmware][1] on them and make them work even better than stock, and no packets leave the LAN.
I can't even imagine a scenario where I'd want a "smart home".
When I leave a room, I turn the light off. When I enter it, I turn it on.
When I leave the house, I turn the heat off. When I enter it, I turn it on. If I had pets, I'd just dial it down to "pet comfort" level while away.
Am I missing something?
EDIT: I do have a Roomba and think it's great. I hit the button when I'm about to go out for an hour, and there's a clean floor when I get back. I appreciate that. No internet required as well.
Here's my (unsolved) use case for a smart home: I've got three heat pumps and a whole house generator; if the generator is on, I don't want all three running at the same time (too much load), and generally, I'd like the set points adjusted, because I value saving fuel (the tank needs to last until the next filling, and I don't know how long the outage will be) over comfort. I've also got a hot water recirculator that I'd like to turn off when the well pump isn't powered (it's not on the same circuit as the house) or is otherwise not providing water pressure --- after a water leak, the previous recirculator burned its bearings moving air instead of water, and it was expensive to replace.
Also, I'd love to run the HVAC fans to equalize temperatures when some rooms are much cooler than others.
This type of scenario shouldn't require "smart" home hardware - Traditionally situations like this were addressed by wired/wireless SCADA networks or BACnet or similar protocols.
The only issue is that these things are typically designed to be engineered along with the systems they control, and are difficult for the average user to implement or use.
The thing is, "smart" home hardware is a lot less expensive and more accessible than SCADA or BACnet. I can go out and buy any number of home thermostats, and some of them are even capable of properly supporting all the functions; unfortunately I ended up with the Nest, because it has the best support for enabling both auxiliary heat and emergency heat; but it doesn't have support for interfacing via LAN, and Google rescinded the APIs. Also, the physical interface is infuriating, but it's better than the awful honeywell touchscreen garbage the previous homeowner had installed.
I have smart lights in the bedroom. They come automatically as a light-alarm and it's fantastic. It would be hard to go back to waking up in the dark and turning on a massively bright light.
It's also remote controlled at night so I can turn it off from bed. These are all useful features.
None of it is "necessary", but having my locks, garage doors, HVAC, doorbell[1], most of my media players, and most of my lights integrated with Home Assistant provides all sorts of convenience.
It all works "offline" (in reality, as long as the house has power, it all works, I just can't connect to it if I'm not home).
[1] The doorbell is a "dumb" doorbell that triggers an IO pin on an ESP8266.
Likewise. Last time we renovated, we avoided "smart" technologies like the plague. No regrets, and so far no obvious disadvantages either, but apparently many dodged bullets when it comes to longevity, privacy, security, etc.
I've had a startup business in the back of my brain for a few years that's driven by this syndrome:
1. It vendors would contract with IoTZombieCo to have their product covered. Like a form of insurance mixed with code escrow. They pay a premium for this and can put a brand label on the product that tells consumers it can't be EOL'ed. Vendor sells more product or can charge a premium price due to customer assurance of continued service.
2. My IoTZombieCo staff are given access to their code and ops resources (in the ideal scenario the vendor outsources running the backend services to us too).
3. When the vendor goes broke or decides to EOL, IoTZombieCo takes over the customer relationship, charges the customer a reasonable fee to continue providing service. Fee to be limited by some covenant that allows reasonable but not excessive profits.
That would solve "It stopped working when the vendor went out of business" but it wouldn't solve "It stopped getting security updates when the vendor went out of business"
Code escrow doesn't guarantee maintenance either, that would be yet another committment for the company and might increase their operating costs over time as the accumulate more and more zombie products to maintain, potentially leading to failure too.
Isn't maintenance what IoTZombieCo would be contractually obligated to do, though?
Here in the UK new homes often have a 10 year warranty from a third-party insurer, who gets to dictate construction methods to the builder. So the idea of products coming with third-party warranties isn't unprecedented.
You're forgetting that consumer products companies want customers to buy new products, not keep using old ones indefinitely. It's in their financial interest to deliberately EOL older products with no recourse for customers so long as the average customer doesn't wise up enough to shy away from the brand entirely. This is really digital planned obsolescence.
A better company would be one funded by donations that reverse engineers protocols for these things and provides an open server to keep them alive, similar to what the Pebble community did.
Without enough instances of people losing support at EOL, consumers won't put a premium on such a brand label and thus I do not see companies willing to pay the premium and contribute the labor needed to complete step 1. By the time that consumers are accustom to the point where they are worried about EOL support, they would likely be looking for systems that are more open, and that is assuming we ever get to the point where consumers care that much (they have notoriously short memories sometimes).
Such a business might have a chance at the point when the cross over from not caring to caring occur.
This idea isn't new. Companies in the B2B and industrial space typically contract out to other companies like what you describe to provide 10+ year support plans for products.
The trick is that for a consumer market, the margins are typically much harder to work with -- especially with startups.
In fact, there's a zombieco that took over running Vector's upstream servers when Anki died, so definitely not new.
I think the most important hire for this startup would be a brilliant marketer. It would be an uphill battle because a vendor isn't going to want your service if their customers don't even know to ask about it. It would be 99% educating people about this service so they can look for your logo on the box when they buy their next smart doorbell.
