We expect web services to go away eventually. We expect physical products to keep working until they physically break. The problem is when companies sell physical products with a web service as a fundamental part of the product. Companies want to treat the web service like any other web service, while their customers reasonably expect their physical product to act like any other physical product.

Edit: I want to add that this is a really easy problem to solve. Support or create an open standard for the communication between the device and the server. Allow the device to be reconfigured to communicate with other servers. Then you can shut down your service and owners can keep on going. Better yet, make it so that our device can communicate on the LAN without needing a server at all. These problems exist not because they’re some inevitable fact of life, but because these companies see more value in not fixing them.

Are the physical devices not working after this change, or rather some of the applications developed for these devices need to switch to another API?

There is no other API to switch to, as Google controls the protocol, no open standard exists to replicate the Nest API server, and Nest devices don’t support targeting a self hosted API.

If you as a user or third party provider integrated with this API and rely on it for functionality, you have no alternative nor recourse.

Isn't "Works with Google Assistant" supposed to become the replacement API? It seems like they don't offer direct access to the data or the device but plan to allow "some" level of control? (I don't know the space, apologize if my questions are naive...)

> As a Works with Nest developer and partner, you will not be able to access or control Nest devices once the Works with Nest APIs are turned off on August 31, 2019. Moving forward, our team will focus on making Works with Google Assistant the most helpful and intelligent ecosystem for the home, enabling all of the products in your users’ homes to work together. We encourage all smart home developers to visit the Actions on Google Smart Home developer site to learn how to integrate your devices or services with the Google Assistant.

Part of the issue here is that many IoT products use web services as part of functions that don't need them at all. The cynic within me suggests that this is an intentional choice to allow for enforced obsolescence.

Beyond this, apps can be updated to remove or decrease support for older devices, and here our focus on trying to insist on auto-updated apps to fix security problems, or on apps that are designed to receive server-side updates, causes significant problems.

At some point, for example, I'm going to have to stop using the Philips Hue app, because I have an older Hue bridge. Support for the older bridge was at some point removed from the main app (and accompanied by an advertisement for the newer bridge), and moved to a legacy one, which seems to have a significantly slower response time. Now the legacy app has popups noting that web services will no longer support older hubs, and other services will be degraded... along with advertisements for the newer hubs. I expect at some point the app will simply be intentionally broken, entirely.

Fortunately, in this case, I can just use other apps, and other systems entirely, if need be, and at least the older hardware supports entirely local APIs (unless Philips decides to push a firmware update to break them?). I'm certainly not going to replace hardware that works with new hardware that provides no benefits, and throw away the old hardware; this would make no sense.

I’m really glad I bought a pantelligent because it only cost me $40 to learn that IoT is a giant scam.

I think you'll find it's Western Capitalism that's the scam and that IoT is still a reasonable idea.

I haven’t found that at all. Why don’t you explain, because I’m curious which non-western non-capitalist system prevents IoT devices from becoming worthless without warning when the vendor ceases support or fails altogether.

> this is a really easy problem to solve. Support or create an open standard...

We call it the tech industry but 90% of it is actually politics. Big money influences things the most but everyone basically looks out for their own interest, right down to programmers shilling their favourite language/framework. Thats what makes this a difficult problem.

I think that many people naturally expect Google to work like a lot of other companies, which are focused on satisfying individual customers.

Google, though, is unusual that most of their users are not Google customers. They are the product. So Google's relationship to most individuals is statistical, not personal. If they care about me as an individual, I have never seen the slightest sign of it.

I also think that a lot of things at Google get started and killed not on the business merits, but due to shifting internal politics. Most of my Google friends have left, so I don't have recent examples, but from what I hear Google Plus is a fine example of something launched because a powerful executive got excited. As the political winds shifted, though, it got neglected, and then eventually killed.

Another interesting example for me is Google Reader. I liked the product, but was never clear to me why they built it. They ignored it for a long time and then shot it in the head. Giving it away crushed the market for commercial RSS readers. Once they stopped, that market came back, proving that Google could have had a successful commercial product if they had bothered to run it like a real business.

So no, I don't think we expect them to continue all things indefinitely. But I would love to see a more clear logic for what they will continue to support, because from my perspective it's erratic. I would be very unlikely to build a company that depended on Google; despite having used them for 20 years, I still don't think I can predict their business strategy. In contrast, I've been using Amazon for a similar period, and I have a pretty good idea of what they'll do. I'm not worried they'll EOL the AWS services I use, for example.

