I agree but in some cases, you end up with such a restrictive product because it's part of a system package.
For example, I recently bought a new Trane air conditioner system and it comes with a fancy system controller.[1] It has wifi so Trane promotes the capability for you to control the temperature remotely from your smartphone. Well, the IoT rub is that it requires the cloud for that capability. When I first heard that explained to me by the installer, I had to have him repeat it twice because it didn't make any sense.
"You mean my iPhone that's on wifi doesn't communicate with this thermostat that's on the same wifi -- directly? You mean the thermostat app on my iPhone has to send the "increase temperature" command out of the house to the mothership cloud, and then round trip back to my thermostat?!? Yes, that's how it works." That's crazy -- If the internet is down, my iPhone can't control the thermostat that's 10 feet away!
Needless to say, I just forego the smartphone remote app and simply walk up to the thermostat and change the settings on the wall.
[1] https://www.trane.com/residential/en/products/thermostats-an...
smartphone at the office --> cloud --> thermostat home
The problem is that the engineers made the command flow with the cloud as the only method of communication instead of enabling a secondary communication that lets 2 wifi devices talk directly. Obviously, direct wifi communication is technically possible since people buy inkjet wifi printers and their wifi laptops print directly to them every day. People would be livid if they couldn't print their paper just because the internet down or the vendor's cloud server had an outage for "maintenance". The data sent from the computer to the printer does not roundtrip up to the cloud and back down to the printer.
Yes, you reminded me of a very good point about local wifi communications:
If 2 devices are on wifi, it's easier if both are configured as clients that talk to a 3rd server (such as cloud). When both act as clients, they don't need to know each other's ip address to communicate.
The other option of configuring one of the wifi devices as a server is more difficult especially for non-techies. Now, you have an extra complication of discovering the ip address of the printer/camera acting as a "server" to receive commands. This requires a weird dance where the server device has to "broadcast" itself and you have the client computers "find" it. For my Canon wifi printer, I had to press and hold down the power button a certain way so the computer could "discover" it. For my wifi camera, I had to go the DOS prompt and enter an obscure ARP command[1]. The difficulty is compounded when the "server" device has no builtin LCD display so you're basically "flying blind" as to what ip address it got from the wifi router's DHCP.
However, the HVAC wall controller has a large LCD display and touchpad so configuring it as a server on local wifi shouldn't be that difficult.
Total bonkers. Silicon Valley only knows how to make web services apparently.
If your IoT device depends on the internet for functionality it really should be able to connect to cell networks, not just wifi. The only problem is the security risks with that, but it doesn't expand the attack surface that much more than having a cloud-connected device as it is.
Some proposals for future 'smart cities' is having CAT-6 LTE (which can be as fast as most home broadband connections) broadcasting everywhere and having all the devices connected to that, so you don't have to depend on your home wifi staying on or being in range.
The Chromecast could be discontinued tomorrow, but it only cost £30, so I'll just replace it with whatever. Relying on your expensive TV's probably terrible and badly supported software is madness, but it's what most people seem happy to do. It's getting harder and harder to buy TVs that are not smart, and I find it maddening.
I take the same approach with Speakers.
As soon as you do you'll get nagged for software updates, you'll get told the privacy policy or the terms of service have changed (which you have to acknowledge) and apps for things like Youtube and Netflix tend to be inferior to those in boxes (eg Roku, AppleTV).
I buy my TV to be a dumb screen. Then I plug a Roku into it. If at some point I decide the Roku isn't working I throw it away and get something else. Worst case I'm out $120.
Non-smart TVs basically don't exist anymore (Sceptre is the only brand I can find). You can usually get a smart TV and keep it off the internet, but I've seen horror stories of some TVs not working at all until being "updated" after unboxing.
A similar pet peeve of mine is that the huge infotainment screens on modern cars cannot be turned off the majority of the time. Only in the VW Golf I rented did it actually have a way to shut the whole thing down.
I think a dumb screen option will always be available because it's always going to be a value player in a world of Sony and Samsungs. The question is will the screen be a nice enough panel.
But I could have saved a lot of money by going with a conventional Smart TV, and it would have been easier to shop for it. Commercial displays aren't hanging on the wall at your local Best Buy, although they might be hanging on the wall at your local bar, restaurant, bank, etc.
It is not possible to buy a high-end panel without it anymore.
AS long as you just use it as a dumb screen, physically disconnected from the network, it'll keep doing it's job for year and years.
My TV is a Smart TV. I've never use that part of it. In fact, it came with a second remote control for managing the Smart features and I've never even put batteries in it let alone given the panel network access.
