My vision is that if you want to automate your home, all of the parts need to be able to talk to each other. The Nest could not (and still can't).

It's the same with WeMo, Hue, and every other recent "home automation" product we've seen.


I really have no idea. I'm not sure what would be the best route, either. The existing standards all have their own problems.

What I'd love to see is for Google to make a new, unencumbered standard that anyone can use (like Android). License fees seem to hurt the prices of devices based on current standards, but I don't have anything to back that up...

When Google creates stuff it can still have 'encumbrances'. One is that Google later loses interest and subsequently shuts down or abandons the effort.

Another is that the standard is so 'open' that multitudes of variations and not-quite-compatible sub-standards appear.

It can be quite difficult to create a well rounded and substantive standard that is not reliant on a specific funding model. Would Google have interest in a standard which did not allow Google to peer into the data?

We like standards, but 'we' often want them to be 'free' (that is, somebody else is paying for the creation, maintenance and future proofing of said standard).

That's all true, but here's the thing: someone has to do it, and why not Google?

The fact is, there are already several standards for home automation. I use ZWave, but there's also Insteon, Zigbee, X10, and others. None of these standards talk to each other, but they've lasted as long as they have because enough companies have created products around each standard to outfit a home to the extent that people are happy with.

We might as well give Google a try here, because the current trend is to create businesses around a product and not a standard. When that happens, the path to an automated home full of products that talk to each other is slowed or blocked completely.

Don't get me wrong, I have very little optimism that Google can pull this off. It's going to be a HARD thing to do. But there has to be some way to integrate the products in this space because it's only then that you can do some really cool things with this stuff.

Aye, there's the rub. Most of the negative press about the buyout is precisely in answer to that question. The negative opinion states that Google is the last company you want to be exposing your privacy to. Here is a well-stated opinion piece on that topic. I'm unsure whether I agree or disagree with that opinion yet.

Given that it appears Tony has been told he can continue to run his own ship, I don't see them adopting anyone else's standard; the fact that there's little else convincing out there supports this position. Making their own though, is another thing.

In the Verge interview with nest's Matt (the one in the fire truck) he clearly hinted that they were working on integrating other devices into the nest "ecosystem" (hue bulbs were mentioned); this sounds like integration up at an API level like others are doing - Revolv, Zonoff, etc - vs direct device to device. Personally I think that's the best way to go anyway, because you stand a chance of building something that actually works with a decent number of devices.

...and yes, I got stock for my work. Very happy about how that worked out, which was a long way beyond my wildest dreams. Of course, if I'd joined I would have done a lot better, but then I wouldn't have had all this fun working on electric imp

Yes, I fully realize that. It's been impossible to miss because that's ALL that people are talking about with this. That's why I wanted to take a break from that privacy scare stuff to talk about how this could play out in some sort of positive way. I'm bored of the privacy scare stuff. Really, how much can we say about it? At some point people are just going to have to deal with the fact that the company got bought and it's done. Lets move on.

When I asked "why not Google," it was to ask who might be interested in creating something that addressed those concerns I raised earlier. Everyone is saying that the company that should have bought Nest was Apple. Yeah, like the product couldn't have gotten any more siloed that it already was.


Correct me if I'm wrong, but Revolv isn't working with an API. They're basically hoping that Hue, Nest, and the others, don't break their product. It's a bit of a crap shoot.

I should have clarified this point, too. When I said there should be a standard, I should have said that I'd be fine if that standard was basically cobbling together a bunch of APIs from the various companies. But it didn't seem like these guys were that interested in doing that. And Nest joining Hue didn't excite me because that's still extremely limited.

Revolv has the right idea, but it still has very limited support. I've checked, and most of my ZWave modules and switches are not supported by their product, not to mention that they don't even have an Android app yet. Just because they say they support ZWave doesn't mean it's anywhere near FULL support.

Then you have efforts like CoAP.

