Slashdot Mirror
Linus Renames 2.6.40 Kernel To Linux 3.0, Announces Release Candidate
An anonymous reader writes "Linus just released the first -rc of the next kernel series, but rather than continuing development as the Linux 2.6.40 kernel, he has renamed it to be the Linux 3.0 kernel." And he's tacked on a second dot and another zero (3.0.0), at least for now, because many scripts expect and rely on a three-part kernel version.
There's never been a large enough jump in features to justify a major release increment, yet 2.6.40 is more distinct from 2.6.0 than 2.6.0 was from 2.0.0
Clearly 3.0.0 is 0.4.60 more advanced than 2.6.40.
signature is pants
I like his 3.0 commit message
"Version numbers? We can increment them!"
Thankfully, Linus hasn't rewritten the kernel in VB.
Also this version has codename "Sneaky Weasel"
--- a/Makefile
+++ b/Makefile
@@ -1,8 +1,8 @@
-VERSION = 2
-PATCHLEVEL = 6
-SUBLEVEL = 39
-EXTRAVERSION =
-NAME = Flesh-Eating Bats with Fangs
+VERSION = 3
+PATCHLEVEL = 0
+SUBLEVEL = 0
+EXTRAVERSION = -rc1
+NAME = Sneaky Weasel
I'd tell a UDP joke, but you may not get it. I'd tell a TCP joke, but I'd have to keep repeating it until you got it.
Maybe this year is the year of the Linux desktop too?
Oh shit! Are you telling me there is a regression in how LInux does arithmetic?
LFS user here. Will 2.6.39 get the LTS treatment just like 2.6.35 down to 2.6.32? Would be nice to have a stable target for years to come. I have a box that's still using 2.6.16 (formerly LTS) and another that's 2.4.37. Moving up from "minor" releases, e.g. from 2.6.35 to 2.6.36 haven't really been as minor as they used to be. They tend to be somewhat nerve-wracking experiences. Personally sticking to 2.6.35 as long I can.
Clearly 3.0.0 is 0.4.60 more advanced than 2.6.40.
Hmm, are you forgetting to carry the overflow from minor digit to major digit? In this case, 3.0.0 would be 0.3.60 more advanced than 2.6.40, naturally.
Those who can make you believe absurdities can make you commit atrocities. - Voltaire
There's one more thing worse: the people who complain about how annoying the people who complain about the meaningless bump in version numbers are. Boy those guys are real jerks!
What a fool believes, he sees, no wise man has the power to reason away.
He wrote on the lkml: "the real reason is just that I can no longer comfortably count as high as 40."
No, 2.6.40 + 0.3.60 = 2.9.100
I think he meant 1.-6.-40 more advanced.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
That's a bold statement.
Usually a change in major version number means major incompabilities, that's what the "major" means. I.e. Python 3 is incompatible with Python 2. GTK 3 is incompatible with GTK 2. etc. Doing so when it is in fact not incompatible upsets people's expectations but is a pleasant surprise when they find out they don't actually have to adapt everything to it.
This means: 3.0.x - expect disruptive changes?
Questions raise, answers kill. Raise questions to stay alive.
Gnome 3.0 deserved the update of the major version number because the libraries it depends on have been extensively revised (GTK+ went to version 3.0, for example).
But does anyone else get chills when thinking about the 3.1.1 version somewhere down the road?
You know... as in for workgroups?
Mit der Dummheit kämpfen Götter selbst vergebens
I hear ya bud, but do you know what really grinds my gears?!
On second thought - lets just drop it... it's jerks from here to infinity.
My understanding is that the jump to 3.0 is simply that they no longer want to have the second digit even means stable and odd means unstable versioning any more. So rather than going to 2.7.0 and having everyone assume it's unstable or skipping 2.7.0 and going straight to 2.8.0 just to maintain an old and unused version system, they have went with 3.0.
No, Linus mentioned that it is an evolutionary (not revolutionary) release, and that the APIs and ABIs wouldn't change.
But I guess the marketing mentality somehow, somewhere, has taken over.
Hardly. It was already broken, the "2.6" part of the number was completely irrelevant, and whereas it might not bother you, if you're talking about version numbers all day every day, having superfluous data in there will get annoying. So yeah, the "upgrade" is misleading but from now on the version bumps more accurately reflect the scale of change in the kernel.
Anyway, who markets the kernel? Distros are marketed, nobody cares about the kernel who doesn't already know what's going on.
This is far more a case of developers wanting a version number system that makes sense to the current kernel development model than anything else.
This is a complete outrage. Not only will it require extensive re-testing but distros will need to change as well.
I believe it's time for us to fork the GNU/Linux kernel to a more appropriate versioning scheme, while removing all non-libre blobs at the same time. Only then can we depose this dictator Torvalds and his pro-capitalist kernel.
Everyone knows minors don't overflow into majors. (And if you thought Naturally was the first baseman, then you don't know Who.)
Now the 3.0 Linux branch is just plain about shiny numbering.
Yup, and is all the better for it. What you don't mention in your list is the fact that the development model changed in 2.6, from a break-> stabilise->break-> stabilise model to one of continuous stable development. The version number system stayed the same, which suggests the same development process of stabilisation with no new features, so this is a newer system that fixes that.