Right. Not having an anchor customer is what's prevented me developing the concept. I think it would be best done by starting as a service outsourcing provider then adding the insurance component.
Well, I'm involved with a project that takes that approach, so...agreed :) But realistically vendors mostly won't do this and customers mostly won't be able to run their own back end. And there are often dependent services needed (weather data for example).
Well, even if everyone doesn't want to run their own back end, if it's at least possible to do so the chance of someone offering to run that service for a small fee is infinitely greater than if not.
I just bought an OpenEVSE electric car charger and this is what it does. They run a service for it, but if you prefer you can change the url and run the service yourself.
I agree with the pain point to be solved but let me phrase this pitch another way:
IoTZombieCo: give us x% of your revenue and we’ll acquire you for free when you die. In exchange, you can put our logo on your box as if we’ve already acquired you.
Companies don’t shut down for fun. Usually it means the economics didn’t make sense. Having a large portfolio of failed products wouldn’t change the economics for IoTZombieCo.
The only long-term solution is products that aren’t cloud addicted. Make everything local-first with some progressive (optional) enhancement with cloud functionality that users pay for to keep it sustainable.
Disclaimer: I build local-first IoT products at Hiome (hiome.com)
But unless IotZombieCo can ensure it too doesn’t go out of business isn’t it just the same risk to the customer however with 2 companies instead of one? I think that 2 companies with the ability to keep a device running is better but the real solution is fully open sourcing the code of these devices since the company is no longer going using them or supporting them. Instead the company is offering a disingenuous offer where you can get some Ring equipment for a discount but really that discount is a minuscule fraction of what one would pay just going to the store and if I were to guess they probably made some back room deal with Ring as well that they could convert ‘x’ amount of customers to their service and got some sort of back pay. This was just a shit deal but it is good it happens from time to time because we need people to start to become aware of this and demand open source.
1. IoT device stops working when the vendor drops support.
2. IoT device stops getting security updates and ends up on an DDOS botnet when the vendor drops support.
If you want to solve 2, whoever takes over support needs control of the auto-update mechanism, something to deter them from abusing that control, and no unpopular devices or cheapskate customers to fall through the cracks.
Turtles all the way down. But the difference is that IoTZombieCo has an obviously viable business model as long as consumers want their devices to work and are willing to pay the cost. Many of the original vendors don't have viable business models which is why they go out of business.
There are many devices used in business scenarios for example where the consumer is likely to take a keener interest in long term viability and also has a higher tolerance for paying for that.
"as long as consumers want their devices to work and are willing to pay the cost"
This is what is upsetting the current customers. All they have to do is buy a Ring device and they essentially get their service to continue. Or you propose they continue to pay IotZombieCo and they get to have their software run on their device. So either they pay for the device or software in order to keep a service that worked perfectly fine before, and really does not need some other person's servers controlling it when a person could set that up on their own network (not against cloud services as they can provide added value but the ability to run on your own network should be an option too).
That's expensive. Much more expensive than basic firmware updates and a couple servers from a system that by design is going to put a priority on keeping the cost of firmware updates low.
The manufacturer should never be able to either deactivate or otherwise disable the functionality of a device they sold in a form that requires a service to function. If the customer is paying a subscription fee and no longer has to pay the fee when the service goes away, that is maaaybe fine (albeit opening a loophole that would most likely be abused (from a customer's perspective).
If the service was priced into the selling price of the device, game or other software though: nope, never. You did the math, you sold the device for that price, now keep on delivering.
That would not necessarily mean that they would have to run servers indefinitely, it could as well be in the form of releasing or open sourcing the related server software or the specifications (and required certificates for signing updates if applicable? That might be thorny though.) in a form that someone else could successfully take their place when they end-of-life a device.
Ideally in a form that would also work when the company closes shop, that is when no longer any form of workforce or capital for anything might be available.
It might be time to actually require that from (also software) manufacturers since there seems almost no change in that regard.
> The manufacturer should never be able to either deactivate or otherwise disable the functionality of a device they sold in a form that requires a service to function.
I would go further: IMO, the manufacturer should never be able to sell the device that requires a service to function, period. They should only be able to either a) sell the device that can be optionally tied to a service - not necessarily manufacturer's - but otherwise works fine without it, or b) rent a device as a part of a service, remaining responsible for its maintenance.
Model a) is how cell phones work - you buy a cell phone, a cellular connection is a separate service provided by multiple competing parties that are not the manufacturer. Model b) is how ISPs tend to work - they lend you hardware to connect to their network infrastructure, they replace it if it breaks, they take it away at the end of the contract, and they absolutely refuse to let you use their hardware with a different ISP.
> Ideally in a form that would also work when the company closes shop
You typically already have to provide full hardware schematic when submitting an electronic product to test labs to get various certifications.
It doesn't seem that far fetched that you would also have to submit server software and the certification house would keep it in escrow in case you go bankrupt.
I don't think it's realistic to expect that will happen though.
Code escrow isn't unheard of in B2B software. Maybe enough people will be burned by companies bricking their hardware that it will happen with B2C, too?
