Linux

Linux 3.0: More important than you think

Linux turns 3.0 and does so with little to no fanfare. Read why Jack Wallen believes this might be one of the more significant (and fanfare-worthy) releases of the Linux kernel to date.

Ah vacation. It was a week of blissful lounging around a breezy beach side and playing in a water-filled world where I was no longer at the top of the food chain. There were no computers, no talk of networking this, security that, or anything in between. But then the hard reality of the world wormed its way back into my mind and I now find myself trying hard to get back into some sort of groove...an open source kind of groove (of course).

And although it's officially next month (the month of my forty-fourth birthday, thank you very much) Linux is about to turn 3.0. And although Linus Torvalds himself has said this is not a big deal, it is. Why? Because of the very fact it is not a big deal.

Say what?

Let me explain. The Linux kernel is about to jump from 2.6 (where it has been for a number of years) to 3.0 and do so without much fanfare at all. I remember the day when Linux jumped from 1 to 2 and one would have thought every Linux fan on the planet had just won the lottery. Yes, the change from 1 to 2 was quite a bit more significant...but any time there is a major change in release numbering, it is generally done so because of a milestone. And from my perspective, going from 2.6 to 3 in such a way is a huge milestone.

Gone are the days when users prayed to whatever God or Gods they depended upon that specific hardware support would be included in the next release. Gone are the days when a major kernel release heralded a more stable, efficient, cleaner experience. Now...those wishes are assumed. Linux (and the Linux community) has finally reached a point where things can be (and I say this daringly, I know) taken for granted. And 3.0 proves that. How?

  • We assume there are going to be new features.
  • We assume things are going to work out of the starting gate.
  • We assume there will be performance improvements.

These assumptions occur whether they are true or not -- even if it has been made clear there are no deal making/deal breaking changes in the kernel. After all, look at the major feature list in the 3.0 kernel:

  • Btrfs data scrubbing and automatic defragmentation
  • XEN Dom0 support.
  • Unprivileged ICMP_ECHO.
  • Wake on WLAN.
  • Berkeley Packet Filter JIT filtering.
  • A memcached-like system for the page cache.
  • A sendmmsg() syscall that batches sendmsg() calls.
  • The setns() a syscall that allows better handling of light virtualization systems such as containers.
  • New hardware support such as Microsoft Kinect and AMD Llano Fusion APUs.

So changes like the above are now considered "minor" and "insignificant". To me, the very idea that those changes have come to be considered minor is significant.

That's right -- from my perspective, Linux turning 3.0 is much more telling than most are giving it credit for. Although the 3.0 moniker, to many, is just a numbering scheme change with some minor updates, 3.0 is a major leap forward for a twenty-year old operating system. Linux kernel 3.0 tells me that Linux is truly grown up and that the old argument for or against Linux being "ready for prime time" has finally been tossed out with the last shakes of beach sand from the shoes.

To Linus Torvalds I want to say "congratulations" on your pet project coming of serious age. To the developers of the Linux kernel I want to say "congratulations" on your hard work. And to the Linux operating system, I want to say "Happy Birthday, you're all grown up!"

About

Jack Wallen is an award-winning writer for TechRepublic and Linux.com. He’s an avid promoter of open source and the voice of The Android Expert. For more news about Jack Wallen, visit his website getjackd.net.

27 comments
ivank2139
ivank2139

I think a large part of what is so much better nowadays about Linux is the result of a disciplined and well managed development process. Lots of regression tests and a good patch management process have improved the stability of the code tremendously.

Zenith545
Zenith545

Doesn't Microsoft do a major version renumbering with every Windows??? Now they ran out of moniker ideas and we will just go from Windows 7 to Windows 8, which in fact will be Windows 7.2 or 7.3. LOL

Jerriais
Jerriais

So will this do anything to explain how to restore sound support in Debian distro that has been missing, presumed cancelled somewhere, since they moved it to 'non-free' optional?

aroc
aroc

I am getting to dread new releases as more and more device support is dropped (or just plain messed up as with recent xorg "updates") for the older hardware I have no reason to get rid of from a functional standpoint - still works under WinXP. I seem to see Linux fanboys (vs "serious supporters") throw out there every so often how wonderful Linux is about keeping old hardware from being scrapped - BS! Oh, sure I can keep the older versions, but then I get behind on the security updates.