It's been happily humming along in dumb mode for about 4 years I think. Just plug a cheap accessory of your choice in (Fire TV, Chromecast, Roku, self built HTPC etc) and enjoy!
Of all the television sets I have owned since 1985, the native tuner included inside the device has been used maybe 2% of the time. The only advantage I have ever seen from buying a display screen as a television set is that it comes with speakers loud enough to hear from across a large room, and has as many as six different input ports. None of the screens I now own currently decode their own NTSC/ATSC/QAM signals.
My first "smart" TV was also my last, because LG made it obsolete within months of me buying it. Now it also functions only as a dumb screen, and is not allowed on my home network. Never again.
You don't need a smart television now. You don't even need a television. You need a monitor with HDMI inputs and speakers.
And if you want a DisplayPort input in lieu of one of the HDMI sockets, that seems to add at least $30 to an otherwise identical monitor, even though DisplayPort is royalty-free. There does seem to be an attempt to segment the market to charge more for displays connected to a computer than for displays connected to a home entertainment appliance.
At larger sizes, I think the assumption is that if you don't have a TV tuner, you're attaching it to a computer, and if you do, you're not. And that breaks down at smaller sizes, where resolution and pixel pitch become an issue for computing, but less so for consoles and DVD players.
I haven't done much digging on this, since 40" is plenty for the longest line-of-sight possible in my house.
I'm using an NEC E-series 55" commercial display as my TV and gaming monitor. It's not a great TV, but it is solid and it's way better than the one it replaced.
If it's never connected to the network....
But in many other cases of smart devices where the core function relies on external services, they'll be entirely useless.
I volunteered to set it up, and what do you know -- the "OS" was gated by a registration / sign in screen. So, no way to access core functionality (including external inputs, settings menu, etc.) without an internet connection. Even if there was some strange ritual that would let me get through (there wasn't), the average user is fucked.
Aside from privacy, I like to think of it similar to musical keyboards. Back when synths were primitive, it made sense to buy a keyboard that had builtin speakers, synthesizer software/hardware - a smart keyboard if you will. Nowadays, the only thing you would want is a good set of keys. That means a midi controller with a baby grand/grand like keybed. [1] Best of all, it should be a lot cheaper than an all in one. And itll be just as good in 20 years as it is now - youll just run a different synth on your computer. All in ones dont make sense for many things. It may be a trend but its an overzealous and oversold one. Ill stick with my dumb midi instruments and my dumb projector. And hopefully theyll still be with me in 5 or more years.
To go with the music world, the best parallel would be the increasingly annoying tendency for music software to move to "cloud" based offerings or other Internet authorization techniques. For instance Roland, with their "cloud plugins", apparently thinks all music venues have super spiffy wifi, as they require network authorization every time the plugin loads in the DAW. (https://www.rolandcloud.com/faq#FAQLink17) If Roland ever pulled this authorization scheme, poof, your plugins no longer work.
I have a 16 year old keyboard (Alesis Andromeda) and a synth that is a bit over a quarter of a century old (Roland D-50), and they both work fine more or less (a few issues here and there but nothing that stops me from using it). They don't need an Internet connection just to work. The companies can discontinue support (or even go bankrupt, in Alesis's case), and the synths still work.
Personally, I certainly do not have great confidence that the Roland Cloud authorization service will last a quarter century.
Hardware that requires the Internet to work is no different than software that requires the Internet to work, it honestly is a risky prospect for longevity. The Harmony Link from what I see was released in 2011. Maybe 7 years is a long time in this current disposable electronics world, but it doesn't feel like that to me.
It will forever try to login into that wifi network, even if you disabled the wifi in the tv ui.
Sure, all TVs will probably not do that, but before this one it wouldn't occur to me, that it could do that.
Lesson learned.
This type of behavior was the cause of new Apple Watches that couldn’t access the net over 3G (they were trapped on captive WiFi networks with no UI to disconnect). If this happened with one device, it could happen with others.
edit: split a run on sentence
Anyhow, you are all wrong to disagree with me. There were just CVEs for radio chip exploits a couple of days ago! These have been coming out quite often lately. The layer at which these exploits can be done does not require the TV to be on any network. Not even ad-hoc. Just the fact that it has a listening radio chip that has old firmware is enough to make it extremely dangerous [digitally.] That TV can literally became a wifi enabled worm distribution center, infecting everything in your home. Not worth the risk IMO. I would stick with a dumb TV.
"targets didn't have to connect to the attacker's Wi-Fi network. Simply having Wi-Fi turned on was sufficient to being hacked."
[1] https://www.google.com/amp/s/arstechnica.com/information-tec...
Dumb TV's with sufficient quality are becoming harder and harder to find. I prefer to plug in my own smarts to TV's since they evolve quicker than the years I'll own a TV.