When has Google made or used an unencumbered standard, and stuck to it without restrictions? Based on Google's past behavior, I'd answer no to K447's question.

Google's track record is not a good one in this regard. So even dismissing the privacy concerns, how does Google regain the trust on the standards side to be able to push forward with a home automation one? Why would the community trust Google when the next version of the standard could just pull something critical from the open side, screwing over say Hue (similar to how Google continues to pull stuff from the ASOP side and bring it into the Google Licensed side of Android).

Well, depending on what you mean by "unencumbered", there's the VP9 video codec. That does have a pretty extensive IP portfolio backing it (as do the video standards coming out of the MPEG group), but it's open and royalty-free (unlike the standards coming out of the MPEG group). Whether or not Google will eventually abandon VP9, like they did VP8, I couldn't say, as H.265 doesn't (yet) have the same dominance that VP8's competition, H.264, did when VP8 was released.

They also developed SPDY, which is being used as the starting point for the HTTP 2.0 standard.

And they certainly use HTML5.

So, off the top of my head, there are three unencumbered standards that Google has made, contributed to, or used. They're also members of the Open Handset Alliance, which develops a lot of standards being used in Android (that are open for use in non-Android phones, too).

So, it's not impossible.

But I do think it's unlikely. I think Nest is going to go the same route as other home automation brands. Their stuff will work with their stuff, and they're not interested in being "integrated" into a system filled with a bunch of NIH devices.

Agreed this is an area where they seem to be on the right path by trying to compete instead of catch up. Though the VP8 history looks bad for VP9 to work out. Mozilla bought into the message Google preached about VP8. Google then never followed through with dumping H.264 from Chrome, which ended up hindering VP8 adoption and widespread use. Having Google owned Motorola sue and lose over standards essential H.264 patents also didn't help.


Ooo, thanks for that reminder. I had forgotten about SPDY, and hadn't realized HTTP 2.0 is based off it. This could be promising if Google's work continues to be migrated to a proper unencumbered standards organization.


They do, but they also actively embraced Flash inside Chrome and Android, just to spite Apple's own HTML 5 efforts with the iPad. I still don't trust Google in this realm. Especially since today I still can't play videos on Google+ inside Firefox properly. (also related to the VP8 issue above).

Personally, I just hope Google exits this blind rage mode they seem to be in, striking out at their big competitors in negative ways. It reminds me too much of Microsoft's behavior. The rivalry between Google and Facebook seems to be very unhealthy for Google. To the point it's lead to the negative reaction of them buying Nest. I can see lots of potential positive from it, especially considering the design angle Nest brings.

Hopefully it works out in a positive way. And congrats to Hugo for getting a small piece of that payout, it's well deserved

Internet of Things: The "Basket of Remotes" problem

Exactly. This is what is bothering me so much about the current trends in this space. These products have great marketing and design, and often have better ideas than the products that came before them. But they don't work together, which minimizes what we're able to do with a whole-home system.

There have been some efforts to get these things to work together, but it's mostly through hacks or IFTTT. I love IFTTT, but it's too slow and limited.

Hue published their API and I believe nest finally did too (or at least they are working with partners on an API). I don't believe this is all reverse-engineered stuff.

Published API can be deprecated or abandoned.

...which then upsets your customers. What's your point there?

Even a published and currently supported API is only as reliable as the company(s) providing it. Which is nothing new, of course.

It is unclear where or how a decent set of reasonably long-term interoperable standards will evolve in this space. The long term part is hard since home automation equipment tends to be adopted and used over many years.

That X-10 is still around could be viewed as testament to how slowly the wider consumer home automation space moves.

Perhaps Google + Nest will surprise me and create something durable that can outlast Google's direct interest and support.

I'm glad to hear that these companies have finally published their APIs. Excellent. Honestly, that would mostly take care of one of the two big issues I have with recent automation devices.