He is Finnish, but his first language is Swedish.
(PS. This is why you should not choose the language for your web site based on a geographical lookup of the IP address.)
"Can someone please explain what is the difference between 3.0 and 2.6.40 ?"
Linus: The numbers all go to 3. Look, right across the git repository, 3, 3, 3 and...
Interviewer: Oh, I see. And most kernels go up to 2.6?
Linus: Exactly.
Interviewer: Does that mean it's better? Is it any better?
Linus: Well, it's one better, isn't it? It's not 2. You see, most blokes, you know, will be running 2.x. You're on 2.6 here, all the way up, all the way up, all the way up, you're on 2.6 on your computer. Where can you go from there? Where?
Interviewer: I don't know.
Linus: Nowhere. Exactly. What we do is, if we need that extra push over the cliff, you know what we do?
Interviewer: Compile it up to 3.
Linus: 3. Exactly. One better.
Interviewer: Why don't you just make 2.6 better and make 2.6 be the top number and make that a little better?
Linus: [pause] These go to 3.
If linux kernels had microsoft marketing setting the names, we wouldn't have decimal points etc.
It would be "Linux NT", "Linux 95", "Linux Server 2003", "Linux XP", "Linux Vista", "Linux 7".
Just think how much more marketable Linux could be and how much more the suits would want to buy it.
Well, under the current development model "2.6" is essentially static, It's like OS X always remaining OS X rather than move to OS XI, OS XII, OS XIII etc. as there's absolutely no work on a "2.7" branch and probably never will be.
The 2.6.x changes are far bigger than a 0.0.1 change should be, I mean it's the main development release. Making them 0.1 changes is more than reasonable. The stabilization team will get to move up from 4th to 3rd digit so 3.0.3 rather than 2.6.40.3. Simpler, shorter all around.
I thought the Linux community wasn't shy of just minor, incremental updates. If it ain't broke don't fix it, don't rock the boat etc. But I guess the marketing mentality somehow, somewhere, has taken over. /looks at Gnome 3.0
Also are you arguing that Gnome 3 isn't a radical enough departure breaking enough eggs to warrant it's version number? Sounds to me like most people complain it's too different from Gnome 2.x. In this case, you seem to argue Linux 3.0 will be too similar to Linux 2.x. Is there a way to win here?
Live today, because you never know what tomorrow brings
IMHO it should have been done back with 2.6.19 or no later than 2.6.25. Better late than never though.
Oh really...
What about removal of the big kernel lock?
What about plug-in resource schedulers?
What about fast ip locking?
What about kernel video mode switching?
What about systemtap?
What about cgroups?
And much more...
When taken in combination, the growth of the Linux kernel since 2003 definitely warrants a major jump.
The issue is whether it should be 2.8 or 3.0. I would side with 3.0.
Because Linux is now ready for serious MP, both on a local and a cluster level. And these features are not "backwards portable".
Just another "Cubible(sic) Joe" 2 17 3061
"I decided to just bite the bullet, and call the next version 3.0. It will get released close enough to the 20-year mark, which is excuse enough for me, although honestly, the real reason is just that I can no longe rcomfortably count as high as 40," said Linus.
"If fifty million people say a foolish thing, it's still a foolish thing."
Well, since there's no predefined definition of version numbers, you have considerable freedom in defining them. For example:
Version numbers a.b.c and d.e.f are added by forming the ordinals (w^2 a + w b + c) and (w^2 d + w e + f) where w is the smallest infinite ordinal, then adding those (note that ordinal addition is not commutative!), giving an expression of the form (w^2 f + w g + h) which then is concerted back to the version number f.g.h in the obvious way
For example, we would get 2.6.40 + 0.0.1 = 2.6.41, 2.6.40 + 0.1.0 = 2.7.0 and 2.6.40 + 1.0.0 = 3.0.0 -- which actually makes some sense. Also note that version comparison reduced to ordinal comparison would give the correct ordering.
Note that the above definition assumes that all version number components are nonnegative (there is no such thing as a negative ordinal). However, the major version number could be allowed to be an arbitrary ordinal number :-)
The Tao of math: The numbers you can count are not the real numbers.
Actually, he was calculating it in his head because OpenOffice Calc was taking too long to load.
That's because you didn't get the reference you insensitive clod!
When Linus states that the ABI won't change, he means the user-space ABI (e.g. that (g)libc uses).
Non-Linux Penguins ?
I could never understand what is with all these digits in version numbers. If it was up to me the kernel would be in version 8.x or 9.x already.
What's with open source and all these version numbers starting with 0.x?. Why are they so afraid of just a freaking number? I've been using mythtv for about 10 years and they just released version 0.24.1 *facepalm*
Linus just realized that version numbers are about marketing more than anything else. Microsoft has been doing this for decades. I should buy me some redhat stock.
HTML is obsolete. It's time for a new, simpler and richer markup language.
Linux 3.1 was a certain ring to it.
echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
I prefer to read in well-written English, rather than see them bastardize my mother tongue.
In soviet russia the government regulates the companies.
No, it's called "I'm an idiot", produced, written and directed by siride.
Right, the ever-feared "divide by 2" error...
First project (at least that I was using) I remember pulling it was Slackware jumping to 7. It seemed kind of odd to me at the time, though even moreso now.