Google's Fuchsia OS Is Rolling Out Widely To 1st-Gen Nest Hubs (9to5google.com) 40
More owners of the first-generation Nest Hub are receiving the update to Google's Fuchsia operating system as it expands beyond the Preview program. 9to5Google reports: Back in May, Google formally released Fuchsia, its effort to develop a "not Linux" operating system from scratch, which has been years in the making. The first device to receive the new OS was Google's 2018 smart display, the Nest Hub -- not to be confused with the second generation Nest Hub with sleep tracking released earlier this year -- taking it permanently off of the existing Linux based "Cast OS" without negatively affecting the UI or experience. At the time, the rollout of Fuchsia was limited to a select few devices that were enrolled in the "Preview program" available to all devices via the Google Home app. Over time, the number of Nest Hubs running Fuchsia was expanded, with Google no doubt watching carefully for any issues with the upgrade.
Late last week, Google updated a support page to reflect that the Nest Hub has received a new firmware update for both the Preview program and all other devices. Specifically, the first-gen Nest Hub is now receiving firmware version 1.52.260996. Google confirmed to us that this update does indeed include the upgrade to Fuchsia. All goes well, this means in a matter of days, all first-gen Nest Hub devices in households around the world should be running Fuchsia.
Late last week, Google updated a support page to reflect that the Nest Hub has received a new firmware update for both the Preview program and all other devices. Specifically, the first-gen Nest Hub is now receiving firmware version 1.52.260996. Google confirmed to us that this update does indeed include the upgrade to Fuchsia. All goes well, this means in a matter of days, all first-gen Nest Hub devices in households around the world should be running Fuchsia.
Sonos? (Score:3)
I wonder if this is in any way related to the IP infringement suit by Sonos alleging Google stole their IP for the smart speakers?
Re: (Score:2)
Do we need another OS? (Score:3, Interesting)
Re: (Score:2)
I have no idea what Google's goals are, but I do think that we need an open source 'not Linux' OS for places where Linux doesn't make sense. I'd love to see something like QNX, for example.
Re: (Score:2)
There's FreeROTS and Zephyr. Doesn't seem like much room for "OpenQNX" with those and Linux.
Re: (Score:1)
FreeRTOS is little more than a scheduler and a collection of synchronization primitives (mutex, semaphore, etc).
AFAIK, QNX is much more complete/mature, featuring even a graphics stack.
I was surprised to learn Qt Quick has a FreeRTOS compatible version though. [doc.qt.io]
Whether extending FreeRTOS/Zephyr is better than "OpenQNX", I don't know.
But it seems to me that there is room either direction.
Linux on a low-power embedded system will probably always be overkill at this point.
Re: (Score:2)
"FreeRTOS is little more than a scheduler and a collection of synchronization primitives (mutex, semaphore, etc).
AFAIK, QNX is much more complete/mature, featuring even a graphics stack."
What is your point? Different things are different. "An open source 'not Linux' OS for places where Linux doesn't make sense" can certainly be something that is merely "a scheduler and a collection of synchronization primitives", in fact that may be precisely what you want.
"Linux on a low-power embedded system will probabl
Re: (Score:2)
It amazes me how resistant modern programmers are to take ownership of their entire platform for embedded applications. My view has always been that "Linux on a low-power embedded system" is always overkill, laziness even.
I couldn't agree more.
Re: (Score:2)
Benefit to Google device users? Little, though an OS optimized to lightweight battery driven hardware and real-time processing of audio and video can have some notable performance advantages. Discarding legacy hardware support could simplify and even security harden the kernel. Whether such potential benefits are of any practical use and justify using a distinct kernel and libc are interesting questions.
Re: (Score:3)
I suppose if Fuschia was a little more efficient or responsive it might have some imperceptible bene
Re: (Score:2)
but consumers still won't care that it is the OS enabling this
Why is that germane? If Google makes the enclosure out of a different type of plastic that's cheaper, and they can drop the price, consumers won't care that it's the type of plastic that's enabling this. They care that it costs less.
In the case of a more efficient OS, they care that it uses less electricity, or gets less hot. Or any of the other advantages Google gains to improve the customer experience like more agile updates to the kernel which
Re: (Score:2)
Why is that germane? If Google makes the enclosure out of a different type of plastic that's cheaper, and they can drop the price, consumers won't care that it's the type of plastic that's enabling this. They care that it costs less.
Thanks for agreeing with exactly what I said.
Re: (Score:2)
The general public doesn't care what kernel / operating system powers their watch, smart device, phone or tablet.
They care how long the battery lasts for a given device size. That's the primary benefit of optimization in mobile devices.
I suppose if Fuschia was a little more efficient or responsive it might have some imperceptible benefit to the user, but the downside is it might crash all over the place or even brick devices
Or it might be more stable. I think that's likely, actually. I remember the days when Linux was rock solid, but it's gotten too big and complex, and those days are now long gone.
Longer term, perhaps the OS is designed run more efficiently on low power devices but consumers still won't care that it is the OS enabling this.
Sure. They care about how the device performs, not why it does so. That's irrelevant to the point that a lighter, better-optimized OS could provide benefits that users appreciate.