I don't want to continue my negativity, but I still have issues with the direction these recent systems are taking. It's still a bit of a mess to get all these disparate systems working together. For one, you're going to end up with multiple controller modules. If you want to control those Hue lights you still need the Hue controller, for example. The Rovolv talks to that controller, not the lights.

Second, unless I'm missing something, all the commands have to be sent over the internet. This seems pretty slow to me. I'd be interested in seeing a full setup and how responsive it would be. The goal of ZWave was to have a protocol that responded instantly and without fail (it's the entire reason I switched from X10 to ZWave). The speed with which this happens can be impressively fast, and doesn't depend on WiFi, which much of these new devices do.

Lastly: cost. This is still a huge obstacle to home automation adoption. In fact, it's even worse now. The Nest is about twice as expensive as other HA thermostats, and the Hue bulbs will bankrupt you if you wanted to outfit your whole home with them. High prices have always been a problem with HA. I only afforded the equipment I have because of a radio shack firesale.

All I'm trying to say is that everyone is acting like Google buying Nest was the worst thing in the world. My counter is that Nest wasn't going to be the company to bring home automation to the masses on their own. It just wasn't going to happen. Why not see if Google can do something in this space?

I also find it humorous that it seems like most of the people freaking out over the acquisition are Nest owners but not home automation enthusiasts, and they're the ones saying how bad this is for home automation. I'll be the first person to say that home automation is in a terrible place right now, and I want to see if someone can do it right. I don't know if Google is that company, but this market has been around for decades without someone getting it right, so lets give them a freaking chance.

Being one of these people, it's been because I wanted to wait. I was a PDA (Palm, iPaq) enthusiast, but saw the nightmare that were smartphones pre 2007. Waiting got me the iPhone. Waiting on home automation got me the Nest. Sure, both were expensive. But because I didn't buy into the bad marketplace before, my TCO for a smartphone, or home automation just for a thermostat was still lower then many people. And I voted for my wallet on what I saw as a superior clean start to the issues in both markets.

I also have the living situation of being in and out of various apartments over the past few years. Installing a Nest thermostat was easy. Trying to install all the automation stuff at every location would have not been worth it to me. Once I have a more settled place again, then maybe.

I'll wait and see how the Google deal turns out. I just don't hold much hope of it being some sort of magical solution in the space. If Google leaves Nest alone for the most part but lets them tap into more resources of the mothership, then perhaps it will work out. But if Google tries to shove Nest into the ad revenue category that the rest of the company operates at, it's game over for me. I prefer to pay for products, not be the product sold.

I get all of what you're saying, but I also don't understand the assumption that Nest was going to turn into a complete home automation company that could hand you all the tools you needed for...well...an automated home.

That's what bothers me here. It was already clear that Nest had no intention of doing that. They were partnering up with a systems integration company. If you thought installing past home automation products in your various apartments was difficult, Nest didn't have a plan to improve that for you. If anything, it would be worse because you'd have to hire a company to do it for you, pay their overhead, and rely on whatever solution they installed for you.

Again, I have no idea what Google is going to do with Nest. I only have a small idea of what Nest would have done on their own, and it wasn't what you seem to be thinking, though I may not understand what you were expecting out of them.

By the way, I want to address one more thing about the Nest: it just doesn't work for everyone.

We've talked about this before. To sum up: I have a completely erratic schedule and never leave or return at the same time. When I am home I'm the only one there, and I spend most of the day in front of a computer. Lastly, my thermostat is in the least busy area of the house (the dining room, where nobody ever walks through).

This means that the Nest isn't a choice for me, and at best the only way I could use it would mean I'd have to buy at least one Nest Protect for my office to show it I was home.

That's a PITA. You know what's easy? Tapping a single button when I leave, and another one when I come back. Actually, I have an NFC tag outside my home that adjusts the temperatures for away mode. I have another tag in my car that I tap when I'm heading home so it can set comfortable temps. This way my thermostat is immediately stopping (I'm not wasting time ramping it down), I have precise control over my home/away schedule, and - most importantly - every other part of my automated home is controlled at the same time. That tag outside my house turns off all the lights and makes certain the door is locked, for example.

