Linux Kernel Developers and Commits Dropped in 2019 (phoronix.com) 37
Phoronix reports that on New Year's Day, the Linux kernel's Git source tree showed 27,852,148 lines of code, divided among 66,492 files (including docs, Kconfig files, user-space utilities in-tree, etc).
Over its lifetime there's been 887,925 commits, and around 21,074 different authors: During 2019, the Linux kernel saw 74,754 commits, which is actually the lowest point since 2013. The 74k commits is compares to 80k commits seen in both 2017 and 2018, 77k commits in 2016, and 75k commits in both 2014 and 2015. Besides the commit count being lower, the author count for the year is also lower. 2019 saw around 4,189 different authors to the Linux kernel, which is lower than the 4,362 in 2018 and 4,402 in 2017.
While the commit count is lower for the year, on a line count it's about average with 3,386,347 lines of new code added and 1,696,620 lines removed...
Intel and Red Hat have remained the top companies contributing to the upstream Linux kernel.
Over its lifetime there's been 887,925 commits, and around 21,074 different authors: During 2019, the Linux kernel saw 74,754 commits, which is actually the lowest point since 2013. The 74k commits is compares to 80k commits seen in both 2017 and 2018, 77k commits in 2016, and 75k commits in both 2014 and 2015. Besides the commit count being lower, the author count for the year is also lower. 2019 saw around 4,189 different authors to the Linux kernel, which is lower than the 4,362 in 2018 and 4,402 in 2017.
While the commit count is lower for the year, on a line count it's about average with 3,386,347 lines of new code added and 1,696,620 lines removed...
Intel and Red Hat have remained the top companies contributing to the upstream Linux kernel.
Re: (Score:1)
Dear user,
Thank you for that insightful link! Once again, it tends to prove that Chris is a victim here. He keeps on reading those SEO, Youtube algorithm, basically get rich quick sites. He doesn't realize that he is the fish for them since they make money off him with their own schemes. Then, he wastes his time trying to implement what those sites suggest and he ends up disturbing people.
I mean, those crooks tell Chris that he has to build personal brands and he goes on the Internet and makes everything ab
Intel?? (Score:2)
Re:Intel?? (Score:4, Interesting)
Because the Linux money is in the server/workstation/cloud market, where sleep states and battery life means very little. According to StatCounter the Linux desktop web market share was 1.85% and on Steam it's 0.67%. There's of course Android too but it runs on ARM, so nobody really works to optimize Linux for x86 laptops except maybe a few from Google making Chromebooks.
Re:Intel?? (Score:4, Informative)
Good point! Also, Linux often use generic power management features defined by some kind of standard. Of course, Linux open source code has some specific tweaks to handle specific vendor hardware but in the end, it is lacking a lot of functionalities implemented by vendor binary blobs available under Windows and Apple. Some vendors provide Linux binary blobs for their hardware and I have used them quite a few times (your kernel becomes tainted). A very few provide their full up to date API in order to handle the hardware.
In the end, it is similar to installing and booting a new OS and using "generic vga" in order for the screen to work until you install more performant drivers. With Linux, sometimes your only choice is to use "generic whatever". Even when there are some optimizations available for Linux, they often trail behind.
Re:Intel?? (Score:5, Funny)
The few of us with linux laptops usually just choose the model with the Beluga whale battery pack sticking out the bottom, and then rarely use it off power.
Re: (Score:3)
Re: (Score:2)
Re: (Score:3)
Re: (Score:3)
I think that is outdated, at least for some brands. I have a Thinkpad Carbon X1 and a Dell Inspiron - both get very good battery lifetimes in Linux (8-12hr). A lot of this comes down to the BIOS and firmware.
Re: (Score:1, Interesting)
Re:Will be interesting to track through 2025 (Score:5, Interesting)
Too late, big developments in Russia, China and India. So probably yeah, you would see a whole lot less coming out of Russia, China and India as they will inevitably host their own forks, no longer trusting the USA for anything what so ever. So contributions from those sources will drop, then there is the EU, how far off is an EU hosted Linux because they also no longer trust the USA. USA hosted tech is likely to suffer in this environment, especially FOSS as the US kernel will mostly only end up with US contributions and will have to seek out the others.
