Google Is Starting To Reveal the Secrets of Its Experimental Fuchsia OS (theverge.com) 75
At Google's I/O developer conference this past week, Android and Chrome chief Hiroshi Lockheimer offered some rare insight into Fuchsia, albeit at a very high level, in front of public audiences. The Verge reports: What we do know about Fuchsia is that it's an open source project, similar to AOSP, but could run all manner of devices, from smart home gadgets to laptops to phones. It's also known to be built on an all-new, Google-built kernel called "zircon," formerly known as "magenta," and not the Linux kernel that forms the foundation of Android and Chrome OS.
"We're looking at what a new take on an operating system could be like. And so I know out there people are getting pretty excited saying, 'Oh this is the new Android,' or, 'This is the new Chrome OS,'" Lockheimer said. "Fuchsia is really not about that. Fuchsia is about just pushing the state of the art in terms of operating systems and things that we learn from Fuchsia we can incorporate into other products." He says the point of the experimental OS is to also experiment with different form factors, a hint toward the possibility that Fuchsia is designed to run on smart home devices, wearables, or possibly even augmented or virtual reality devices. "You know Android works really well on phones and and you know in the context of Chrome OS as a runtime for apps there. But Fuchsia may be optimized for certain other form factors as well. So we're experimenting." Lockheimer provided some additional details at a separate Android fireside chat held at Google I/O today. "It's not just phones and PCs. In the world of [the Internet of Things], there are increasing number of devices that require operating systems and new runtimes and so on. I think there's a lot of room for multiple operating systems with different strengths and specializations. Fuchsia is one of those things and so, stay tuned," he told the audience.
"We're looking at what a new take on an operating system could be like. And so I know out there people are getting pretty excited saying, 'Oh this is the new Android,' or, 'This is the new Chrome OS,'" Lockheimer said. "Fuchsia is really not about that. Fuchsia is about just pushing the state of the art in terms of operating systems and things that we learn from Fuchsia we can incorporate into other products." He says the point of the experimental OS is to also experiment with different form factors, a hint toward the possibility that Fuchsia is designed to run on smart home devices, wearables, or possibly even augmented or virtual reality devices. "You know Android works really well on phones and and you know in the context of Chrome OS as a runtime for apps there. But Fuchsia may be optimized for certain other form factors as well. So we're experimenting." Lockheimer provided some additional details at a separate Android fireside chat held at Google I/O today. "It's not just phones and PCs. In the world of [the Internet of Things], there are increasing number of devices that require operating systems and new runtimes and so on. I think there's a lot of room for multiple operating systems with different strengths and specializations. Fuchsia is one of those things and so, stay tuned," he told the audience.
Pronounced Fucks Ya (Score:5, Funny)
There, fixed that for you!
Re: (Score:1)
Re: (Score:2)
Whatever it is (Score:4, Insightful)
its primary aim is to give Google a monopolistic foothold into as many devices as possible, to turn into spying and data collection machines.
And yes, it'll be really good and very shiny, because people stop caring about privacy and data security so long as the product is good and shiny enough - which is why all Google products are good, incidentally.
Re:Whatever it is (Score:5, Insightful)
+1
If it is anything like the Chromium project, it will be technically "open source" and yet not driven by any real community beyond the campus of Google. Then the production versions, that see any use in real products, will be binary blobs of locked-down mystery. And on top of that, a lot of proprietary services.
At least with Android, the base is Linux and at least some of Google's effort is/was funneled back into an actual open source project that has/had the potential to benefit everyone (not just Google) and they can/could only twist it but so much.
In a world of eroding privacy and rising security concerns, I will put my mindshare even more behind very open and transparent projects/systems (Linux, Firefox, LibreOffice, OpenSSH, Ansible, GnuCash, Kodi, Mplayer, VLC, PostreSQL, Python, Wireshark, VNC, Apache, Audacity, KDE, Gimp, etc).
Not so high risk (Score:3)
On the other hand, I'm not sure if there's so much need to be afraid of it, in practice.
- it's google. that means there's a very high chance that two years down the line they decide to pull the plug and kill Fuchsia.
- the "year of linux" (...on everything but your desktop) has come already a long time ago. The whole world has standardised around Linux as a kernel for most embed application (smartphones, router box, set top boxes, etc. even the firmware running inside Wifi-enabled SD cards for photo cameras)
Re: Whatever it is (Score:2)
Re: (Score:2)
Then the production versions, that see any use in real products, will be binary blobs of locked-down mystery. And on top of that, a lot of proprietary services.