I'm glad you voted with your money and waited for the Nest. In the meantime, I was enjoying years of having a home full of lights I could control from my couch as easily as I could from Europe. If I'd waited for a fancy thermostat, I'd still be waiting on those lights.

*edit*
BTW, I'm not sure what moving around had to do with waiting for the Nest. Every other thermostat that integrates with home automation wires in exactly like a Nest. If anything, the Nest is a more difficult thermostat to swap into a rental property due to its shape...

The security aspects of externally (Internet) controlled and/or monitored home equipment and devices is a largely under appreciated issue.

I found this article to provide an informed perspective. Bruce has been providing thoughtful and deep insight on security related matters for many years.

How the ongoing provision of security updates and enhancements to the multitudes of in-home connected devices is going to be funded and actually delivered (enforced, almost) is an open question.

Perhaps these devices need to have two part funding. You buy the thing, but the thing only works if you subscribe to a service that maintains and updates it. The service needs to actually provide good security/service or the affected devices will develop negative reputations and the users or device manufacturers will shop elsewhere.

And somebody needs to be able to audit the actual security as implemented in those devices and updates. All this has costs. The Internet of things seems to be very much not 'easy' to build properly.

It is not only the (home) user that is affected by breached security. For example, large scale botnets exist primarily because there are so many computers, and now other devices, that were unable to resist being compromised. Those botnets are an actual force in the real world, affecting real users and costing real money.

Up to now most of the compromised devices on the Internet were (presumably) Windows computers. Perhaps as the population of attackable routers and other 'smart' devices continues to expand it will be profitable to compromise and conscript these routers, Nests and other plastic devices.

The compute power inside these things may seem modest, but the average compute power will increase as Moores Law progresses, and legions of such conscripts can provide sufficient parallel computation to be worth pursuing. Since these things tend to be powered and connected 24/7 and their activity largely goes unmonitored, they seem ideal for conscript purposes.

Bitcoin itself is in part threatened by the mining of coins at scale using 'stolen' computing cycles

A recent malware attack which locks computer users out of their own data until a ransom is paid (via bitcoin) highlights the potential for compromised security on a device to become a real monetary cost.

Indeed, the archaic wiring system used for HVAC is a problem. Adapting an ancient hard wired 5-conductor system at the far end of the cable seems backwards.

The furnace itself should have the Ethernet or WiFi connectivity. The thermostat and/or room sensors (temperature, occupancy, whatever) should in no way be restricted by where the original in-wall thermostat wiring happened to terminate. I suggest it be abandoned as if it was an old knob and tube wire.

The classic 5-wire HVAC system carries no smarts at all. No remote diagnostics, no protocols for relaying service or maintenance information. The primary method the furnace has for signaling a problem is partial or complete shut down. When the occupant notices the house getting cold, that is the only signal the furnace could send.*

When confronting the unhappy heating equipment in person, often the only user interface is a single blinking LED. Which can only be seen by peering into a sight hole, and for which the code pattern is erased as soon as the main access panel is removed!

Quite ridiculous, really. In North America the bulk of home HVAC equipment is manufactured by a handful of consolidated companies. It is possible for a far more modern and intelligent HAVC control method to be be defined, implemented, and sold.

* I just diagnosed and repaired a neighbor's natural gas furnace last week. It would sometimes delay or completely refuse the thermostat's demand for heat. Turned out to be a cracked transition assembly which was affecting an air pressure sensor. The furnace would sometimes be able to start, sometimes not. If it could not start four times in a row it would shut down for three hours, then try again.

When the furnace access panel was removed it affected the ambient pressure just enough to allow it to start more often, making accurate fault diagnosis difficult.

