SoylentNews is people
SoylentNews
All going well, the stable release should appear sometime in August.
Introducing the release candidate, Torvalds said it was "right up there with v4.9, which has long been our biggest release by quite a bit in number of commits." That said, the 4.9 kernel was "artificially big" because of a couple of special factors, whereas 5.8 is a "more comprehensive release."
Torvalds said: "The development is really all over the place: there's tons of fairly fundamental core work and cleanups, but there is also lots of filesystem work and obviously all the usual driver updates too. Plus documentation and architecture work." He added: "We have modified about 20 per cent of all the files in the kernel source repository. That's really a fairly big percentage, and while some of it _is_ scripted, on the whole it's really just the same pattern: 5.8 has simply seen a lot of development."
While the code for the kernel is large, only a small part of it ends up in any individual system, since the kernel source contains code for every chip architecture and hardware it supports. In early 2018, maintainer Greg Kroah-Hartman said that "an average laptop uses around 2 million lines of kernel from 5,000 files to function properly." At the time, there were 25 million lines of code in the kernel, whereas now there are over 28 million.
See also: Linux 5.8 Kernel Features Include New Intel/AMD Capabilities, Security Improvements, Optimizations.
(Score: 0) by Anonymous Coward on Tuesday June 16 2020, @10:05PM (4 children)
It's my primary OS not out of preference but out of lack of serious choices. It's like the the US political systems but we get the illusion of three choices instead of two. Let's review the choices we have today:
1. Be spied on and forced to update and reboot "your" system based on another corporation's schedule.
2. Have a closed off ecosystem on both the desktop and the phone and a security theater of pretending to not cooperate with government agencies.
3. Use a sohmorically "designed" system that throws large amounts of code into kernel space (when even Microsoft has been moving away from this practice for over two decades)--a security, maintenance and stability nighmare; often breaks upon update; has weird instability problems on some configurations; and requires multiple proprietary blobs to make the system at all usable (how people have convinced themselves this is even allowed by the GPL is beyond me).
It's disappointing to hear people complain about systemd. If you aren't happy with that, boy are you going to love discovering how the internals of the kernel are cobbled together.
When are we going to get a real solution? Maybe we can get one written in a better language next time, too. People around here seem to hate Rust (for reasons that aren't actually technical, as far as I've seen), so maybe we can all agree on Ada or something. Almost anything would be better than C. It seems like no true Scotsman ever writes code with buffer overflows or any number memory errors, but it seems like we don't have a lot of true Scotsman programmers out there--they definitely aren't working on any of the three main OSes in use today. So maybe we should start thinking about something a little more practical if we care at all about security and technical soundness. But maybe I'm wrong. Maybe none of this stuff is important.
(Score: 1, Informative) by Anonymous Coward on Wednesday June 17 2020, @01:05PM (3 children)
Rust and Ada is not C replacements. C does what C does well. It's a damn stable and minimalist language that will let the developer control exactly what the computer is doing. It is quite possible and realistic to write safe C programs. There are plenty of linters for static checking of compliance to safety standards and compilers have instrumentation flags you can add, if you want runtime safety checks. My reluctance to use Rust (and for that matter C++) is that I can do those things with C in less complex (computational) ways and control how it is done. It's not like you can't do object oriented programming in C or comply to safety standards or have runtime crouches, if you so choose.
It's dissapointing to me that systemd proponents can't explain to me why I need it, yet call me radical when I say I don't want it. It's funny how your point 3. about the kernel is negative, then you equate it with systemd, yet systemd is good and the kernel is not. I think it's real good news that Devuan now has eudev and to Gnome users that elogind enables them to use another userland base than systemd.
I use ALSA, I use sysvinit, I use primarily C based programs. My system is fine and it's simple.
If your goal is safer programs there is no swoop-solution. Developers need to use tooling that ensures that programs have a high standard of security. A coding safety standard, a good linter, static analysis (ex. frama-c), good build settings, fuzzing.
I thought all this freedom nonsense was about choice.
(Score: 1, Funny) by Anonymous Coward on Wednesday June 17 2020, @03:47PM
About time for one of you "no true Scotsmen" types to come out! Just as expected.
(Score: 0) by Anonymous Coward on Wednesday June 17 2020, @06:40PM (1 child)
>It's funny how your point 3. about the kernel is negative, then you equate it with systemd, yet systemd is good and the kernel is not.
It is amazing to me that this is how you interpreted my post. I don't even know how to think about that.
(Score: 0) by Anonymous Coward on Wednesday June 17 2020, @07:28PM
I feel the same about your reply.