The manufacturer then sells the smart tv at a $25-50 premium over the dumb version and it's a decent profit for them.
FWIW - I was always skeptical of the onboard TV smart features, but the Samsung TV that I use a computer monitor actually is actually surprisingly decent.
https://www.amazon.com/Samsung-Electronics-UN55MU6500-Curved...
As an aside, I absolutely love using it as a monitor hooked up to a MBP.
First, price. I think this is the biggest factor. You need a computer (for each TV). I don't think there are too many computers (if any) in the $35-$150 range that can handle smoothly playing HD video and pushing it to an external display. Assuming the machine isn't dedicated and you already own it, it breaks point 2 below.
Second, I haven't seen a good setup for controlling a computer connected to a TV. Granted, I stopped looking a few years ago. For me (and I'd assume most people), "good setup" means ability to turn system on and control it remotely, at a minimum. I'm sure one exists, but it would definitely require even more hardware (cuts into price) and more custom setup.
Third, some people might not have the technical know-how to set up something that meets the above two things. Sure, plenty of us technical people can do it, but it's not hard to understand that we are the minority.
tl;dr there's nothing "wrong" with using a computer and an HDMI cable, but there are plenty of compelling reasons these days to use one of the ultra cheap boxes. They are basically cheap purpose-specific computers, anyway. Ultimately, different people have different needs and priorities.
I have a mini-keyboard with touch pad, but it is still inferior to a roku remote or my harmony remote. My wife won't use it.
Then there are the complications DRM adds, e.g. it's often not possible to get the best quality from Netflix on your PC in the browser, you only get that with the native Windows 10 app (and then not 4K unless you have some very specific GPU and drivers). The Apps on a smart TV or a FireTV/whatever don't have these restrictions.
With Netflix you at least get an app and the possibility to stream with highest quality (4K, HDR, multichannel sound), but the app is extremely basic and clearly only optimized for touch - it doesn't even offer basic keyboard shortcuts beyond start/stop.
You can get a somewhat near-HTPC using Chrome (at least shortcuts work there) and some customizing, but you're stuck with stereo (I'm not sure about 4K/HDR, but the latter is no-go enough for me to not have tried).
The same goes for Amazon Video, except that they don't even have an app outside Android/iOS, so you're stuck with stereo anyway. And so on. It would be downright embarassing, if it weren't so obvious this is intentional...
Does anyone know of a good smart thermostat with an api that can be used without a proprietary cloud?
Company: http://www.radiothermostat.com/
API Doc: http://assistly-production.s3.amazonaws.com/91626/kb_article...
I'm not falling into this mate, unshortened link is:
http://assistly-production.s3.amazonaws.com/91626/kb_article...
The API would be provided by something like Mi Casa Verde running on some hardware in your house.
You can do everything the cloud services enable, locally, such as ‘scenes’ where you press a button and multiple things happen at once.
There are of course some trade-offs with Z-wave. It’s proprietary and expensive (compared to other things) to license. You’ll find a much more limited selection of Z-wave devices compared to BT or WiFi.
They make Z-wave controllers for raspberry pi and such. I’ve seen plenty of projects that used Mi Casa Verde to enable phone support and automation without going to someone else’s cloud.
For what it’s worth, I’ve got z-wave switches for nearly every switch in my house, a few electrical outlets, and a few corded adapters.
The adapters and outlets are great come holiday time. The outlets are paired with light switches in kids’ rooms so when we turn on the switch for their white noise machines some under bed night lights come on.
I used to use a Z-Wave thermostat and it worked just fine, I switched to an Ecobee 3 with room sensors to help better manage hot/cold spots.
I tie it all together with a Wink Hub, so it’s going to someone else’s cloud. I like their app on my phone better and it just works and has no monthly cost. They’ve got an API to integrate with, and have generally been nice to work with.
I specifically chose Z-wave though to give me options if they go under or I decide to cut ties and run my own stuff.
I've been working on moving my Vera controlled zwave network to home-assistant.io and an aeon labs zwave USB gateway, but haven't made much progress because zwave devices all rely on inputting rhythmic patterns of button presses locally to reprogram them.
https://partofthething.com/thoughts/enlighten-your-old-furna...
What if service isn't discontinued, but merely suspended?
What if the company goes out of business? What if it's purchased, and becomes a different legal entity?
What if the service is discontinued as the result of a legal action?
What if the API depends on some proprietary technology?
"Just open source it!" is a knee-jerk reaction with little bearing on the practical realities.
"Why don't we just make bad things illegal?"
Because those are not things you buy to own. Those are things you rent and which only keep working as long the owner is willing to keep them running.