Google Working On New "Fuchsia" Operating System, Powered By Magenta / LK Kernel

[Hi-Angel]

Another bit of info: Magenta have no compatibility with Linux modules.

[Hi-Angel]

Quote from the link, gotten on IRC:

Some useful bits from IRC:
[16:21] <ocdtrekkie_web> Why's it public (mirrored to GitHub even) but not announced or even documented what it's for?
[16:22] <@swetland> ocdtrekkie_web: the decision was made to build it open source, so might as well start there from the beginning
[16:22] <lanechr> ocdtrekkie_web: things will eventually be public, documented and announced, just not yet
[16:23] <@swetland> currently booting reasonably well on broadwell and skylake NUCs and the Acer Switch Alpha 12, though driver support is still a work in progress
[16:24] <@travisg> yeah and soon we'll have raspberry pi 3 support which should be interesting to some folk

[Hi-Angel]

Well, if anybody curious: I joined #fuchsia channel on Freenode, shared the article 3 hours ago (and comments link, because the "Add comment" link was broken). Nobody of them commented, neither here, nor there; link was at the visibility all the time. ≈½ hour ago I asked a few questions, and got a couple of very tight answers, which is that unlikely the system is going to use Wayland, quoting "I would assume, given the code written so far, it will be specific to fuchsia"; and, judging from later comments, (which, though, doesn't seem were from devs), it looks like devs are tied by NDA.

[duby229]

That makes no sense at all. Honestly I never really understood why in the fuck google chose to use java. So stupid.

Anyway, LK seems like it would absolutely need hardware manufacturers to supply proprietary drivers for hardware enablement. I doubt many will backport existing drivers so this is unlikely to be a threat to linux for another 7-8 years at minimum. Who knows what will happen between now and then.

[bregma]

Ah, Google is writing a couple of new kernels, one for IoT (LK) and one for phones and desktops (Magenta) that extends the IoT one, both with Apache licenses just like Android. That will let them get rid of the last bit of GPL software in their stack (which will make OEMs happy) and give them much more control over the entire software stack on every device in your life.

Imagine it: a US corporation subject to US law enforcement agencies and that earns its revenue from watching everything you do, controlling every aspect of pretty much every device around you. It's enhanced by a license that lets anyone, including non-US corporations, governments, agencies, or organizations, substitute their own code and still claim it's original. Makes you feel warm and cozy, yeah?

Also, the common kernel base and the broad applicability, together with the blurring between ChromeOS and Android, is clearly convergence in action. It's clearly the way of the future. You have the choice now on the convergence front: semi-open-source full-stack corporate lock-in with Google, messy closed-source cathedral-style corporate lock-in with Microsoft, or GPL-licensed Free software with Ubuntu.

[rubdos]

A friend of mine said he thought the reason that Anroid was written in Java was to get rid of the Linux Kernel at some point.

I call EEE. It's BSD licensed, so it'll be a lot easier to make things proprietary for vendors.

[UseLinuxNotWindows]

LK looks more like an alternative to FreeRTOS (and other embedded OS's) than a serious competitor for Linux. A huge amount of time and effort has gone into making the Linux kernel what it is today. That is why Apple did not write OS/X from scratch but took OpenSTEP (later released by Apple as "Darwin") and built on top of that. Even if you created a lightweight POSIX compatible OS just to run Java (which is in effect what Android is), Google would have a huge uphill struggle for it to gain traction. I suspect that the new OS is for IOT devices and other light weight devices rather than a replacement for Android's OS layer.

[NateHubbard]

https://fuchsia.googlesource.com/magenta/+/5a5236c1b2a74c77ef196e7f083d0863dc89f667

Interesting. I wonder what exactly they are planning with this.

[Hi-Angel]

I must say, the project is in its very early stage. But looking at the repository, I, probably, a bit mistaken that the ecosystem isn't going to benefit. They're not going to rewrite absolutely everything, I see there's ffmpeg, zlib, freetype, go, cmake, etc.; this also mean that this kernel has POSIX API. I'm curious what protocol windows system going to use (i.e. if its Wayland, or a custom one), but code of Compositor hasn't yet a single line of code.

That's all, not much to see yet, actually.

[Hi-Angel]

Michael "Add comment" link in the article is broken