There are many interesting things to discuss about Fuchsia. In this article, you will get a taste of how Fuchsia OS works through a deep dive into some of its core features. We’ll also run the Fuchsia emulator on our systems and try running some example components on it.
A great resource if you’re considering getting started with Fuchsia.
This article has already been covered by the video posted here on OSAlert: https://www.osnews.com/story/133763/two-googlers-offer-a-tour-and-coding-demo-of-fuchsia-os/
Does anyone know if “Geist” still works on Fushia at google? (The originiator of NewOS, that became the Haiku kernel base)
His twitter says so https://twitter.com/tkgeisel
Interesting to know that there’s lineage between Haiku and Fuchsia, even if it’s more by “marriage” (developer) than by “blood” (code)
Perhaps if Fuchsia is a evolution of what became Haiku/NewOS they could rebase Haiku to use it. That would help with the hardware support.
I think that ship has sailed long ago. I imaging the differences in the codebase are so phenomenally different now, that any attempt at moving the kernel of either would be wholly counterproductive. Also, i really doubt that Fuchsia has that much better hardware support than Haiku does, let alone the fact that Fuchsia is designed to target completely different hardware to Haiku.
There may be milage in designing a driver compatibility layer for Haiku, based on the Fuchsia driver API/ABI, much like Haiku has at the moment for FreeBSD network drivers. But then again, i expect most of those drivers will be proprietary, and targeted more towards smartphone/tablet hardware than generic PC hardware.
I expect Fuchsia to replace Chrome OS so that means graphics drivers which is the most difficult part.
The123king,
I suspect that Fuchsia will have a much easier time getting drivers simply because it’s backed by Google and Google’s market share assuming Google positions Fuchsia to replace Android, However I agree that Haiku won’t be driver compatible with Fuchsia unless Haiku’s devs decide to pivot towards Fuchsia’s driver model. Also like you said, they don’t currently target the same hardware. but maybe it could be valuable for an ARM port.
jgfenix,
Regardless of if it’s replacing Android or ChromeOS, it will have drivers, the question is whether those drivers will be open source. When Fuchsia goes into production, it’s very likely that the drivers will be written by manufacturers instead of Google itself, and just like with Android they will refuse to provide the source code.
Incidentally, this has lead to the notorious stalemate in the industry between FOSS developers (ie linux) and manufacturers, resulting in android and other devices being permanently locked to manufacturer kernels for it’s entire existence. I think this stalemate is a strong motivating factor for google to build it’s own kernel replacing linux because it can finally have a stable ABI…
https://fuchsia.dev/fuchsia-src/concepts/drivers/fdf
This should allow both google and owners to finally be able to update & replace the OS independently from manufacturer builds, which has been a serious con on the android + linux end of things. Manufacturers have been aweful about releasing new kernels and it’s been a huge obstacle for those who want to run their own linux builds on ARM. Phones are readily available, ready to be saved from the trash even, and could make awesome DIY platforms except it’s all for naught when our own updated builds don’t have the necessary drivers. Fuchsia may finally offer a glimmer of hope because the manufacture drivers will work even with our own DIY kernel builds!!!
This is a huge game changer for Fuchsia devices assuming artificial restrictions aren’t going to be built in. Whether google and/or manufacturers end up locking down Fuchsia devices is an area of uncertainty. And while I can see the benefits Fuchsia could bring already, I am worried that google might be temped to implement restrictions that take control away from owners, which could be devastating for independent development..
I’m really excited about Fuchsia! I’m hoping it’ll have nicer performance on the same hardware compared to an Android installation.
But I’m most concerned / cynical about this:
“All software that runs on Fuchsia receives the least privilege it needs to perform its job, and gains access only to information it needs to know.”
If I understand the current state of Android correctly, ALL apps are given FULL access to the internet any time they want, and unless you do something special (root, custom ROM, firewall…) there’s nothing you can do about it. This is the #1 reason I root my devices. My keyboard app, photo gallery, video player, alarm clock, and so on do NOT need access to the internet.
drcouzelis,
I doubt it will have any noticeable performance differences, but will be interesting to see. I think Fuchsia will bring a much needed alternative to Linux’s monolithic design and hopefully we will no longer be stuck running only the manufacture’s kernels like we’ve been with linux. It’s been a gripe of mine for decades.
I haven’t looked too closely at Fuchsia yet, but I just hope google doesn’t pull another half closed/half open model like they’ve done with android. The closed bits of android have been detrimental to alternatives like lineageOS.
Also, I hope the owner is given explicit control over security policies. I will be so disappointed if google gives themselves privileged access on Fuchsia devices while keeping owners in a restricted walled garden. This would be detrimental to owner rights.
I hope it’s easy to sandbox applications, this is something I wish linux did better. Linux’s ability to isolate processes has gotten more capable with namespaces but something like this works best when it’s designed with sandboxing from the start. In unix it’s been hard to get away from the original UID/mode based permission model since so many unix tools just make assumptions about them.