And whose fault is that? Google who open sourced it? Or the community who doesn't give a damn and then complain about it.
Re: (Score:2)
Highly Secretive Open Source? (Score:5, Insightful)
So it's a highly secretive Open Source project? When do they spring it on us? Are the exploits and malware deeply embedded already? Where's the source repository? Is there even a reference implementation source tarball anywhere we can download?
Re:Highly Secretive Open Source? (Score:5, Interesting)
As you would expect from a group inside a Web company, most of their improvements seem to be in the APIs. Don't expect the high-graphics performance of BeOS. The APIs are pretty, but they come with reduced functionality, and as a side-effect, not compatible with POSIX or.....anything.
The 80s was when OSes went commercial, the 90s when they went open, and now it's really, really hard to make money on a new OS. When they first started several years ago, their messaging was "A new OS that will replace Linux in Android!" Now the messaging has been scaled back to, "pushing the state of the art..." by putting the OS in smart home devices, or wearables, which apparently no OS has ever done before.
There are a number of embedded OSes that are better than Fuschia, built by smarter people than the Fuschia developers. This is not a wheel that needs reinventing, although it can be very fun (and you learn a lot) to do so.
Re: Highly Secretive Open Source? (Score:5, Funny)
So what you're saying is there's no fuchsia in it?
Re: (Score:2, Informative)
No, it's not. If you bothered to look it it, you would see:
"The Fuchsia docs repository is no longer being mirrored to GitHub."
It's now behind a google login.
Re: (Score:2)
It's still public and open, though: https://fuchsia.googlesource.c... [googlesource.com]
Re: (Score:3)
It's a group of people inside Google who got frustrated with the realities/difficulties of dealing with Linux, and thought they could do better. [...] The APIs are pretty, but they come with reduced functionality, and as a side-effect, not compatible with POSIX or.....anything.
This sounds suspiciously like one of those perfect systems you get before your ststem has to interact with the messy, imprefect real world. Linux isn't perfect, and neither is the POSIX API. On the other hand they're both solid, battl
Re: (Score:2)
This sounds suspiciously like one of those perfect systems you get before your ststem has to interact with the messy, imprefect real world. Linux isn't perfect, and neither is the POSIX API. On the other hand they're both solid, battletested and have withstood the test of time pretty well. I think part of the problem is that the Android developers used the Linux kernel simply because it was there without any understanding of Linux or unix, used it badly, and have been slowly, badly, reinventing all the bits of Linux which they chucked out at the beginning.
Yeah, that's what I thought, too.
Like Redox ? (Score:3)
It's a group of people inside Google who {...} thought they could do better. {...} But somehow they managed to get Google to pay for their work.
{...}
Now the messaging has been scaled back to, "pushing the state of the art..."
So basically it is becoming what Redox OS is to the Rust comunity ?
Nobody in their right mind is expecting that Redox OS will suddenly replace Windows (or even Linux) on the desktop, but it's a fun experiment to push the limits or Rust, and prove that is indeed a system-level language, which in practice helps the rust language it self improve by drawing light to some limitations ?
Re: (Score:2)
that, and being Google (Score:2)
I expect Fuschia to disappear after a few years, based on the downgrade in messaging.
Well, that. And also being a Google project. A cancel-happy company with a long track record of killed projects :-P
Re: (Score:2)
There are a number of embedded OSes that are better than Fuschia, built by smarter people than the Fuschia developers. This is not a wheel that needs reinventing, although it can be very fun (and you learn a lot) to do so.
TempleOS [templeos.org]
rip terry
Re: (Score:3)
Re: (Score:2)
Tarball?? Dude, you don't get it. It's a new OS. It's new technology, CUTTING EDGE even! It's files aren't available to other operating systems because those operating systems are inferior in every way imaginable. And you think this OS will stoop down from it's high seat, to read your stewpid OS's files? pa-leeze.
As far as we know, Fuchsia doesn't even run on electricity, but rather on cool-points.
A new take (Score:2)
I'm willing to bet that one of the dazzling new features is the inability to uninstall bloatware.
Real secret is... (Score:2)
The real secret is that it's not fuchsia but purplish red.