What happens to domain.com once $business is now $noLongerBusiness?
You can provide a method of flashing firmware onto the device itself (possible security risk?), but where do you get the files from? How do you verify they're safe?
Then, you need to host the server software itself (depending on how it's coded, that ranges from quite simple to your worst nightmare), and you expect someone to keep that secure? The EU dictates GDPR compliance, which adds additional cost to maintenance.
It's just easier and cheaper to just not bother with servers if the company can't keep them online.
Let user change server domain. Firmware must be signed by certificate with that domain. So now user can change server domain to his own server (or someone he trusts) and sign firmware with that key. It poses no security risk.
That's a very idealistic but unrealistic perspective. Often times the source code can't or won't be released - it could put further customers at risk if a vulnerability is found and the update servers have gone away.
Some devices might not be possible to update due to hardware/software configuration (perhaps certain variables are hard coded?).
Whilst what you're saying is right it just doesn't work that way. I'd love for all the Android devices to get years of OS and security patches, but it simply doesn't happen.
Any IoT device is pretty much the same. It's not going to change any time soon.
Only invest in products you can be sure of (which is hard these days). Expect them to break, expect nobody to care, and be prepared to lose money or get a terrible ROI - it's why phones don't get updates, the OEM doesn't make money from providing updates which are a direct cost VS selling a new phone.
This is capitalism, produce crap, sell many units, offer no support, rinse and repeat.
Market forces have clearly failed to address this issue for an extended period, so this seems to be a situation where statutory regulation is indicated.
Isn't that tying? The manufacturer sells both the product and the service, and they can't be separated. Many countries have laws against it, or at least some forms of it.
Furthermore, many countries now have some regulations regarding e-waste. Either now or in the making. And I expect "bricking" to be punished in some way. Right to repair comes to mind.
The manufacturer should never be able to either deactivate or otherwise disable the functionality of a device they sold in a form that requires a service to function.
Exactly, why can’t I still dial in with my 2400 baud modem to get virus definitions for Norton Antivirus for my 386 that I bought in the 80s?
You can dial your 2400 baud modem if you find another endpoint or operate one yourself, you're not bound to a specific provider. Fax machines still work after all.
You can run norton antivirus for DOS in whatever state it was when it was last sold.
So now all I have to do is find someone who keeps a bank of 2400 baud modems and an updated list of virus definition files.
As far as using as it was, what good is virus protection software that only works for viruses created in the 80s?
Sure I can use a standalone gps device that I bought 10 years ago that doesn’t get updated but that doesn’t do me too much good navigating in a suburban area where most of the roads, houses and businesses didn’t even exist then.
A general-purpose computer with a general-purpose operating system and a general-purpose network adapter tied to it absolutely does not require Norton Antivirus or its virus definitions to function. Correspondingly, Norton is not involved in manufacturing any of these components.
That seems like a distinction without a difference. I hypothetically bought a product that requires a service for functionality - Norton Antivirus. According to the parent poster, Norton should be required to keep their service running.
Should Sony also be forced to keep their service running for PS1 games that had an online component - forever?
> Should Sony also be forced to keep their service running for PS1 games that had an online component - forever?
On the other hand, would that even be particularly difficult?
For most companies making games with an online component, if they set aside 0.1% of revenue into a fund that would take over server costs once support ends, that fund would last for a very long time.
At a certain point the player base drops enough that you only need one server and a few hours a month of labor. (Less than one server, actually, once you factor in virtualization. Even if you want to have some redundancy and support multiple regions.)
Technically you didn't buy a Norton Antivirus, you entered into a license agreement. Software is, unfortunately, almost always leased and not bought. However, I would very much be into forcing companies to either a) keep their service running forever, or b) allow the user to use a third-party service - in this case, virus definitions. This is in line with my comment elsewhere in the thread - companies should be allowed to either a) sell a product that doesn't require a service, and if it uses a service, it should support compatible third-party services as well, or b) lease the product as a part of a service. They should explicitly be banned from selling a product that requires their particular service to function.
As for Sony, PS1 games and on-line: similarly, whatever middleware there is that facilitates on-line gaming, endpoints should be user-configurable. Fortunately, at least PS1 the device doesn't require Internet to run.
(Come to think of it, did PS1 even support an Internet connection?)
You’re right. It was first introduced with the PS2 in 2000. I do however have a vague recollection of early cartridge based systems where you could download games over cable.
You're probably thinking of Sega Channel [1] which was functionally like Microsoft's Game Pass that is available now. You knew that you would only have access to the games as long as you subscribed, and that they could be changed at any point by Sega.
First thought being that the "Zigbee Alliance, an industry group that develops standards and certifies products to ensure multi-vendor interoperability" essentially certified something that after all does not allow "multi-vendor interoperability".
I can well understand how each vendor would try to create its own "walled garden", but the whole idea of such "standard" (actually most of the time "pseudo-standard") associations should be exactly that of contrasting the single firm attempt to cut out the other vendors in the same area.
Zigbee interoperability isn't what you might think it is. Think of it like a layer 1-4 interop, while layer 7 (application layer) can, and in many cases is, up to the device vendor. That means Zigbee radios in your house can use each other for mesh networking, but may not actually be able to talk to each other in any meaningful way. ZLL and then ZHA are intended to address this, but aren't always widely adopted by device vendors for a bunch of reasons, one of which might be the "walled garden" you mention.
