Google's Mysterious Fuchsia OS Makes its Public Debut (techhive.com) 68
Big under-the-hood changes are coming to Google's original Nest Hub, even if most users won't ever be aware of what's happening. From a report: Starting today, the open-source Fuchsia OS will start rolling out to first-gen Nest Hub displays, according to 9to5Google. In the works since 2016, Fuchsia will land first on Nest Hub devices enrolled in Google's Preview Program, before arriving more widely on non-Preview Program displays. Don't expect the user experience to change much, though. 9to5Google notes that the look and feel of Fuchsia OS-powered Nest Hubs will be "essentially identical" to what it was before.
OK, so what's the big deal about Fuchsia, then? It's a new, open-source OS that's decidedly not based on the Linux kernel, as Android and Chrome OS are. Instead, Fuchsia is based on Magneta, which (as we described it back in 2016) is "combination microkernel and set of user-space services and hardware drivers" with a "physics based renderer" that can power graphical user interfaces. Because it's an open-source project, Fuchsia's existence has been well publicized over the years, although its purpose has been harder to fathom; "out in the open" yet "shrouded in mystery" is how we aptly put it. With its arrival on the original Nest Hub, Fuchsia is taking its first tentative steps out of the lab and into the hands of actual users, even if those users aren't aware of the new OS.
OK, so what's the big deal about Fuchsia, then? It's a new, open-source OS that's decidedly not based on the Linux kernel, as Android and Chrome OS are. Instead, Fuchsia is based on Magneta, which (as we described it back in 2016) is "combination microkernel and set of user-space services and hardware drivers" with a "physics based renderer" that can power graphical user interfaces. Because it's an open-source project, Fuchsia's existence has been well publicized over the years, although its purpose has been harder to fathom; "out in the open" yet "shrouded in mystery" is how we aptly put it. With its arrival on the original Nest Hub, Fuchsia is taking its first tentative steps out of the lab and into the hands of actual users, even if those users aren't aware of the new OS.
Polls are open for when it will get dropped (Score:5, Insightful)
New Google Product. Limited Use Cases. Based on new code, in which Google will need to invest time and money for a larger set of support.
I am guessing Google will drop this product in July 2023
Re:Polls are open for when it will get dropped (Score:4, Informative)
Well it is open-source [fuchsia.dev] and after all isn't one of the principles that no one could take it away from you, not even Google.
Re: Polls are open for when it will get dropped (Score:3)
Comment removed (Score:5, Insightful)
Re: (Score:3)
TVs with Android have been a failure, for example. It's a decent mobile phone or tablet operating system, but beyond that it's not been very successful.
Google should have known that even the premium brands would put extremely slow CPUs in their smart TVs. But the OS is not really optimized for such slow hardware.
It doesn't help that most TVs that implement Android TV have essentially two different OSs. If the TV functions were all implemented under Android TV userspace it would be a little more seamless. As it is, most I see flip back and forth between "TV mode" and "Android mode"
Re: (Score:2)
Google should have known that even the premium brands would put extremely slow CPUs in their smart TVs.
For the most part the CPUs in smart TVs aren't slow. The reality is they are bogged down with bloat and unimaginably poorly written code for what should be a very simple piece of embedded software.
Re: Polls are open for when it will get dropped (Score:2)
They already support two distinct operating systems, Android and ChromeOS. Neither are remotely alike, the only thing the two have in common is that they both use the Linux kernel. ChromeOS also has a sometimes-implemented / sometimes-not Android compatibility layer, but that's as far as it goes.
False. Chrome OS is actually based on Android, even uses the same surface flinger renderer, only a lot of Android specific bits (like ART/dalvik, GMS, and a few others) are removed. Adding support for Android apps only requires adding a subsystem that includes all of that.
Supporting a third doesn't seem extreme if it's for a different purpose. Android has been traditionally their "Put it everywhere" OS, and, well, it doesn't work like that.
It's meant for consumer devices, and it does pretty well at that.
TVs with Android have been a failure, for example. It's a decent mobile phone or tablet operating system, but beyond that it's not been very successful.
Based on what? It's the third most common OS used in smart TVs, and in Europe and Asia it's the most popular OS for cable boxes. And what's easily the best streaming box out
Re: (Score:3)
Android TV is a huge success. The market is flooded with hundreds of devices, from the cheapest STBs to Nvidia Shield to many big TV brands. Every streaming service supports it.
Some cars come with Android too now, and while it's early days it's been well received.
