MGLRU Looks Like One Of The Best Linux Kernel Innovations Of The Year

[ATLief]

Otherwise what?

Otherwise they'll (eventually) switch Android and ChromeOS to Fuchsia, stop contributing to Linux, and potentially pull other users and contributors away from Linux.

[jorgepl]

With this out of the way, one of Linux' biggest & most obvious shortcoming will finally come to an end!

Thanks Google for actually caring about Linux as an interactive OS, unlike most other contributors which are mainly focusing on server (throughput) performance...

[Not you of course, Valve!]

Despite my dislike for Google, I agree. They do care about interactive interfaces at least, certainly so for ChromeOS and Android devices. Actually, if ChromeOS wasn't so badly and stupidly limited, it'd make the best operating system out there: open source, free, clean, quick, responsive, small footprint, focused on your content. Much less intrusive than macOS or WIndows.

But... for some stupid reason Google only wants you to run HTML there...

[jorgepl]

That’ll still take several years at least. Part of me thinks that announcement was just to keep Linus/x on their toes.

”Maybe you’ll accept our patches sooner rather than later. Otherwise…”

That won't take years simply because it won't happen. Everyone that has read the source code has said it is damn incomplete and pretty unimpressive despite initial hype. It solves nothing appart from the stable driver ABI (which Treble on the Android side solves, and which doesn't seem to be a problem from ChromeOS devices which have an excellent lifespan for a commercial consumer device). Everything that Google wants to do, they can do (and pretty well so) with Linux. They even are able to avoid licensing issues and close their devices more than Apple does (at least on computers).

Fuchsia has been WIP since at least 2016. So far it has only been released in 2 EOL devices without any (big or important) public acknowledgement from Google. Not even a sad blog post.

And Google keeps on working on mainlining the Android kernel. Am I to believe they want to replace Linux with Fuchsia? Well, it doesn't look like...

[jorgepl]

I know, since I was typing that very post from my (responsive & overall very smooth) cheap Chromebook...

PS:
I stopped worrying about Fuchsia ever since it became clear that Linux is actually capable of being turned into a hard real-time OS!
[Source: https://elisa.tech/]

Also, there is simply no way that Fuchsia with its microkernel design will be able to compete with Linux in the performance department, either.

Fuchsia feels more like vaporware the more time it passes. It serves no purpose and is not able to do anything better than Linux currently does UX wise. Security could theoretically be better, but security is already pretty good in Android and ChromeOS devices.

[jorgepl]

Otherwise they'll (eventually) switch Android and ChromeOS to Fuchsia, stop contributing to Linux, and potentially pull other users and contributors away from Linux.

The Linux Foundation must be terrified

[sinepgib]

Despite my dislike for Google, I agree. They do care about interactive interfaces at least, certainly so for ChromeOS and Android devices. Actually, if ChromeOS wasn't so badly and stupidly limited, it'd make the best operating system out there: open source, free, clean, quick, responsive, small footprint, focused on your content. Much less intrusive than macOS or WIndows.

But... for some stupid reason Google only wants you to run HTML there...

No stupid reason. Their business model is having you on the web to profile you. Native applications are less desirable for them.

[cynic]

Otherwise they'll (eventually) switch Android and ChromeOS to Fuchsia, stop contributing to Linux, and potentially pull other users and contributors away from Linux.

Good luck with that!

[pal666]

That’ll still take several years at least

that will take infinite amount of years. no single company can outcompete linux

[pal666]

Otherwise they'll (eventually) switch Android and ChromeOS to Fuchsia, stop contributing to Linux, and potentially pull other users and contributors away from Linux.

what resources? their linux contributions are tiny

[ATLief]

that will take infinite amount of years. no single company can outcompete linux

That's why I said "at least"; never happening is longer than several years