> This document proposes a mechanism for running unmodified Linux programs on Fuchsia. The programs are run in a userspace process whose system interface is compatible with the Linux ABI. Rather than using the Linux kernel to implement this interface, we will implement the interface in a Fuchsia userspace program, called starnix. Largely, starnix will serve as a compatibility layer, translating requests from the Linux client program to the appropriate Fuchsia subsystem.
So they will be able to run unmodified Linux binaries which should allow for a smooth transition from Android/Linux.
I'm not even sure that they'll rebrand Android at that point, as kernel for users is invisible. My guess is they'll just replace Linux with Fuchsia.
Hell, even their approach is to retrofit Linux binaries rather than trying to modify Android to run on Fuchsia. Android is a different beast itself, I'd rather see a better, reworked Android runtime and HAL that is idiomatic to Fuchsia rather than this one. The same could be said for ChromeOS too.
But that seems to be exactly what they are doing. They're surely porting the Android runtime to Fuchsia, or even already ported it.
Remember there are Android APKs containing native Linux banaries (made with Android's NDK) that you should be able to run. That is the whole main point of Fuchsia' starnix. I think this is even mentioned on starnix's RFC.
Both. If CastOS (which is Linux-based) for the Google Nest Hub was already replaced via an update, it is completely obvious that Fuchsia is not only here to stay, but will definitely replace Android AND ChromeOS.
I think Unix has had its day and its quite refreshing to see a new OS that is designed for security and is actually optimised for modern hardware and CPU architectures from the start.
Downvoters: So Google did NOT just replace the Linux-based CastOS on the latest Google Nest Hub update with Fuchsia and it is still an 'experiment' that some were still hoping for? [0] Are you prepared to present evidence against my answer to the parent's question?
There is no need to be scared of change, especially when Google inevitably replaces Android and ChromeOS with Fuchsia.
I'm not scared of "change", I'm scared of Google. I'm scared of losing power over my computers. Android is already a huge step back - only its Linux underpinnings allow it to act as a "real computer" in a pinch, and Google is rapidly encroaching on that too [0]. If Google is permitted to execute, unfettered, its vision for what end-user computing should look like, I do not expect it to be terribly empowering or respectful of my ideals, however technically impressive it may be.
Keep in mind that there are many teams at Google and they will need to be convinced. The Fuchsia may have broader ambitions but they seem to be starting with Google’s lesser-known OSes on embedded devices. This is in itself a win at a big company since there are fewer obscure OSes to maintain. That doesn’t necessarily mean that the Android or ChromeOS teams decided to commit just yet, as far as we know.
To be a good kernel. One imagines that, insomuch as a good kernel is part of an OS, that it could 'replace' Android and Chrome OS. But the days of platforms getting sunset are a day of the past, the Android runtime survives and will survive for years.
Maybe. I think they’ll continue rolling it out to simple devices for a few years while they do extensive testing and fix any issues that arise but I do believe their long term goal is for it to be a replacement for all their various Linux based OSs (since it is supposed to be able to run on simple devices all the way up to smart phones and PCs.
Treble already provides that stable ABI, since Android 8 Linux drivers are considered legacy on Android and all newer drivers must use Treble IPC for certification.
PS 4 uses an open source OS, how much have you seen of it?
Obviously this is a reason otherwise Fuchsia would have the same licence. However Google probably doesn't care that much about having an option to keep it closed source. Antagonism to the GPL comes mainly from Hardware vendors.
They cannot make anything of Fuchsia, which has been released under a permissive license retroactively open source. If Fuchsia replaces Linux on a large scale, they cannot take that away. What Google could do is having a closed source branch or release custom builds of Fuchsia, which have closed source in it. While there are certainly some szenarios which could be worrysome, it depends on the open source community on how much Google can get away with that. If only the open source parts are used, closed source version would just not have much meaning.
And of course, there would be some very legit reasons not to use GPL - like embedding ZFS into the kernel. Which would be no license issue for Fuchsia.
I'll be surprised if it's still around (as in not killed by Google) in a few years. Besides, I don't see it ever becoming popular enough to be an alternative to Linux outside of Google.
Why? It was just officially rolled out to its first devices less than two weeks ago. It isn’t like a google service that can fail/not get enough traction fuchsia already has a bunch of users (anybody with a google or android device) that it could be rolled out to. I imagine it will not replace android in the next five years but simpler devices it’s almost guaranteed to be the main OS shipped going forward.