Solutions like Z-Wave also have a tightly-defined application layer which allows nearly all devices to have some level of meaningful interaction with each other. For example, a Z-Wave motion sensor from Vendor A can turn on a Z-Wave light from Vendor B in response to motion. This was largely possible due to all Z-Wave radios being made by one organization. There are now promises of opening the stack a bit so partners can also develop radios, and it remains to be seen what impact this will have on the interoperability of Z-Wave devices in the future.
If you like tinkering, look into Zigbee2mqtt (=software). It supports loads of Zigbee devices so you can mix various devices together. Everything you need to do is pretty well explained.
IMO home automation is currently anything but smart. It's fun if you like to tinker with things. That'll allow you to get great results. It's still too cumbersome though for non technical family members. Everything needs to be made easier by several factors.
Why? Did ever normal people install electric equipment (wires, fusibles, switches etc) in their homes? The biggest problem here is that most electricians are not keeping up with the tech and, as always, tech is atill young and moving too fast. But in most cases you shouldn't worry about zigbee, z-wave, mqtt etc, you should just tell the electrician what you want to achieve, the basic workflow and that's it. Then you could customize it if you are a bit skilled/interested, or call them again (paying a fee).
I close analogy that more HN readers would understand from the linked article would be imagine a pile of tablets connected to a wifi network, and they want to video conference. Every tablet implements the wifi standards such that they connect quite successfully to the AP at great speeds and range.
However, one mfgr, for the sake of argument, Apple, stuck a MAC address software firewall in their OS 'to ensure a consistent user experience' such that its tablets at the MAC address layer can only communicate with other Apple MAC addressed products.
Older HN readers will recall a vast variety of floppy disk using 80s home computers such that a CP/M Z80 box, a IBM AT 80286, an apple II, a TRS-80 model 4, and a commodore 64, all technically use standards compliant 5 1/4 floppy disks such that any double sided double density disk will work in any drive of all those devices... I assure you the filesystem layer is utterly incompatible and at best you'll need very special utilities.
Somewhat younger HN readers might recall "stupid flash drive tricks" under linux like formatting a USB attached flash drive as ext2 format simply to see if you can do it. And it does work, perfectly. And nothing other than a linux box could mount a ext2 formatted USB flash drive. Technically thats a standards compliant flash drive both at physical interface and firmware layer that will not work on a mac or windows box.
There's essentially an implementation bug in the standards documents for zigbee such that the RF layer will interop but the application layer most certainly does not have to interop and often enough does not, whereas zwave does NOT have the standards document implementation bug and any rando zwave device will mostly pretty much work with any other rando zwave device.
The only technology I'm aware of where 100% of devices will interop is SCADA current loops. Some techs come close like DMX or zwave which never seems to fail. Some semi silo'd tech like insteon is pretty interoperable. Some stuff like zigbee essentially doesn't work IRL unless the mfgr put effort into it, which some do... and some don't.
I think of it as WiFi Alliance. You would expect several WiFi devices connect to the same AP, but you probably wouldn't expect a random widget from company A to interoperate with random widget from company B.
There's a lot of details to chase around though. Like were the actual devices in question labeled as certified, or was the Zigbee compliance just an implementation detail of the hardware they shipped.
I don't think the Zigbee Alliance requires their members to only market compliant products as certified, so they could be marketing them to consumers as compliant and to companies like Charter as easy to integrate.
A requirement that products carry a clear statement about relying on a service to function, and then some rules about how the end of life for such products and services should be sequenced would not be onerous and would make the situation much better for consumers (even the ones that didn't read the statement would get the minimum period of service specified in the EOL rules).
The message you get (or at least the one I get) is:
> WHY Leading companies choose Zigbee
> Interoperable
> Our world-renowned certification program and our large focus on interoperability assure consumers and developers that devices will interoperate.
I don't think that if I have something certified/stamped by them and it comes out that it won't interoperate with something else they also certified/stamped it is excessive to say that their assurance is m00t.
Tech Enthusiasts: Everything in my house is wired to the Internet of Things! I control it all from my smartphone! My smart-house is bluetooth enabled and I can give it voice commands via alexa! I love the future!
Programmers / Engineers: The most recent piece of technology I own is a printer from 2004 and I keep a loaded gun ready to shoot it if it ever makes an unexpected noise.
It's ironic (contrary to expectation) that those who work in tech/web/software, myself included, are often wary of adopting new technology like the Smart Home.
I've used a "dumb phone" for a long time, and only bought a "smart" one due to work necessity. Even now, I barely use the latter personally, or carry it around with me, since I know it's basically a spying device for third-party apps.
I love the concept of the Internet of Things, and have started studying it deeper as a hobby. If my home is going to be "smart", I want to minimize its dependencies, to understand how it works, what services it's connecting to, what data it's sending, and make sure everything is under my control, repairable and extensible.
Like the HP printers that stop working if they can't phone home (without informed consent, I'm sure) - or these Charter security devices that will simply become useless after the company shut down.. I believe the reluctance on some part of the tech-savvy population is justified, because we/they understand the risks better than naive tech enthusiasts.
