OMG! Ubuntu! reports
The arrival of the Linux Kernel 4.12 at the weekend brought a boat load of big changes (including two I/O schedulers) but do you know how big it is?
Well, it's easy to see in this chart shared by kernel maintainer Greg Kroah-Hartman which details exactly how big the release is. graphic
"Linux 4.12 is big, really big, like bigger than you thought big", [Greg] says in an update on his Google+ profile.
It took 63 days to create Linux 4.12, during which a total of 14,570 commits were made across 59,806 files.
With 24,170,860 [...] lines of code in the Linux kernel 4.12, that works out at a boggling 795.58 lines of code added per hour.
Linus Torvalds commented on the size of the latest stable release in his mailing list post to announce the release, saying:
"As mentioned over the various rc announcements, 4.12 is one of the bigger releases historically, and I think only 4.9 ends up having had more commits [...] 4.12 is just plain big."
(Score: 0) by Anonymous Coward on Saturday July 08 2017, @02:09AM (8 children)
Big != Good
(Score: 0) by Anonymous Coward on Saturday July 08 2017, @02:11AM (4 children)
Big != Bad
(Score: 3, Insightful) by kaszz on Saturday July 08 2017, @02:17AM (3 children)
Seems you believe in luck ;-)
Big means big chance of more bugs than for less code.
(Score: 4, Insightful) by frojack on Saturday July 08 2017, @04:41AM (2 children)
Exactly. Where are all these sets of eyes that are supposed to be Reading all this code?
If each of us able to to read and comprehend enen the simplest module were to adopt one single program, we still couldn't even gloss through the kernel and drivers, and almost certainly couldn't find any significant errors.
Lots of code is lots of problems.
No, you are mistaken. I've always had this sig.
(Score: 2) by kaszz on Saturday July 08 2017, @11:50AM
In other words, simple and small code is good code (usually).
(Score: 2) by http on Saturday July 08 2017, @06:16PM
That's a wild misinterpretation of the Linus's Law (coined by ESR), which is "Given enough eyeballs, all bugs are shallow."
Nowhere does it say everyone has to read the code, or even that they should, or even that it's a good idea. Or even most coders on a project. The point Raymond was underlining is that with open code, it's easy to collaborate and find the origin of a particular bug. This task is at right angles to both auditing commits and fixing bugs.
I browse at -1 when I have mod points. It's unsettling.
(Score: 0) by Anonymous Coward on Saturday July 08 2017, @02:13AM (1 child)
You just keep tellin' 'em that.
(Score: 1, Troll) by Ethanol-fueled on Saturday July 08 2017, @06:55AM
Fucking Niggers.
(Score: 1, Touché) by Anonymous Coward on Saturday July 08 2017, @02:13AM
You realize you don't need to select "yes" for every option and every driver in the config?
(Score: 1, Offtopic) by realDonaldTrump on Saturday July 08 2017, @02:31AM
The spreadsheet says there were 1,202,920 lines added, not 24,170,860 or 24,170,860 million. It works out with 1,202,920. It doesn't work out with those other numbers. Too big! 24,170,860 is the total, all the lines ever added. My budget for the #USA was off by $2 trillion. Mnuchin fucked it up, big league. So I'm very careful about numbers now. 🇺🇸
(Score: 2) by black6host on Saturday July 08 2017, @03:59AM (3 children)
If it's that big (of a change) then I'd most likely not run it for a while. Lots of changes to code mean lots of new possibilities... Good or bad. I've written enough code to know.
Note: I'm not currently running Linux, but I have, way back when. And I'm not arguing that any other OS is better. Just that big may or may not be better.
(Score: 2) by RS3 on Saturday July 08 2017, @04:21AM (2 children)
I've had 4.12 running for 4.4 days on CentOS 6.9 on one of the webservers I admin. Absolutely no problems, seems fast, RAM well managed. It's a "stock" kernel-ml-4.12.0-1.el6.elrepo.i686
(Score: 2) by black6host on Saturday July 08 2017, @05:42AM
Well, I must applaud your uptime!!!!!!! J/K hope it continues working out well for you. I guess I just grew up to be a cynical bastard, regardless of the venue :)
(Score: 0) by Anonymous Coward on Saturday July 08 2017, @10:13AM
And you run spanking brand new kernels in production webservers? Prfff.
(Score: 1) by Acabatag on Saturday July 08 2017, @04:09AM (2 children)
The kernel will probably take longer than 63 days to build, on my 386sx. It has all 16 megs the motherboard can hold.
I once tried to build FVWM2 on NetBSD on my Mac SE/30. (yikes)
(Score: 2) by Soylentbob on Saturday July 08 2017, @08:12AM
Maybe there is an Android-app [virtualbox.org] to cross-compile it on your Smartphone; that should be way more efficient.
(Score: 1) by toddestan on Saturday July 08 2017, @10:56PM
Just so you know, 386 support was dropped a few years ago. You might want to consider upgrading to that 486.
(Score: 0) by Anonymous Coward on Saturday July 08 2017, @04:28AM
Full of sound and fury, signifying exactly what?
(Score: 5, Informative) by KiloByte on Saturday July 08 2017, @05:21AM (1 child)
If anyone bothered to read Linus' post, you'd know there wasn't anything special: the commit count is above average but nowhere close to 4.9, the commits/day ratio is wrong (you divide by the length of the previous cycle as the merge window is at the beginning), and the massive line count comes from headers of a single bloated driver. All that idiotic "Wow"iness isn't a fault of Greg either.
That whoever that "OMG Ubuntu" guy is happens to be illiterate is not surprising: Sturgeon's law is not a new thing and millenials really push it to its limit. But, aren't we supposed to have an editor weed out crap and at least take a glance at the article that's being posted?!?
I can understand posters not having read TFA -- it's Slashdot/Soylent's tradition, but if even the alleged editor shirks this duty...
Ceterum censeo systemd esse delendam.
(Score: 0) by Anonymous Coward on Saturday July 08 2017, @07:30PM
I didn't want to lay the blame on millenials, but the writing style reminded me of typical ad infested short lived online e-zines of yesterday's early internet. filled with gee whiz claims and name dropping people only geeks would have a clue about, citing facts and figures that mean nothing to anyone that is hard core.
meaning, even if you couldnt tell from the name, it's an enthusiastic enthusiast who could even just be some sort of paid marketing shill, and not a real hard core/smart person that knows what he or she is doing.
no one claims the stuff that person claimed, if trying to impress other techs. instead, it reads like a canned template, like cosmo's steps to make someone go wild, repeated every few years with a few new cultural references to keep it relevant to new readers since all the old ones have left.
(granted, i would like to have someone get me to go wild... but ubuntu is not the way to do it, nor with the facts sited... geek I am, dweeb I am not)