ShawnS427
ShawnS427

Unlike for-profit OS developers, when we all work on our branches, we change the version number for revision control, not market appeal. I veiw the development of linux as a evergrowing mountain, as it ages and developes we say "here is good place to stop" , mark down where we are, and give it out. But since you climb a mountain one step at a time 3000 feet is only one step more than 2999ft , which are arbitrary marking points and not naturally signifigant. The fact that we can assume stability, support, and funcion has been a building proccess as well. It could have been 2.7 , but the 3.0 choice was made to generate articals like this, and pump up the public image, and more free advertising :)

bistra
bistra

I really don't care what it includes...all I know is that the Linux community provide me with a genuine option to avoid using the products of the despicable Microsoft. Whether it has every function, indeed the functions described are more or less unintelligible to me, I really don't care. It does the job and does it well. A heartfelt thankyou Linux community.

PVBenn
PVBenn

I learned a long time ago never trust anything where the version starts with 1dot or ends in dot0. I'll wait for 3.0.1.

realvarezm
realvarezm

Great something for....nothing is free! This is going to be a major turn to the linux comunity and for IT world too. This is great!

stephanie.white
stephanie.white

My clients have been running key apps on Linux for years. In 2004 one of my clients opted for linux over AIX. It has been considered "industrial strength" for at least that long by many others as well. I think it was significant when Big Blue went linux. Even IBM's BCU for linux (aka appliance calculator) goes back to 2006.

mrdt
mrdt

Since I started using Linux, there have been a multitude of changes to the kernel, and only small insignificant version number changes. It probably could have, or should have reached 3 some time ago. But I guess that perhaps Linus was being conservative with his version numbering.

opensourcepcs
opensourcepcs

The Linux kernel has been incidentally improved for years. It really is high time that we rolled the odometer on the Linux Kernel version numbers. While version 3 is not so much a change in functionality, it is an acknowledgment of the change in perception from the work-in-progress OS to the undisputed master of the super-computer world.

Adam S
Adam S

Linus has it right: Arguing over numbering is silly and time is better spent discussing other things. We use a lot of software that stays in beta or sub-1 versions for years. The only reason we assume that a number change means improvement is because... it is supposed to mean something has changed for the better. Would we expect it to get worse? This doesn't just apply to the Linux kernel. It applies to all software. Why did they take the kernel to 3? Considering how few changes were made, why not 2.7 or whatever? (To be honest, I don't even know what version I have installed right now. That's how little I care.) They could have taken it to 5. Oh, but that's twice the improvement, right? Why think that? If the boss man wants to rewrite the numbering scheme and relate it to release date, he could do that and stop all the hand wringing by developers. The point is this: It is just a number. There's nothing significant in it and, thus, no hoopla.

rfolden
rfolden

Again, the whole discussion re: "Linux" is a non-starter.

john.mountcastle
john.mountcastle

If I remember correctly, my first running kernel was 0.11. I've used and studied Linux since then and I'll be eager to install 3.0. Just for the pure beauty of it.

seanferd
seanferd

Let's see if it stays as intelligent, interesting, and amusing for the next day or so. Best wishes.

Spitfire_Sysop
Spitfire_Sysop

"The whole renumbering was discussed at last years Kernel Summit, and there was a plan to take it up this year too," Torvalds added. "But let's face it - what's the point of being in charge if you can't pick the bike shed color without holding a referendum on it? So I'm just going all alpha-male, and just renumbering it. You'll like it."