Either that, or, we're getting older and suspicious of new things - but for good reasons, as our trust has been betrayed countless times by tech companies.
I bought an 'Audrey' years ago - early touchpanel tablet meant for the kitchen counter. Had three buttons for 'channels' which meant url-for-specially-formatted-web-page, for news and weather and music.
Almost immediately 3Com discontinued support for those channels. So I thought "No problem; I'll just go into settings and change the URLs". Nope. No way to do that.
Early internet stupidity, where some marketing type figured "Capture the customer to our data stream so we can spam them". So screw the customer.
Couldn‘t you make a business out of supporting abandoned IoT devices. Make a small server that rewrites DNS requests and URLs so you can serve your own content to these devices. Or in the case of Zigbee emulate different Zigbee routers?
I know the chumby devices, and to a lesser extent the Dreamcast have been given a second life due to people reverse engineering and rewriting DNS requests. Those were consumer products with a passionate community around them though.
As painful as it is, I think these cultural lessons are necessary, for a few reasons.
The first is that people should not buy something just because someone made promises about it; that's how snake oil sales work. They promise the moon, take your money, and run. Instead people should think critically, ask questions, do research. That may mean becoming more technically literate, or not buying it if they can't understand it. (Is a "smart home" really smart just because "smart" is in the name? Or is "smart" just another "miracle cure-all" ?)
And that follows to the second reason: we shouldn't depend too much on technology that we don't understand or control. If your car breaks and you can fix it, you don't have a $20,000 paperweight. If your car breaks and only the manufacturer can fix it, and they go out of business/stop supporting it, yikes. Not to mention if you're in a bad situation where you need to get yourself back on the road in a hurry.
But really, all that applies to all parts of your life: being able to read a map, mail a letter, plant a garden, repair things in your home, cook a meal, etc. Basic knowledge that I'll bet is becoming rarer with each generation. We need to encourage people to deeply understand the world around them so they don't end up as helpless consumers expecting someone else to take care of them.
Slightly out of topic. I'm actually looking into smart home devices for an apartment I have that will be rented half of the year on airbnb when I'm not there. Specifically, I want to be able to control lights, the air conditioner, the heating and rolling shutters remotely.
I mostly want to do thzt because I want to be able to turn off heating and air conditioner remotely, but I would also like to have a master button to turn off all the lights when people live the apartment, a button to turn of lights on a path to the toilets in the bedroom and an alarm that automatically opens the shutters in the morning at the scheduled time (always found that waking up by sunlight is better)
Does anyone here have any recommendation of which devices I can control directly with something like a raspberry pi that's always on? I've read that Z-Wave's specifications avoid the Zigbee specifications pitfall that make them to vendor dependent..
Zigbee lights are generally not vendor dependent. You can buy Ikea lights and they will interoperate with a Philips hub just fine.
To use a Raspberry Pi as the hub, you can use Home Assistant and a USB or GPIO Zigbee adapter such as the Conbee or Raspbee. Home Assistant generally requires a Pi 3. Automations are all defined within Home Assistant so they do not depend on having homogeneous vendors.
Thanks, so I would also need to use zigbee switches everywhere and use zigbee to turn the lights on and off? Are there any good websites that review zigbee switches to find which ones are decent...
I usually go to Wire Cutter for this but their specific article is not really useful since they don't recommend any pure zigbee switch but only proprietary solutions.
Yes, you would. I am actually working right now on a way to use an ESP8266 or ESP32 as an interface between dumb switches and Home Assistant (via MQTT), but it requires substantial rewiring and, being all DIY, is probably not suited for something that you rent.
For Zigbee switches, depending on your lights it could also be power plugs. Ikea and Philips are again the easy choices if they sell what you want. Depending on where you are based you could also find something that fits in your existing switchboxes.
HomeKit stuff will do that with an Apple TV acting as a hub. Once configured, you can then remote-control stuff from iOS devices connected to the same account or 'family' (once you turn on 'Allow Remote Access'). The list of supported devices isn't huge, but the certification comes with some useful guarantees, like not being dependent on the manufacturer's cloud service to work.
The devices on the list that act as hubs also carry that support across to anything else they control; for example, the Philips Hue Bridge lets you treat a lot of Zigbee 3.0 lights/switches as if they were native HomeKit devices.
Well, I obviously wouldn't do anything when they are staying in the apartment but once they check out, I would use that to turn off the lights, air conditioner and close the shutters.
My parents have an apartment they rent and it's happened that guests forget to turn off the air conditioner at the end of their stay... It's not cheap to pay for that for multiple days not to mention that it's bad for the environment
I'd love it if there were legislation requiring products that include an "online" component to be required to explicitly advertise the minimum number of years of guaranteed service, and be forced to prominently advertise "0 years" (prominent like cigarette cancer warnings) if it isn't guaranteed.
And be forced to post a financial bond at a level set by law that will be forfeited and used to keep the service running (e.g. auctioning to software companies who will bid to run the service) in case the company decides to cancel, goes bankrupt, etc. (As long as the company runs it, they get the bond back at the end of the guarantee.)