There will be much forking and we can all guess exactly who will end up dominating FOSS in order to globally crush US tech firms, China (the Huawei ban was the single stupidest possible thing the US government could have done, ramifications across the whole planet, US tech can no longer trusted not one little bit, the cost to M$ huge).
Re: (Score:2)
Chrome, Firefox, Microsoft Office and Windows are shining examples of developers constantly fucking with things, and fucking them up in the process.
That seems to me all part of the plan.
Re: (Score:3)
That Linux is a mature OS is part of an explanation and evolution of hardware and software comes in waves. NUMA as well as other strategies were developed some generations ago.
It's possible that the new generation of AMD processors spawns new development caused by new instruction sets and also when you get new chipsets and new devices.
What could be done is more on the matter of security improvements and device drivers. There it's of course a balance between performance and security. In this case security is
Re: (Score:2, Insightful)
If working on a computer project for free not much fun?
It was fun in past years? What is the part that changed?
The OS? CPU? RAM price? GPU use? Network speed? The "cost" of a new computer/desktop/laptop/ISP account?
A CoC was enforced?
A once fun hobby becomes very political and the fun is gone....
Are the really smart people looking at new projects like more "work" under a new CoC.. ?
To code under some new CoC? What for? Fun? Hours of wo
Re: (Score:2, Informative)
Quit playing with your CoC.
Re: (Score:1, Troll)
The people who can code did not join a political group to code for.
They wanted to work on fun free code.
A CoC placed over a project is not fun and takes away from free limited time to code pre week...
Their projects have now become political.
The really smart people can just walk away... its their skills and their free time..they dont need the projects...the new politics, the long list of changes to another new Co
Re: Will be interesting to track through 2025 (Score:1, Informative)
Yup. A while ago a project I use was taken over by reactionary Progressives. First thing they did was force a CoC down the community's throat. No vote, no democracy, just corporate money power and coercion. After forcing their CoC, of course - of course - they began harassing contributors who didn't kowtow to the Official Narrative.
I stopped contributing. Many other people stopped contributing. That project is still alive, but it's not what it used to be.
Re: (Score:1)
See now I'm not even reading it as "C.O.C." I'm simply reading it as "cock".
Locked-down desktop-replacement devices (Score:2)
What is the part that changed?
The OS? CPU?
That might be part of it: a shift to wider uses of devices that run a graphical Linux OS and attempt to substitute for previous uses of a graphical Linux OS but can't easily be used to develop apps for a graphical Linux OS and certainly not the kernel itself. This blocks people who rely on an Android device or a Chromebook from getting into coding without purchasing another device.
Because, like quarterly profits... (Score:5, Insightful)
A CoC will (Score:1, Troll)
Return to code productivity.
Work on a CoC does not add the computer code needed for the productive parts of a OS.
Obviously... (Score:2, Funny)
More stuff is going into systemd so there's less need for code in the kernel. /s
Yeah, infinite growth in a pathologic thing. (Score:3)
Our capitalist views may be skewed, but in nature, there are only two types of processes: A) Exponentially growing ones, and B) *surviving ones*.
Because anyone with a leftover bit of common sense, quickly realizes that in a world with finite resources, the former means resources will run out, or waste will drown it, and the process will die. ... Great job! A+! --.--
Unless some "friction" counter-force starts to emerge, to balance it out, and make it *stable*.
Except, of course, when some moron (*looks at TFA*) comes along, and yammers "Oh noes! Teh stagnation! It is death! Do away with teh friction!" until the balance derails and we're right on track towards death again.
Low hanging fruit (Score:2)
A long time ago I encountered an irritating issue in the Linux soft realtime scheduling. Some hours of bug hunting later, I submitted a 1 line patch that fixed the issue.
Some years later, an issue in a netcard driver caused me some grief. I emailed the maintainer for that driver and he fixed the issue.
These two small contributions were both fairly easy - in each case the code involved was small enough to understand with only a little time spent studying it, and the problem (and its solution) wasn't that obs
Comment removed (Score:4, Informative)