Why should anyone let a Google self-driving car on the road? They will just end support for it whenever "internal politics shift."

Let's ban Google from working on life-saving medical devices, because they will just get bored and stop supporting the software.

On the other hand, EOLing an IoT pacemaker with short notice might bring some light to this issue.

> Why should anyone let a Google self-driving car on the road? They will just end support for it whenever "internal politics shift."

So what? That's between the car buyers and Google. Why should they be disallowed?

Some countries have laws which require devices to be fit-for-purposes that were advertised prior to sale. If consumers purchased a product based on claims made by Google, and those claims are not true, will regulators get involved?

US Federal Trade Commission, https://www.ftc.gov/system/files/documents/public_statements... (1983)

> The basic question is whether the act or practice is likely to affect the consumer's conduct or decision with regard to a product or service. If so, the practice is material ... because consumers are likely to have chosen differently but for the deception ... the Commission will find deception if there is a representation, omission or practice that is likely to mislead the consumer acting reasonably in the circumstances, to the consumer's detriment.

My girlfriend recently bought a sleep tracker device, from the store, new with shrink wrap, only to take it home and find out the service required to operate it had already shut down.

I've been seeing things like this in TJ Maxx lately.

I don't know if anyone at TJ Maxx even understands that these are not items they can sell, as they don't work anymore. I think they think they're just fancy watches.

They know.

They will get a fraction returned. So, they still profit.

Big companies like that need fines big enough to risk closing them down for doing stuff like this. No reason for them to care otherwise.

How do you know? Do you work at a TJ Maxx?

Every major company has company wide inventory. They have specialist buyers for different departments.

They either know, or have purposefully arranged things so the people who know enough aren't burdened with knowing everything. Neither is acceptable.

They could be idiots who don't know what they're doing, but then they wouldn't still be in business.

I returned cheap electronics to a UK store recently, it was very clear the entire batch didn't work, indeed it was probably why they were on offer. People will say "oh it was only a couple of quid, won't bother taking it back". Still on sale. I wouldn't be surprised if they hadn't reshelved the non-working ones I returned.

The thermostat I grew up with was retired (voluntarily—it still worked) when it was ~40 years old. If the thermostat manufacturer had somehow actively broken all the light switches, lamps, etc. in my parents’ house, even 40 years after the purchase of the thermostat, they would have been criminally liable.

Safety-critical() physical infrastructure is different than internet services. Applying laws for one to the other makes no sense.

() remember when all the Nests’ batteries died at the same time one winter?

I expect a "smart hub" to interoperate with a broad variety of devices. I expect existing banner features to not be revoked over petty turf wars. Google just made themselves irrelevant in yet another market.

Maybe they could develop an AI that predicts when Google services will get shut down so that users can make sounds decisions about how much to invest in them. Maybe it could even predict its own demise...

Google kills off products and services with wanton caprice and a complete disconnect from the needs of their users. While all services end, Google is liable to pull the rug out from under you at any moment. This is even more egregious if they kill off anything that physical devices use, as they basically just reaches out and broke a product you bought in good faith.

If they choose to make it dependent on an external service then yes, I expect them to keep that service available indefinitely, for free.

Consider it an incentive to avoid dependencies on external services.

Alternatively, given that you already know that they won't reliably keep those services available indefinitely, you could consider this incentive as a consumer.

Do we want to let the low morals of companies hamper technological development?

Perhaps there's a legislative solution. Although in the UK the Consumer Rights Act has no time-limit, so in theory of your thermostat stops working because the company decided to pull the plug then you can get refunded, at least.

It's hard for me to buy into the idea that allowing a company to launch a product with a given feature hampers technological development. Requiring that a company runs an API forever if a product depends on it (and commit to refunds if that doesn't work out) seems like it would raise the barrier of entry to release such products, thus resulting in fewer products in this space (especially from smaller companies). That sounds like it would hamper technological development to me.

The product could be "a thermostat that works until 2021" though, then companies wouldn't be hamstrung with supporting old devices beyond the promises they'd made? Like giving a warranty, or making a parts/repair promise.

I'd like to see something like a requirement to publish interoperability details, or code that runs on a major OS that operates as a server, to facilitate long term user-led support, too.