Consumers should know they're getting an actual guarantee, or be 100% aware that there isn't any guarantee at all, so they actually know what they're buying in advance.
Without customer protection laws you're left to the whims of your vendor. Trusting the promises in the sales brochure (or even taking the sales pitch as a promise) is either courageous or naive.
Isn't this symptomatic of a much deeper problem with the tech industry today, though? Consumer protection laws are useful up to a point, but even in places with relatively strong consumer protections, the problems of closed systems, ephemeral external dependencies and built-in obsolescence are widespread. Almost nothing is made to be long-lasting any more. Compatibility and interoperability are often seen as dirty words that imply loosening the stranglehold you have on your customers. Standards are great, so we should have our own because they will be better than everyone else's.
There's an old saying that you can't implement a technological solution to a social problem. I'm increasingly of the view that this holds the other way around as well.
Unfortunately, there is a vicious circle here. To support makers of products that can be repaired and upgraded, and that don't depend on remote facilities that can disappear but can't be replaced, and that respect user privacy and control, and that play nicely with other products, we need to buy their stuff. That stuff is likely to be more expensive up-front, even if it might work out better and cheaper in the long term, so buyers need to perceive some advantage to justify the extra cost. But until those buyers have been stung by the deliberate limitations and user-hostile behaviours of the competing products, which typically only happens later, they often won't be very aware of the choices they're really making. They might experience buyer's remorse further down the line, but by then it's too late. This effect makes it difficult to compete if you want to offer good quality, future-proof products, which in turn limits how many people have those products and how many people see that those products can be advantageous.
All our ceiling lights, aircon, outlets are IR-controlled.
Then I got generic Chinese IR blasters, and use homebridge to connect it all over the LAN to Apple HomeKit.
If there is a shift in tech, someone will always be selling new IR blasters due to the installed base. Worse case, I'll always have the IR remote controls (or can build something out of an raspberry pi/arduino)
Wow, that's pretty cool. What level of iOS device do you need to run it? I try to avoid using my phone too much so I'm wondering if I could control it all with a really old iPad or a cheap old iTouch. Any recommendations on the IR blaster? I've been wanting to have some sort of non-remote control but I don't trust alexa/google home so this looks like a good option that would let me use my existing devices.
HomeKit was first supported in iOS 8, so I think anything newer than that?
The IR blaster I use is a Broadlink RM Mini, it's cheap, has been reverse-engineered, and after you've set it it's happy running it on an "IoT" VLAN with no internet access.
Haha, yeah IR isn't foolproof either. We bought some bottom-of-the-barrel Chinese humidifier, and it reacted to absolutely anything IR, even sunlight sometimes.
Curious: what tech stack did you land on? I haven't implemented much "smart" tech yet, but would be nice to know if there are some good open options out there.
Home Assistant. If you're comfortable with a bit of tinkering that's overall a good choice.
Currently using a tinkerboard, but would probably be looking to virtualise this
>"smart" tech yet
Smartplugs and smart lightbulbs are pretty easy. Things start getting a bit ugly on presence detection though - i.e. who is home and where. Haven't seen anything reliable on that front. The other headache to think about is standby mode. Many things these days can't be controlled via smartplugs because they go into standby mode...not on mode
2nd comment because I remembered some additional tips:
1) If you're considering a device...google the device name & github. Generally if I can find some github code of someone managed to get this to talk to python then we're probably OK to buy
2) The Xiaomi equivalent (Yeelight) of Philips Hue Lights are really good quality/value for money. If you can swallow the associated question marks on brand I'd go that route.
3) Basically everything is gonna require an app on phone initially and it will phone home...so you'll need a FW to block this after you got it to talk to local home assistant server
4) Some of these things like zigbee...frequency depends on country so look carefully what you're buying.
I only hope their customers have learned to never trust another company again whose devices only work via a connection to their servers and not otherwise.
My Zigbee (Hue) lights are controlled with ZWave switches and nothing changes when I have no internet or Phillips or whoever ends up dead. The only thing that needs an internet connection currently is the voice control, but I have plans to change that as well. Everything is connected with an rPi 3 running home assistant.
I have a similar setup. I really appreciate that Hue bulbs are standard Zigbee bubls, then with a layer of fancy Phillips UI on top. If Phillips ever go away, easy to replace with a different UI, and direct control.
It is amusing and sad to me that person is going to "spend thousands more to get a Ring system"
So you have already been burned by 1 cloud service, so you want to get burned again? As there is nothing to day Amazon will not shutdown the Ring service at some point
Smart home cloud services disappearing and bricking devices seems analogous to pre-Internet devices losing their circuit boards (due to heat, tin whiskers, etc.) which are no longer made by their manufacturers. Or a small plastic part breaking which is similarly no longer available.
It's just another step in the long sordid history of intentional or not planned obsolescence.
Maybe it seems more egregious because of the rate it happens at and that the company has to turn off a service by unplugging the servers.
Ultimately, regardless of cloud features or not, you want to buy things that are durable and repairable, whether that's self hosting a new cloud service or fabricating a new metal part.