Why didn't they buy Blackberry? (Score:2)
If they wanted a microkernel, wasn't the QNX-based kernel already in every target market?
Designing microkernels does not appear to be easy, as anyone who has watched the HURD over the years will know. Why reinvent the QNX wheel?
Re: (Score:2)
Because why not add another one to the list [osdev.org].
The primary motivating concern behind Fuchsia seems to have been security, and the Google engineers felt that they could do it better than the alternatives.
Re: (Score:2)
Creating Micro-kernel OS's is trivially easy. What's hard is to make the result something other than a steaming pile of shit performance.
Every micro-kernel developed to date spends 90% of it's development time poking holes in the micro-kernel to improve performance above big pile of shit. Some of MS's best OS designers spent a decade poking holes in the NT kernel to improve it's performance enough that literally anything else including DOS didn't eat it's lunch.
MK's are easy to create, they are a mountain o
Re: (Score:2)
The purpose of FusciaOS is for Google to get rid of the GPL licensed Linux. That and Google constant not is house syndrome where they constantly reinvent the wheel because the previous wheel wasn't developed by Google.
FusciaOS is the result of Google lingering issues with NIHS in both Android and ChromeOS along with their dislike of GPL software (or preference for BSD license).
But I tend to agree, it will be around a couple years, they will realize the level of investment needed to bring it even 50% of the
Re: (Score:2)
That's a really big IF. Like impossibly big IMO.
The linux kernel has several billion dollars invested in it (probably double digit billions at this point) Microkernels are easy to start but end up costing just as much as Linux has to serve the same markets with equal performance.
Like others have said, they would have been better off to buy QNX, BSD license it, and take the hundreds of millions spent making that micro-kernel work as a leap ahead. Instead they took a project with a grand total of a few hundre
Re: (Score:2)
Micro Kernels are mostly developer OSes. They excel (-ed) in developing modules that traditionally would reside in kernel space, e.g. a network stack or filesystem. Because they avoided the need to build a whole new kerne with that module and boot it (and how actually do you debug a kernel?) for testing.
With a micro kernel filesystem, you just launch that file system process and debug it in user space or one level higher than the kernel.
However with much faster machines, creating a new kernel, is no longer
Re: (Score:2)
With rolling this out to even a single Nest product, it already has a larger installed base than GNU/Hurd.
Re: (Score:3)
Re: (Score:3)
None of the links to "open-source" so far have lead to a license. (The one in the article is totally borked, starting off "hhttps:mobile....",) So I'm not really convinced that I can trust the license. And since the company is Google, I'll take some rather thorough convincing. The GPL would do it, but even the BSD or MIT licenses would be acceptable. Something they've had their own lawyers whack together wouldn't even rate looking at.
Re: (Score:2)
Re: (Score:2)
Well, BSD and MIT are acceptable. I'm not sure about the Apache license, as I've never really investigated it seriously. Of course, that definitely makes it "open source" rather than "free software", with the implicit position that it may be closed as some point in the future...but that wouldn't affect code already released.
OTOH, a mix of licenses is a bit worrisome. Even if each one is acceptable, one doesn't really know how they interact. If Google had the reputation they had when they started this pr
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
I am guessing Google will drop this product in July 2023
This seems like it's a trial by fire... but on a small scale. If it proves to be more trouble than it's worth then it will be put on life support before pulling the plug. If it survives the trial then it will receive additional focus. The original endgame for Fuchsia was to replace the Linux kernel in Android because of pesky driver issues that come with a monolithic kernel. Since it's not POSIX-esque it needed to build out replacement libraries that needed some missing functionality. I'm assuming the
Open source? (Score:4, Informative)
Re:Open source? (Score:4, Insightful)
All that illustrates is the difference between hardware and software. As well as the letter and the spirit of the law. Hence Tivoization [wikipedia.org] and GPLv3.
Re: (Score:3, Informative)
This keeps "n00bs" away from rooting and all the fun stuff, but gives more adventurous people all the power they need.
Re: (Score:2)
Re: Open source? (Score:1)
So it translates to "TiVoization".
And this is why people who oppose the GPL3 are actively conteibuting to destroying the open souce community and freedom.
I dream of completely free software with no licenses too, but I know it is an unrealistic dream, and regulation means nobody stops the biggest troglodyte from using his "freedom" to hit you with a club and take all your mammoth meat and freedom.
Link hideously malformed, delete and start over (Score:2)
I can't even post it here because Slashdot thinks it's ascii art (what year is it)
Re:Link hideously malformed, delete and start over (Score:4, Funny)
It is 2021. A year after where Slashdot was littered with racist posting, because for some reason people thought it was cool to just get those lefties mad.