Re: (Score:2)
Re:Do we need another OS? (Score:4, Informative)
The way it was explained to me was that they wanted to be able to get control the OS away from the silicon vendors. Right now, if you make a phone with a Qualcomm processor, you're stuck using an old version of Linux that will never get updated, because that's all qcom will provide.
So they made a microkernel OS that they think will allow them to use hardware drivers from qcom that never get updated, but still be able to update the rest of the kernel.
Re: (Score:1)
That makes sense to me.
A micro kernel does seem like the right solution to the lazy/greedy vendor problem.
Not that I'd trust Google, or that they wont can this in two more years...
Another solution would be to have OSS kernel modules for all those SOCs, so we can recompile the kernel and be on out way.
Re: (Score:1)
And to push the UI in Flutter, which they widely across their products.
Re: (Score:3)
Also it allows them to pick any license they want for drivers. There will only be proprietary licenses for the drivers, and they will only be available in binary form.
To get Linux support for a new device running Fuchsia, someone will have to reverse engineer everything. This is unlikely to happen for the majority of devices.
Re:Do we need another OS? (Score:4, Insightful)
The net benefit to Google? Not having to publish all of their source code. See also Chrome.
Re: (Score:2)
Bingo.
Re: (Score:2)
>"Not having to publish all of their source code. See also Chrome."
More like see also Chrom*, because now it is essentially all multiplatform browsers that are not Firefox.
And it is not only not having to publish code, but having almost absolutely control over it all.
Re: (Score:1)
Lockdown. Complete control and no open source like Linux/AOSP. DRM and TPM up the wazoo and then some.
Oh, you meant benefits for *you*?
Re: (Score:2)
Linux was not designed as a mobile operating system and while they have done a decent job of using it as one, a lot of the work has been working around old models for how things should be done. For example Linux networking was not in good shape to cope with highly mobile devices switching between different WiFi access points and mobile networks while maintaining TCP level connections, and they had to do a lot of work to improve it.
Over the years more and more of the Linux stuff has been replaced. For exampl
Re: (Score:1)
It's just got to the point where they don't rely on Linux (you can run Android apps on other operating systems, including Windows 11) and they have built a kernel that is more suited to their needs, rather than adapting Linux.
Do you think Google plans to replace Linux with Fuchsia in Android? That's been my assumption, but I haven't seen much corroboration either way.
Re: (Score:2)
Eventually they might.. It needs support from SoC manufacturers.
Re: (Score:3)
I am trying to understand the net benefit here vs Linux.
I'm not sure about Nest's rationale, but I can tell you that the Android Security Platform team has a long-running joke that we want to "drown Linux in a bathtub". Linux is by far the largest source of serious security vulnerabilities in Android, because Linux is huge, complex and monolithic. If you find a code execution vulnerability anywhere in that enormous codebase -- including in any driver, many of which are proprietary -- you own the entire system (well, mostly; making the system secure in the face
Re: (Score:1)
I find this fascinating.
Maybe if Nest proves that Fuschia is a solid platform, we can begin to have serious discussions about moving Android. Maybe. I'm not holding my breath.
The thing is, Fuchsia is a big bet, no? It must have one or more goals in mind, right? If so, is replacing Linux in Android a goal? A non-goal? A happy result if it happens? As an outsider, I'm just so curious about this!
Re: (Score:2)
I find this fascinating.
Maybe if Nest proves that Fuschia is a solid platform, we can begin to have serious discussions about moving Android. Maybe. I'm not holding my breath.
The thing is, Fuchsia is a big bet, no? It must have one or more goals in mind, right?
Google does a lot of things just to see what's possible. Also, goals change as more is learned. And as people join/leave.
A core premise of Google's business has always been that if you build something useful, there will be a way to monetize it. This strategy started with search, which was fully operational and the basis for founding a company well before they hit on advertising as a way to monetize[*]. Projects often have no explicit business plan beyond "let's build this and see if it's something people
Re: (Score:2)
> I am trying to understand the net benefit here vs Linux.
Google doesn't have to convince anybody that their ideas have merit.
Re: (Score:2)
Waiting for their next iteation of the device, code named Uptias before I pass judgement.
Re: (Score:2)
start the clock (Score:4, Insightful)
INB4 *extreme* lockdown. (Score:1)
In a way that just was not possible with Linux and all that pesky AOSP freedom.
Re: (Score:3)
Re: (Score:2)
Google could have always kept Android closed source even if they used a Linux kernel. Most of the user land is Apache / BSD licensed but that was their decision which had nothing to do with the kernel.
Yep. I see no interest in Android to closing anything. Quite the opposite, I see more and more Android teams doing all of their work directly in AOSP, rather than doing it in internal repos. The code becomes public either way, but if it's done in internal repos it doesn't see the light of day until the "dessert" release (annually, basically), as compared to everything being published immediately in AOSP.
Of course, things could always change, and I don't see everything, but from where I sit, Android finds
Obligatory Star Wars Quote (Score:2)
Re: Obligatory Star Wars Quote (Score:2)
And what, Android is not one?