> Smart home cloud services disappearing and bricking devices seems analogous to pre-Internet devices losing their circuit boards
There used to be an electronic repair shop in every neighborhood. These were small businesses that would be able to diagnose and repair faults in random electronic devices.
I wonder if at one point this problem of cloud services disappearing will become widespread enough that we'll start to see businesses that will take your bricked device, flash their own firmware to it and provide their own, generalized cloud service as a replacement.
Obviously reverse engineering each device and making a custom firmware would be a large initial investment, but maybe at a point where just about every trivial thing will depend on a network service it would be economically feasible.
This will not work if the bootloader on your device does not allow upgrading unsigned firmware and would take some laws prohibiting this behaviour to be in place.
One of the greatest faults of Linux was sticking to the GPLv2, GPLv3 actively sought to deny "tivoization" as they called it, as Tivo used Linux and provided the code, but required signed software to boot, rendering the access to the code effectively pointless
I wonder if the fixed cost of reverse engineering would ever overcome the sheer number of devices you’d have to do this with through scaling geographically.
I'm not surprised by this ridiculous and bold move by Spectrum. Where I live they have monopoly and my bill is going up 100% every year. I'm not kidding. Upon inquiry, they lie that you had limited contract with Brighthouse signed and now the promotion is over so they can be rising it up. The only way to stop that is to kill your cable for at least 90 days, then apply for the new one with their current discounts that supposedly doesn't expire. If you think I'm singled out, there has been lots of upset about how Spectrum conducts itself but without class-action lawsuit, nothing will happen.
"I like escalators, because an escalator can never break; it can only become stairs. There would never be an escalator temporarily out of order sign, only an escalator temporarily stairs. Sorry for the convenience."
- Mitch Hedberg
As much as possible, smart tech should be operable mechanically. This seems fairly easy to do with electrical things like light switches.
Stockholm recently had to shut down a huge new train station for months, because it turned out that the escalators would break down and start going in reverse.
Sure, but that failure is on the end of the spectrum. For electronics it'd be: Online and working / Offline and core functionality is working / Battery fire
A minor example in my house that makes a massive difference during winter storms where we lose power: my gas fireplace is electronically, remotely controlled. Except you can manually start it up if needed. Makes the basement super cozy even if the furnace can't run.
Except escalators never merely become stairs; they are blocked off, torn apart, and under maintenance, making them unusable for weeks at a time here in Massachusetts -- presumably because a malfunctioning escalator could be seriously dangerous.
Having a usable fallback to smart tech's smartness is nice, but it can propose complications (and dangers) the manufacturer doesn't want to deal with because making the gear to deal with them then becomes cost prohibitive. The best thing is dumb tech, to which smartness can be added if desired by the user.
I’ve been a user of HA ever since I got a X10 firecracker plugged into my Slackware server under my dorm room bed. I’ve hated that they’ve turned HA into a service. A lot of these cloud HA companies are going with the cheap razors, expensive razor blades business model. And when things don’t work out leave you will with a pile of expensive broke equipment.
I have been using X10 and Insteon devices for years. I have no plans to upgrade the system to newer zigbee/zwave hardware; though I have added some IP Based devices too for things like smart bulbs such as Lifx.
I control everything with a Universal Devices ISY-994i controller and have set up a Polyglot node server to tie it all together. This gives me flexibility because I can always expand the isy-994i to support Zigbee or Zwave; and polyglot node server is open source.
It’s very nice to be protected against forced IoT obsolescence due to cloud API Reversioning, Company failures and mergers, etc. some of my smart home IoT devices are over 20 years old; no plans to upgrade them. Best of all, nothing requires the Internet to work.
> Open source home automation that puts local control and privacy first. Powered by a worldwide community of tinkerers and DIY enthusiasts. Perfect to run on a Raspberry Pi or a local server.
It's things like this that companies do that should really make one consider setting up their own simple NVR system. It's really nice being able to manage and upgrade your own system this way. Even when you feel like upgrading you still have all your cat6 infrastructure in place.
The upcoming wave of IoT products will all be local-first because too many users (and thus founders) have been burned by the old cloud-addicted products that literally die when their supply is cut off.
In the meantime, just say no to cloud-only products!
Games stop working, lifelong memories are erased (yahoo groups,photo sites), watches stop being smart (pebble) and even smarthomes become a complete waste of an investment.
Bare-metal will always be sexy (and higher maintenance), but prosumer tech is not there yet to fill-in all the gaps in smarthomes.
IMO, once we start seeing inexpensive 10TB-100TB home storage appliances, we'll start to see more tech engineered for decade+ operations.
Over 99% of tech is disposible. Just the data is carried over, so you're lucky if you get 5 years use out of any device nowadays. 10 years is pushing it, even for smarthomes.
I don't think the storage is the problem. Everything in computing scales up with available storage - apps, documents, media, all just get larger to compensate. If we couldn't solve this with readily available 100GB drives, or 1TB drives, why would we suddenly be able to with 10TB drives?
The problem is that normal people no longer have a machine at home that’s always left on. As much as I dislike apple, their Apple TV might be one of the things that pushes everyone in the direction of owning their IOT network.