Biggest secret (Score:5, Informative)
I'm going to reveal Google's biggest secret right now, keep this under your hats ok? Developing a full operating system that handles thousands of different kinds of hardware, storage, file systems, thousands of well defined interfaces, sophisticated memory operations, every network situation known to man, etc, etc, etc... is hard as fuck. Yes, that's the big secret, folks.
Re: (Score:2)
There's going to be some fancy abstraction layers, there.
Comment removed (Score:4, Interesting)
Re: (Score:2)
I've been seeing these comments all over slashdot. The "karma" or whatever is usually referring to the Linux code of conduct.
Basically a troll campaign to make OS development as Balkanized as US politics.
Time for a clean OS design (Score:1)
I don't blame Google for writing their own OS.
As others have said, this gives them control over its continued development.
Another plus is that they are free to design it from scratch to solve their problems. Linux is a pretty good OS, but it has become damn bloated compared to 20 yrs ago. This is acceptable for Linux because its target environment is at this point servers or powerful machines. It is certainly not optimal for small or realtime devices.
Strip out 50 years of design decision accretion and start
Re: (Score:2)
It's not, it's got a Dart JIT as a first class citizen.
Re: (Score:1)
There is nothing new in this startling revelation. As the source first released nearly 3 years shows, Fuschia is designed to be a lightweight secure OS for any connected device. They are trying capability-based security on a simple micro kernel. They picked Little Kernel because it's very small and Travis Geiselbrecht used it for Android's Boot Loader and TEE. An alternative would have been seL4 , which is capability-based, secure and mathematically proven to be correct.
Zircon might be leaner than the L
Re: (Score:2)
Re: (Score:2)
It is certainly not optimal for small or realtime devices.
Um...
well how small? I mean sure, it's not tiny like Free-RTOS tiny. It does go pretty small by modern standards, with quite a lot of effort (a few years ago) it could hit 1.6M RAM and 8M flash:
https://events.static.linuxfou... [linuxfound.org]
And for hard real-time, the PREEMPT_RT linux variant can actully do that. There are other RTOSs that can do better but modern PREEMPT_RT is actully a pretty respectable hard-realtime OS.
If you're using Intel You need to put WA
There's ALREADY a good alternative to Java (Score:2)
There's ALREADY a good alternative to Java for Android: C#, via Xamarin. And it's (sort of) cross-platform with iOS... you can write native UI code & share business logic, or go full-on cross-platform using Xamarin.Forms.
As an added bonus, you can still link to "Java" Android libraries (but sacrificing cross-platform compatibility).
In general, C# is a lot nicer than Java. Can we all say, "real, native, unsigned bytes"? Yay, no more endless casting and stupidity for using OpenGL-ES (which is downright EX
Re: There's ALREADY a good alternative to Java (Score:2)
So... why did Java add support for unsigned ints, instead of unsigned bytes? I can think of multiple concrete use cases where Java's lack of unsigned bytes make programming stupidly tedious & error-prone (OpenGL ES in particular), but can't think of ANYTHING major that benefits from unsigned ints that couldn't have easily been autoboxed by the compiler from arrays of unsigned bytes.
The fact is, C# got it right the first time... most primitive integer types are signed, but bytes specifically are unsigned
Android gives too much power to the user (Score:3)
I remember reading a blog about the internal fight at Google between the programming team and the avertising team. The programmers wanted to do a good job and create a secure OS with the user in control of the settings. The advertising team wanted a Swiss-cheese OS so that the secuuriity vulnerabilities can be exploited :
https://archive.fo/DLRqF [archive.fo]
"So far the ad team prevailed" is all we need to know.
Will be interesting to see if Google's actions will present an opportunity for things like Lineage OS or Sailfish.
Who are they kidding? (Score:2)
Re: (Score:2)
Fuchsia is about replacing Linux and Java in Android, Chrome OS and Wear OS.
During Google I/O, Google announced that Android should now be considered a Kotlin-first development target. So, they're already on their way to deprecating and replacing Java, with Kotlin.
Re: (Score:2)
Java and Kotlin are basically the same, they both compile to Java Byte Code before they get dexed to Dalvic code and both rely on the java foundation, aka java.lang.* and java.util.* classes etc.
nuts to invest in this (Score:2)
you would be nuts to invest time in this for anything serious.
before long google will find it takes too much effort or whatever and abandon this OS, like it does with so many apps.
next to that, you'll probably have next to no support from google if you would want to use fuchsia in your own (non google service powered) devices.
Re: (Score:2)