I don't want to "interject" on this, but I think when you start to use Android to bump the OS market share of Linux it starts to be pretty dishonest. Android is nothing like the desktop...
Indeed, thanks to Android, Linux is the most popular end-user operating system, nudging out Windows by 4 percent (39% vs. 35%).
I don't want to "interject" on this, but I think when you start to use Android to bump the OS market share of Linux it starts to be pretty dishonest. Android is nothing like the desktop Linux-based distros.
Agreed. A mobile OS is very different from a desktop OS. Windows learned this the hard way with Windows 8, trying to cram the two "experiences" together into a single clusterfuck that failed on...
Agreed. A mobile OS is very different from a desktop OS. Windows learned this the hard way with Windows 8, trying to cram the two "experiences" together into a single clusterfuck that failed on both platforms. A mobile OS is designed to be a stripped down, simplified interface with touch and swipe gestures and minimal screen real estate being central considerations in their design, whereas a desktop OS is designed with a mouse and keyboard as well as a much larger screen real estate in mind and therefore has a much more expansive feature set and layout. Take any Android or iOS device and compare them to any Linux distribution or macOS version and the experiences are wildly different. They simply cannot be equated and attempting to do so to pad numbers is disingenuous at best. It's even more absurd when you look at their source and see that desktop Linux distributions have a market share of a measly 0.85% as of August of this year.
It's fine if they point out that Linux-based systems take a market share of 39% because of Android. That makes sense since the article is about the new Linux kernel update. Even more so since the article points out that Android is allegedly going to be tracking the latest stable updates from the kernel rather than maintaining a fork. So I have no problem with them bringing up the Android market share and pointing out that it essentially makes Linux "the most popular end-user operating system". It's a factual, relevant statement to make.
But to basically compare Android market share to Windows market share is just comparing apples to oranges. Yes, they're both fruits, but we're not making apple pies out of oranges.
Nor is it like a supercomputer or a mainframe, but it's still Linux at heart. And, in the particular case of Android, it's--not Windows--is now the top end-user operating system.
Nor is it like a supercomputer or a mainframe, but it's still Linux at heart. And, in the particular case of Android, it's--not Windows--is now the top end-user operating system.
It's interesting to see all these projects warming up to Rust. The Linux kernel seems like a fine place for it, and I hope they find ways to interop without introducing too much complexity. In a...
It's interesting to see all these projects warming up to Rust. The Linux kernel seems like a fine place for it, and I hope they find ways to interop without introducing too much complexity.
In a more pessimistic view, I almost feel that Rust is too late. We already have these big established projects like the kernel and Chromium that want to integrate Rust, but there's so much existing code that it'd be impossible to rewrite. It has to be done piece-by-piece, module-by-module, when and if there's time.
From that perspective, it feels like we'll never reach "safety nirvana". A place where the code running our OS and primary software has a near-guarantee of being memory-safe.
But I do recognize that even small increments can be beneficial. If 50% of bugs occur at a security layer, then rewriting that layer would still offer substantial improvement. It doesn't have to be whole hog to be useful.
So as much as I'd love to see moonshot attempts at ground-up rewrites (Fushia might be a good candidate), I think this iterative approach is the best way forward. I hope these projects find success in integrating Rust.
I don't want to "interject" on this, but I think when you start to use Android to bump the OS market share of Linux it starts to be pretty dishonest. Android is nothing like the desktop Linux-based distros.
Agreed. A mobile OS is very different from a desktop OS. Windows learned this the hard way with Windows 8, trying to cram the two "experiences" together into a single clusterfuck that failed on both platforms. A mobile OS is designed to be a stripped down, simplified interface with touch and swipe gestures and minimal screen real estate being central considerations in their design, whereas a desktop OS is designed with a mouse and keyboard as well as a much larger screen real estate in mind and therefore has a much more expansive feature set and layout. Take any Android or iOS device and compare them to any Linux distribution or macOS version and the experiences are wildly different. They simply cannot be equated and attempting to do so to pad numbers is disingenuous at best. It's even more absurd when you look at their source and see that desktop Linux distributions have a market share of a measly 0.85% as of August of this year.
It's fine if they point out that Linux-based systems take a market share of 39% because of Android. That makes sense since the article is about the new Linux kernel update. Even more so since the article points out that Android is allegedly going to be tracking the latest stable updates from the kernel rather than maintaining a fork. So I have no problem with them bringing up the Android market share and pointing out that it essentially makes Linux "the most popular end-user operating system". It's a factual, relevant statement to make.
But to basically compare Android market share to Windows market share is just comparing apples to oranges. Yes, they're both fruits, but we're not making apple pies out of oranges.
Usually I'd agree, but in this case I think its fine because the article is about the Linux kernel in particular
Nor is it like a supercomputer or a mainframe, but it's still Linux at heart. And, in the particular case of Android, it's--not Windows--is now the top end-user operating system.
It's interesting to see all these projects warming up to Rust. The Linux kernel seems like a fine place for it, and I hope they find ways to interop without introducing too much complexity.
In a more pessimistic view, I almost feel that Rust is too late. We already have these big established projects like the kernel and Chromium that want to integrate Rust, but there's so much existing code that it'd be impossible to rewrite. It has to be done piece-by-piece, module-by-module, when and if there's time.
From that perspective, it feels like we'll never reach "safety nirvana". A place where the code running our OS and primary software has a near-guarantee of being memory-safe.
But I do recognize that even small increments can be beneficial. If 50% of bugs occur at a security layer, then rewriting that layer would still offer substantial improvement. It doesn't have to be whole hog to be useful.
So as much as I'd love to see moonshot attempts at ground-up rewrites (Fushia might be a good candidate), I think this iterative approach is the best way forward. I hope these projects find success in integrating Rust.
There's also Redox OS. But, yea, the piecemeal approach is probably the best (and only) option for integrating Rust over time into massive code-bases.
Don't forget about io_uring!
https://thenewstack.io/how-io_uring-and-ebpf-will-revolutionize-programming-in-linux/
https://dzone.com/articles/the-backend-revolution-or-why-io-uring-is-so-impor
I'm not sure how uring will work out in the long run, but I'm keeping a close eye on it.
There's been a lot of new good stuff in the Linux kernel lately, and there's more open-source goodness coming.