Alpha_Dog
Alpha_Dog

Linux 3.0 refers to the kernel, not the OS. It doesn't break any significant conventions, so from a user's perspective it just works as they have come to expect. From the coder's perspective, things are cleaner, and there's a bit of new to learn, but it's not a complete revolution. not a big deal here either. The 3.0 moniker is a mark of the code maturity and revision, nothing more.

sysop-dr
sysop-dr

I am having no problem with sound in Debian. In fact it allows me to use 2 sound cards, one for applications and gaming, the other for music, in surround sound with a subwoofer and I have actual physical channel db gauges. If you are having issues with sound in Debian try using wrappers for your applications such as padsp or alsa-oss. Debian works great with sound, but like other things with opensource software you have to invest some time to get it going sometimes. Good luck.

james.vandamme
james.vandamme

...which is 4.04 better than Windows 7. If Mozilla had numbered Linux kernels they'd be up to 17 by now

bornbyforce
bornbyforce

I totally agree. It is all about giving options. Unless you are one who desperately needs one of the newly added ones, you just use it and trust them. Linux is full of problems but it is good enough to be my operating system of choice. And choice is what I care about most.

shryko
shryko

There is the psychological impact, and that is why, in my mind, 3.0 is a lot better than 2.7... Let people feel like it is so much better, and for marketing, 3.0 is much easier to sell to people who have had bad experiences in the past. You can say the new version is fixed (from the much older versions, it really is), and then it's not as much of a fight to get them even to try it. This really only affects marketing and the devs, not the current users. Easier to market and sell, and new under-the-hood stuff for the devs... but the users will not really see any difference. Much nicer engine under the hood is easier to sell the car on, and it's nicer for the mechanics to work with... but the user still sticks in the key, turns it on, and drives the same way they did before.

AnsuGisalas
AnsuGisalas

It's like he's making fun of the very concept of waffling over unimportant issues forever...

Justin James
Justin James

Looking at Jack's improvement list, they don't exactly scream "major version" to me. Wake on WLAN? Yeah, that's worth revving the big number over... I think maybe Jack should have read the full quote from Torvalds. They revved the major because they felt it was time to rev the major, that's IT. If it signals anything else to anyone else, that's their own fault for not doing any kind of research. No one who would go within 10 feet of a Linux box cares if the kernel is at 2.X or 3.X, because the distro version is so much more important anyways. J.Ja

shryko
shryko

and sarcasm can't be read online :P at least he's got humour about the BS of bureaucracy. :P

pgit
pgit

I voted "no." I was rather surprised to see this, I remember Linus saying he'd never do such a thing just for the sake of a bigger number, but here we are. In fact I remember he said there may never be a 3.x kernel, as there wouldn't be significant enough change to warrant it. I just hope this doesn't break anything. I can see where a lot of devs are going to have a lot of make-work on their hands. That'd be great if they got paid for the work. Somehow I doubt too many people are going to be particularly excited about this jump. That's not to belittle the changes coming along, but none of them are serious game changers so far as my ops go.

Justin James
Justin James

Sure, those are both important changes... but really, they aren't exactly "major" changes for the Linux kernel. The Linux kernel has had plenty of similar changes over the course of the 2.X branch, and no one was begging for a major rev. Ditto for the BSD Mach kernel, the Windows kernel, etc. In the world of kernels, truly "major" changes are actually pretty rare. I would argue that someone like memory address randomization would be a much more major change, since it breaks so many things (well, it breaks things that shouldn't be doing what they are doing, for the most part...). J.Ja

Spitfire_Sysop
Spitfire_Sysop

If you don't pay much attention to the kernel then this change is obviously transparent. I think that: "Btrfs data scrubbing and automatic defragmentation" is a huge step for a kernel. Being on the cutting edge of file system technology is something that Linux has always done well. What about "A memcached-like system for the page cache."? New ways of handling memory is also huge and could have performance benefits in the future. I restate that these things are both transparent to the user but let's not act like they are below us.