Re: (Score:2)
Ok Mr conservative.
Nike Shoes!
Kneeling at a football game!
People with body parts going to bathrooms!
Fair and Open Elections!
Re: (Score:2)
irrelevant to having a shoddy amateur ui that can't even handle cutting and pasting backquotes, nor unicode support
and people point to slashdot to claim Perl is still relevant, hah. Relevant to low effort hacks.
Re: (Score:2)
Re: (Score:2, Informative)
If you posted it as a link, rather than plain text, you wouldn't have that problem.
Google's Not-so-secret New OS [slashdot.org]
And so it begins... (Score:2)
magENta you dufus (Score:1)
Re:magENta you dufus (Score:5, Funny)
No error! Fuchsia is a color, sure. But I'm pretty sure Magneta was an even-more-evil girl version of Magneto is a parallel Marvel universe. She was designed back when feminism wasn't as much in vogue, so she that didn't make it past the focus groups. (RIP)
Fuschia? (Score:2)
Does that mean I'll have to learn how to spell the color?
Pretty sure that's a hard stop for me.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Also a character from The Rocky Horror Picture Show [imdb.com].
Re:Fuschia? [sic] (Score:2)
The proper pronunciation isn't possible in English as the sounds you'd use for Fuchs can only be approximated in most English dialects. And it's also useless as a guide as Fuchs is pronounce closer to English "fox" than how the vast majority of English speakers pronounce fuchsia.
People will have to accept that we're not going to get this right. That the general population in the English speaking world won't be speaking German (or Latin) probably ever. And that because we butcher the pronunciation we'll also
Everything open? (Score:2)
Re: (Score:2)
You could have open source firmware on a device and not be aware of a secret microphone in the hardware. You're not likely to find it just by looking at source code if we never initialize the hardware, or just blast some magic values into the hardware that puts it into a reasonable power-on-reset state. Of course once we do turn it on you'll find out, so that's something at least. Doesn't keep the NSA or Google from exploiting your "open" device to do nefarious things. I wonder how many times law enforcemen
Re: (Score:2)
The headline doesn't make sense (Score:3)
It says the OS is open, and shrouded in mystery? So can you not view the source?
Re: (Score:2)
https://fuchsia.googlesource.c... [googlesource.com]
Because you "many eyes" Linux folks don't bother actually reading it. The heartbleed bug in SSL proved many eyes to be false.
Re: (Score:1)
... whereas thousands of exploited bugs in Microsoft Windows and in Windows apps by Microsoft and others prove there's little security advantage to keeping the source closed?
Anyway "many eyes" is an idea, a principle to be applied, not something to be proved or disproved. You can even use the same principle within a closed source shop, with whoever does have access -- the more people are reviewing code and testing for bugs, the more bugs are likely to be found (and fixed) before any damage happens.
Re:The headline doesn't make sense (Score:5, Funny)
It says the OS is open, and shrouded in mystery? So can you not view the source?
I'm a developer. The code I wrote LAST WEEK is often shrouded in mystery to me.
Re: (Score:2)
It says the OS is open, and shrouded in mystery? So can you not view the source?
I'm a developer. The code I wrote LAST WEEK is often shrouded in mystery to me.
Ah! You must be a C programmer! Yes, good old C, the "write-only language". :)
--
.nosig
Re: The headline doesn't make sense (Score:2)
I always hold up Perl as write only. C just isnâ(TM)t that bad.
Re: The headline doesn't make sense (Score:2)
Re: (Score:2)
I think they mean they don't know why Google would create a new OS when it already has two successful ones.
Looking at it they seem to be trying to improve the security model. SE Linux isn't the be-all and end-all. The new model also makes updating parts that were in the Linux kernel but are now micro services easier, independent of the device manufacturer.
Re: (Score:1)
The headline doesn't make sense
Msmash is pleased to make your acquaintance.
Switch Names With Other Google Products (Score:2, Funny)
Google is attempting to escape the GPL (Score:2)
Google is attempting to escape the GPL. That is what this is all about.
I'm in favor of FOSS all the way. We're almost there with fully open source hardware like RISV-V etc coming on too. This is good for privacy and security.
There are still attacks possible though, even on fully open source hardware. There are even sub-gate level dopant attacks that can't be seen even if you decap and scan the chip with a STEM! So ... don't rely on computers to secure anything that can cost you your life or liberty.
For eve