Weave and SmartThings

Looks like Nest/Google is getting serious about Weave. I know a lot of ST users are using Nest and have one foot in that ecosystem. Here’s hoping to a deeper integration with Weave.

1 Like

This part is very confusing, and I’m still not sure if the author understands this correctly – well… certainly more detail is called for, but you can’t expect much from lazy journalists these days:

Update: This story was corrected to reflect that the Nest Weave protocol is different from the Google Weave protocol that was announced at Google I/O.

I’d hardly call Stacey Higginbotham a “lazy journalist.” Plenty of tech industry scoops have come out of that column.

Although it’s not a perfect analogy, to some extent the difference between nest’s weave and google’s weave is a bit like the difference between control 4’s proprietary Zigbee protocol and the open standard Zigbee Home Automation protocol. They’re related, but not interchangeable.

What this means is that there will be weave based devices that cannot work with Nest and will not sport the works with nest logo. It’s like Deja Vu with Amazon’s version of Android. Get ready for Amazon Weave devices that won’t work with any Google Weave or Nest Weave devices. Only this time it’s a sister company to google that is making the proprietary thing.

Holy cow this is only going to confuse consumers.


True, except I think nest is betting on “works with nest” as being enough of a compatibility indicator for mass market consumers. It’s actually the next level up of technically-educated consumer who’s most likely to get into trouble. That’s zigbee’s problem right now, of course.

1 Like

I wish all of the companies would just drop the Works With X badges and build products That Just WorkTM . That said, unless google prevents devices makers from sticking a weave label on the box, I don’t know how this won’t confuse typical consumers. Too many silos.


Well… You were able to provide a helpful explanation and link in just a few minutes. I would have liked the journalist to have dug into this issue. I certainly would have, as it is highly relevant to the headline.

This is, indeed, a disappointment. Perhaps the reasoning will come to light and the details may make it less of a problem than we fear… Or worse.

What the Hell is going on? Why in the world would Google do this?? I knew from the beginning Thread was going to be based on Nest protocols but why in the world would Google allow two separate stacks to go to market? We just cant be getting the whole story?? This has to be a mistake! If its not a mistake Apple will kick their ass. I just cant say how disappointed I am. Heads should roll.

@urman ?

1 Like

For as many things as Google does right, there are 10 things they do wrong. They do baffling things like this at times, which pretty much guarantee the failure of the protocol.

Fortunately we have zwave and zigbee.

Why would Alphabet do this?

I think the short answer is that they quietly made the decision that they need to put a security wall around the thermostat and smoke alarm features.

I’ve mentioned in the past that one of the few areas in home automation where I think random malicious mischief from techno punks could do real harm is in fooling around with the thermostat. That, and garage doors.

If some hacker hit 30,000 thermostats on a February night in Minnesota, they might actually kill people. Alphabet doesn’t want nest to be the brand in that story.

So while the weave standard allows compliant devices to talk to each other, “‘works with nest’ weave” severely limits which devices have access to which features of other devices. In particular, what other devices can request of the thermostat and the smoke detectors.

I think it makes sense in terms of their desired outcome, I just think the terminology is unfriendly for customers. If they had called it the “works with nest API,” developers would’ve understood there might be limitations that weave itself does not have. and I do think “works with nest” is a customer – friendly logo they can put on things.

So I think I understand where they were trying to end up, I just don’t really understand the path they’re taking to get there.

Then I’m not sure what Weave is bringing to the table. This is one of the reasons the light link profile exist in ZigBee. Phillips and others didn’t want their devices sitting on a network where the key is openly available (ZigBee Home Automation key is public). So they bought their way into the ZigBee council and came up with a new security model to meet there needs. Similar things happened to meet the needs of smart energy. So now we have all these different profiles based on ZigBee but they don’t route each others packets. Its an absolute mess in my opinion. The ZigBee organization has even admitted this was a problem and the new ZigBee spec is going to do away with profiles. It sure sounds like the Weave guys are repeating this same mistake.

Thread was supposed to address this and not follow the multi profile model it was one of its main charters. Maybe Weave on top of thread is not the way to go. ZigBee Clusters on top of Thread is starting to look better and better.


Surely they could implement and market this much better than saying “Nest Weave and Google Weave are different protocols”.

If certain devices need to be on a separate Weave “subnet” with some sort of firewall or VPN in the mix for extra security, then I think it is essential for the future of the Weave that Alphabet eliminates the confusion.

1 Like

Scary. That’s the world we’re about to live in too.

Actually, what we are all a part of right now is the beginning stage of a huge HA war. Remember betamax??? yeah, not many do… let’s hope ST and Zigbee HA 1.2 are the next VHS and the proprietary blends go the way of the laser disc movies.

Of course it may be that SmartThings and ZHA 1.2 are the next VHS… In a world of Blu-ray (HomeKit) and streaming (Weave)…

Just sayin’… :wink:

The fact that SmartThings is thread-capable may turn out to be the most important thing about it.

1 Like

Very true on all of those points… but just like the battle of Blu-ray and HD, the only reason an inferior product (blu-ray) won was nothing less than outstanding marketing and using that marketing skill to get the right financial backers with the biggest names… basically the same way that VHS beat betamax…

I’m just really hoping that Taco Bell does NOT win the restaurant wars that are coming soon.

1 Like

I bet marketing folks at Sigma Designs have huge grins on their faced this very moment. There’s certainly a lot of confusion in the enemy’s camp. :smile:

1 Like

In the past, yes, except the next generation of zigbee, zigbee 3.0 combines all of the profiles into one and gets rid of all of the nonsense. It’s been in beta for a well and is scheduled for release by the end of this year. Early reports are all good except for some concern on the part of the green energy group and it may be that there just isn’t backwards compatibility with that profile, but there is forward going compatibility.


It’s interesting that the alphabet onhub has dropped zigbee from their spec sheet but that may just mean it’s going to work as a thread device and A zigbee controller will have to meet up with some thread/zigbee bridge to talk to it. OnHub won’t be a zigbee controller on its own. That’s not terrible.

The real question remains how and if Zigbee and thread are going to coexist. Thread needs Zigbee’s extensibility and the low-power sensors if it’s going to be able to maintain a price advantage versus HomeKit. But can they really play nice together? It looks like onhub was going to settle that, and then they backed off of it. It’s going to be an interesting year, no question! But I don’t think the Sigma design people are feeling comfy right now.