The thermostat was replaced as the initial diagnosis suggested it may be a factor. Installed a Honeywell color LCD thermostat with WiFi connectivity . Of course this fancy thermostat still had no idea what was going on with the furnace since it was hamstrung by the antiquated control wiring system.

That furnace was manufactured in 2010. I certainly hope the next furnace will have nothing to do with that dumb five wire system.

Do any furnaces have an OBD2 type port? Or do they all do the flashing LED as their peak form of diag output?

According to this article ;

Unclear whether there is any standardization on the smart link(s) between furnace and controller, or proprietary in each case.



Here is the Energy Star 2014 most efficient furnace list

My goals are different, and having a better thermostat was a key one at the time. Turning on or off lights isn't a priority to me. The thermostat was in control of much of my monthly bill in Texas. My belief is that Nest wanted to reinvent common home items, adding smarts to them. In time, it would grow more into the general automation space. They also were able to push for integration with many utility companies.

Kinda like the iPhone. The first version didn't even have apps, cut and paste, and many other features. But yet, by starting from scratch and growing, it is now a dominate platform and pivoted the entire industry.

To me, most classic home automation approaches for the thermostat would have meant some ugly box to replace the previous thermostat. Then for the motion sensing if they wanted the auto away, individual detectors wired around the house somehow.

Nest instead built the detectors into the smoke/CO alarms. While also again improving the product they wanted to replace, instead of just wiring the same old stuff into a network. You see it as a PITA, I see it as a nice benefit that tackles two problems in one (Annoying alarms and the need for more sensors).


Lighting, I'm much more interested in something like Hue, since it does something more then just replace my light switch. It also allows some interesting mood/color stuff, which could be extended into a light based notification system. Again, without bringing the baggage of home automation attempts from the past few decades forward.

I look at old home automation efforts the same way we see floppy disk now. Some people tried to extend the floppy (Zip, LS-120). Others rethought it from the ground up and made USB thumb drives. Was Nest going to be the only one pushing forward? No. They also helped pivot the industry, thus improvements across the board.


Nothing does.


Doesn't sound automated to me. In Texas, if I were to forget, it's a noticeable hit on the AC bill for the month. I understand this setup works for you, but it doesn't work for everyone


I already said it was easy, but it seems you didn't believe I meant what I said. So I'll expand.

It was easy for me to swap, as they actually designed it to be so. From the details on how the wires mount into the back plate, to the adapters in the package to cover up odd paint spots due to larger units on the market. Again one of the reasons I appreciated Nest, they had such a strong sense of design. I didn't need any tools outside what came in the box and was up and running in 15 minutes.

No, it seems you didn't understand what I said. You implied that one of the reasons you finally installed the Nest was because you were moving around a lot. I was asking why that had anything to do with it, since it's as easy to install as any other thermostat. I like that they supply an adapter and good wiring diagrams, but thermostats really aren't difficult things to replace. Armed with a screwdriver, anyone can do it.


As for the rest, I'll keep it brief. I still prize interoperability. Just because you don't want or need automated lighting doesn't mean I shouldn't push for it. You've been assuming a lot about Nest's intentions before the buyout. The only evidence I've seen (partnering with Control4) indicates they did not have these grand designs that you're ascribing to them.

In the end, this is currently moot. The controller I use for my home automation setup is the Vera Lite. It has user-made plugins for both the Nest and the Philips Hue. Thanks to the type of thinking that "old style" automation brought us, I can bring some of these products into my system... but I don't want to. The $150 I saved on my ugly ZWave thermostat means I can pick up 2-3 light switches.


Okay okay, I should have been more clear on that. How about this: for some people, the Nest offers nothing over a wifi thermostat. How about that?

Tom, I think I'm just confused about that you're arguing for. This is not at all an "iOS vs Android" thing here. I'm not arguing "vs," I'm arguing "and." I think Nest is a really nice product (for some people). Why can't Nest do what it does and also talk to other things in your home? You seem to be arguing against that, and the only reason I'm hearing is that you're worried the other products it could partner with would have bad design.