There’s also (IMO) a serious lack of open source tools for managing IOT devices. Right now a lot of the cheaper ones speak PnPP but there’s only one functional client I was able to find for this (a buggy and obscure GTK app.) There are a lot of good libraries though so maybe that will change.
> The problem is that normal people no longer have a machine at home that’s always left on.
They do, their router. The better models are half a home server anyway. They have wifi, USB for storage, do voip, have direct internet access (no NAT issues!). They would be an ideal platform for home automation.
Except that they usually don't own it, they rent it from the ISP.
I think the concept of home server - some box you buy separately from Internet connection - is a concept that needs more marketing, until it becomes normalized.
I've had the same happen to Wifi plugs bought as recently as three years ago. They all seem to use different (buggy) apps and route the signal to the plug through some obscure website (usually in China). Of course these sites and domains will go down from time to time. Plus, they all use different, undocumented protocols so homebrewing a new controller app is a no-no.
Lesson learned, I now carefully research any smart home stuff I buy now to make sure it has an accessible API I can use if the cloud stuff goes down.
Any home relies on some external service to function at all was dumb to begin with. People need to stop buying these gadget/service combination things.
Has anyone looked at Fedora IoT? They're trying to make updates for this class of device easier.
https://iot.fedoraproject.org/
That doesn't itself solve, or prevent, the problem of abandoned products. But I wonder to what degree making updates easier and more reliably deployed would help.
They are really going out of their way to not say that any IP camera is any IP camera - it does not matter that their "security hub" is zigbee, the cameras are IP protocol and any Internet video streaming software will connect and stream these camera's video quite well. Just get the FOSS iSpy app on your home network and you're operational, locally.
This is why you might consider getting services from companies where the service is their core, rather than home security from a media company. Lower risk of them deciding “oh, that junk we tried isn’t making enough money, let’s kill it off.”
Spectrum partnered with Ring and Abode to transition the customers - they were offered free equipment with the purchase of 1 year monitoring. Glad that customers got to choose between systems and weren't just left to one.
> A big question is why there's no way for Charter customers to keep using these devices, given that they rely on the Zigbee specification that allows multi-vendor interoperability for smart-home products. Why can't Charter customers connect their security devices to a Zigbee-enabled smart-home hub or use them with another alarm-monitoring service that supports Zigbee? One user on DSLReports pointed out that years ago, Spectrum devices "were firmware coded to prevent them from being seen and usable within the normal universe of Zigbee devices."
I imagine this was a pretty short conversation between the engineers and PM:
- Engineers "Why do we have to lock them down? There's no technical reason to do this to our customers."
More realistically how it works in giant mega-corporations is a contractual arrangement with another company works until it doesn't, and after the contract is no longer in force... well... its all over.
There's a subgenre of the above where mega-corporations set up and later dissolve an entire department. So any employee who could unlock those devices was probably downsized months or even years ago when the project was internally cancelled and the shutdown was planned. First, fire the engineers because we're cancelling so no possibility of R+D, whoopsie that means no possibility of unlocking, oh well.
As someone who tried to create a company that produced smart home tech that didn't even have the ability to phone home, this is why we can't have nice things.
About ten years ago the market was perfect for a smart home renaissance. Hardware was becoming cheap enough and powerful enough that your average bootstrapping solo founder could start a business with off the shelf components. Everything was in place for the old guard of laggy and clunky consumer tech to become fast, efficient, and offline.
But giant corporations and VCs smell an opportunity like a shark smells blood in the water. Everyone jumps on the bandwagon with a creeping sludge of inferior offerings, smothering the ecosystem. Only the billion dollar surveillance capitalists can penetrate the toxic market pollution and reach consumers.
And so, while you may have a handful of niche/maker products that hang on, the consumer is led to believe that everything is an Internet of Shit. And this is why we can't have nice things.
You can always try with a zigate (https://zigate.fr), still, it won't change the fact that some manufacturers do interpret the protocol in their own ways ...
maybe there is a case to not have all this smart cack in your homes. The amount of e-waste that is going to be made, because its not the latest and greatest product/service is enoourmous. Not everything needs to be smart and connected to the internet.
Nura, a young Australian company that sells smart headphones, blocked my pair in a vain way to circumvent banking agreements in case of fraud. They are mixing their terms of use with their terms of sell and reserve the right to brick your device anytime they want: in other words they reserve the right to ban you from their service (which in fact is a calibration service) but do it in in such a way they basically trash your property rights away.
My comment was too long, so I just dumped it on pastebin:
A long time later, I accidentally stumbled upon this puppy right there in the closet. Took it out of the box, and to my first disappointment, it wasn't wireless. You had to plug an Ethernet cable.
Hit the power button and the screen lights up with the word, initializing.
When you have waited long enough, the screen changes to Network error. I tried plugging it in directly on the modem to no avail. I tried to debug it by analyzing the network, the domain it pointed to was no longer resolving.
No internet feature for me I guess. At least I still had a Radio and CD player....
Let's tune in to the radio. Automatic station search. Apparently, that requires the internet. So after you wait long enough, Network Error.
Well let's go with CDs then. When you insert a CD, it looks for the album info online. Oh you have no connection? Network error it is.
Left it on my desk as decoration for a couple years before I threw it in the garbage.