Laying all my cards on the table I'm kind of ideologically opposed to a 300 million person government forcing any regulations whatsoever upon the entire population rather than allowing for local control of laws, but if I were to put myself in the shoes of someone who accepted the basic premises of American society I would see your proposed minimum EOL statement as a totally reasonable consumer protection rule with relatively little burden on the seller. This isn't what I took the other posters in this thread to be proposing, though, and I still think it would have a slight hampering effect on technological innovation (going back to GGP's argument)[0], I can just see arguing that this would be worth the consumer protection provided.

The requirement to publish interoperabilty-details/code/etc. is something I see as far more burdensome, perhaps in some cases to the point of being impractical, which is sad because I actually value these properties quite highly. I don't own any products that rely on the API of a specific company to continue functioning, and as an individual consumer I reward companies that do release such details (albeit in a statistically insignificant way). The simple, obvious reason that this is burdensome is because the release of code helps competitors, and requiring a company to help competitors makes it more likely that they will decide to invest in some less-regulated industry instead. There's a logical reason that most software companies aren't open-sourcing their products, unfortunately. Also, and I disagree with this yet don't want to force my opinions of the world on other peoples' infosec decisions, some people argue that closed source software is more secure than open source software due to security through obscurity. More rarely, I can imagine edge-case IoT devices that rely on highly specific infrastructure where interoperability just isn't reasonable, and the question of which layer interoperability should be provided on is non-obvious. Imagine if SpaceX launched a swarm of satellites measuring some data about the Martian atmosphere and released a product that could tell you the measurements when asked. Do they need to provide the details of the API between the product and the Earth-based relay? Do they need to provide the details of the connection between the Earth-based relay and the Martian satellite swarm? Do they need to provide the details of the internals of the satellites themselves? What if they're the only company anywhere near being able to supply this service, and the technology will almost certainly be obsolete by the time another company manages to launch a Martian satellite swarm? I feel like this probably isn't the best example, but there might be better examples I'm too dumb to think of at the moment, and not at least considering those known-unknown potential scenarios might be harmful to society as a whole (but I think I've articulated this last point worse than the rest of my comment, and I'm not entirely convinced of it myself).

[0]Edit: on further thought, this hampering effect should only really exist if there is a minimum time from date-of-sale that the minimum EOL must extend to; if I can meet the requirements while selling a product that is only guaranteed to work until yesterday then it shouldn't have any real effect other than consumer awareness, but it also kind of takes the teeth out of the proposal.

The fact that there is a Google Graveyard does not imply that someone thought that every service that Google creates should last indefinitely. I view it more as a useful reminder of how frequently Google does this - a good idea if you're about to make something Google does part of your workflow or habits.

There are also some products in there that were killed in a fashion that seemed capricious and difficult-to-explain in terms of popularity, usage and continued cost to Google - notably Reader and Code Search.

If I am the product, I demand it be there forever!

No company is, and expecting any company to continue spending money on maintaining a service even if they've decided it isn't profitable is pollyannaish.

To that end, I'm not sure I see the value in framing this specifically as a Google thing. Google is maybe the most visible because of how many products it has, and how popular many of them are. But really, this sort of thing is a problem that's inherent to any IoT devices that rely on proprietary cloud services, in whole or in part.

There's a difference between an IoT manufacturer going out of business, which causes related software services to terminate, and an IoT manufacturer failing to deliver on device claims made before the sale took place.

There's also a difference between how we want things to work, and how they actually do.

These services typically include some language in the EULA to the effect of, "We can disable parts of this device from our end, for any reason, at any time." And agreeing to those is almost certainly precondition for using the device.

Which absolutely sucks. But it's the legal and consumer protection environment we live in right now. Pretending it isn't is -- hang the downvotes, I'm gonna double down on this -- pollyannaish. Until things change, your only real recourse as a consumer is to refuse to buy products that are sold under those sorts of terms. Which, unfortunately, does mean steering clear of IoT devices that rely on proprietary cloud services.

EULAs do not override national or state regulations on device fitness for purpose.

I expect a company to continue a service for the foreseeable future (lets say the normal duration of the warranty on the product) if it is necessary to use a product i just bought. Otherwise I expect a refund or it feels very much like fraud to me. Legislators should get involved here sooner then later.

There is a huge difference between the product naturally breaking a while after the warranty expires, and the product deliberately self-destructing as soon as it's no longer the manufacturer's problem.