Any other normal thermostat, perhaps. But no on any other thermostat with WiFi at the time. They all required the extra C wire with 24V for power. None of the apartments I've lived in had this wire. And trying to install a C wire for voltage jumps install time from 15 minutes to involving building management to make modifications to wiring.


I assumed Nest would continue to make more smart products for the home. And they have hints of home automation just from the aspect of being more smart then previous products in the space. Your home automation efforts let you set a temperature from somewhere other then the thermostat. Nest gave me that home automation feature too.

So far, my assumption is true with the release of Nest Protect. Never anything more then this. I made one comment that the Nest could be the start to my own home automation efforts. Looking back to refresh my own memory, you and Bruno were having the home automation talks in the initial Nest thread.


As answered before in that old thread in 2012, it is likely due to Nest's startup nature. Startups (from personal experience) have to run lean. Trying to interoperate before shipping a product is a potentially startup dooming waste of time. Even with the second Nest Protect product out the door, Nest was still in investment seeking mode.


That may be the problem. I wasn't arguing, I tried to give you the perspective I thought you asked for as a Nest owner, but not a home automation enthusiast. Then it turned into an argument. I personally never had the assumption that Nest was going to be a full home automation solution, and my posts dating back to their initial announcement reflect this. It might have kickstarted me more into it, had I had a stable living situation. But until that happens, I haven't done a deep dive into the space and simply have a casual surface view dating back to the X10 stuff.

I probably should have just stopped when this was said:

Fair enough. Sorry to get you riled up again I guess we both should have been clearer about where we were coming from and what we were debating. I realize now that I was adding too many secondary thoughts about my opinion of the Nest product, and that was getting me away from my primary thesis.

I still think Nest had no intention of expanding much farther than they had, and even if they did I didn't see them interoperating with other systems, but as I said it's starting to not matter. As long as these companies publish open APIs, other people will take on that "startup dooming waste of time" for them, as people have already done with Revolv and Vera. We're both getting what we want. I also (to get back to the original question) don't see why Nest couldn't continue down the road you envisioned for them.


I swear I'm not being antagonistic here, I'm just curious because I don't understand. The Nest doesn't need a C wire? It doesn't run on batteries so I assume a C wire is necessary... Sorry, I'm just still curious how the wiring is different from other thermostats.

I didn't have a C wire behind my original thermostat either when I moved in, so I didn't know what I was going to do for power. I'd read a few things about running an AC adapter up to the thermostat, but clearly that would be really ugly. Finally I read a tip that I might actually have the wire, just not visibly. Sure enough, I pulled the cable further out of the wall and found that they'd cut and pulled back extra wires and left them wound around the outer sleeve. I picked a color, added it to the C Wire terminal on my new thermostat, and connected the other end to the power terminal at the furnace. Worked like a charm!


Oh good god, I'm very sorry. X10 is the worst. I used X10 for the first 8-10 years that I was playing with automation, and I pretty much always hated it. It was completely unreliable, and "scenes" were handled in the worst possible manner. I remember having a scene in my condo that turned on 12 lights to specific dim levels. It would send the on command to the first light, then slowly dim it to the desired level, then repeat with the second light, and so on. It took about 2 minutes to turn them all on. That's if all the signals got through and there was no noise on the power lines. Yikes.

I may think home automation hasn't gotten very far, but it's still way better than X10. Thank goodness

Thermostats can leech small amounts of current from the other 4 wires in the absence of a C wire (depending on the furnace), usually enough to operate on.

They do this by inserting a resistive load across the "fan relay" wiring, such that it doesn't supply enough current to actually turn on the furnace fan, but still causes enough current to flow to power the thermostat.

I didn't do this for my own setup -- adding a C wire was simple enough and easier for a non hardware geek (me) to figure out.

Cheers