Kernel 2.6.12 Released

Mad Merlin writes "Linux kernel 2.6.12 has been released! Kerneltrap has a brief summary on it. The changelog is only partial however: 'The full ChangeLog ended up missing, because I only have the history from 2.6.12-rc2 in my git archives, but if you want to, you can puzzle it together by taking the 2.6.12 changelog and merging it with the -rc1 and -rc2 logs in the testing directory. The file that says ChangeLog-2.6.12 only contains the stuff from -rc2 onward.' As always you can find the changelog and the source at kernel.org"

Comments

Just after ATI...

[xafan]

Just after hell froze over and ATI released new video drivers for Linux specifically supporting 2.6.11, 2.6.12 gets released.

Let me start off the collective "ARRGGGHHH!"

Re:Just after ATI...

[Golradir]

O how I like nVidia...

Re:Just after ATI...

[Jarlsberg]

Wow, every negative, but valid, comments immediately gets marked as a troll. ;=)

But, does 2.6.12 really break the new ATI drivers? That`s kinda wacky...

Re:Just after ATI...

[hungrygrue]

I know that the ATI Direct Rendering Manager is currently broken for 2.6.12 - at least building for amd64. Oh well, no TuxRacer for a couple days until it gets sorted out.

Re:Just after ATI...

[mr_z_beeblebrox]

Just after hell froze over and ATI released new video drivers for Linux specifically supporting 2.6.11, 2.6.12 gets released.

As someone who specifically uses 2.4.x kernels due to certain support issues, I give you permission not to upgrade. Matter of fact to go further I give you this checklist to decide any and all software upgrades in the future:
Does your current software solve your needs?
Does the upgrade mess with something you care about?
Does the upgrade fix a vital security issue?
Are you a developer?
I would discuss the answers in an if.. then... else sort of way. But, if you can upgrade your kernel you should be able to figure it out. Oh, one more thing, if you do not know the answer to any of these questions, you shouldn't even think about upgrading. Do not run code simply because it has been written. Code is written to address needs, use the code that was wrtten for yours and be happy that there is code for other people to.

Re:Just after ATI...

[neon-fx]

Would "I fogot to compile the APM power managements into my last kernel and have been to lazy to do another compile with 2.6.11" be a good enough reason to upgrade to the new kernel?

Re:Just after ATI...

[End11]

laziness is a good enough reason to do anything.

Re:Just after ATI...

[David Hume]

laziness is a good enough reason to do anything

I thought laziness was a good enough reason not to do anything....

Re:Just after ATI...

[petermgreen]

btw are there any sites that list kernel security issues and more importantly which versions are safe from all known issues?

i.e. a site that will tell me which versions are safe to use without worrying about backporting security fixes.

Re:Just after ATI...

[Anonymous Coward]

Hard work often pays off in the future.
Laziness always pays off now.

Re:Just after ATI...

[Anonymous Coward]

I thought laziness was a good enough reason not to do anything...

Nah, it's an excuse not to, and a reason to do something.

Re:Just after ATI...

[SQLz]

I'm sure they will work with 2.6.12 fine, after all, its still Linux isn't it?

Re:Just after ATI...

[Anonymous Coward]

a freaking men

Re:Just after ATI...

[htd2]

I guess this also explains why Linux is doomed

Re:I've been waiting...

[the_sidewinder]

The Penguin has landed!

Wow!

[Anonymous Coward]

Let's see, Open Solaris, the latest Linux Kernel ... what next, Longhorn?

Re:Wow!

[Anonymous Coward]

Would GNU Hurd be too much to ask? :-)

Re:Wow!

[frodo from middle ea]

Umm, wouldn't armageddon come first before Longhorn...and if armageddon does indeed come then we can safely say RMS taking a bath is not far away...

Re:Wow!

[Dolda2000]

I think that it would be a fairly safe bet to say that there will be another twelve revisions of the Linux kernel before we see Longhorn.

Now, there's the right message

[NitsujTPU]

The full ChangeLog ended up missing, because I only have the history from 2.6.12-rc2 in my git archives, but if you want to, you can puzzle it together by taking the 2.6.12 changelog and merging it with the -rc1 and -rc2 logs in the testing directory

Nothing instills confidence in those who are not convinced that Linux is mature enough for their application like the messages: "I was too lazy to download these files to put together a changelog" and
"the changelog wasn't in our CMS."

Re:Now, there's the right message

[tvon]

I agree, that's just embarassing.

Re:Now, there's the right message

[syynnapse]

I take it as a good indicator of who is working on the code... people who cant get their lives together often write great code.

Re:Now, there's the right message

[Anonymous Coward]

"the changelog wasn't in our CMS."

Are you living under a rock or did you probably notice that the core linux developers switched their cms?

mod him down (-3 flamebait)

Re:Now, there's the right message

[NitsujTPU]

Whoooaaa buddy.

I'm a card-carrying member of the FSF, a Linux user, a supporter, and didn't mean to HURT anybody. I meant to make an obvservation, and hope that it perhaps HELPS somebody.

Re:Now, there's the right message

[Chiisu]

Sums up Linux perfectly

Re:Now, there's the right message

[caluml]

That's why those people should employ other people to tell them that it's mature enough.

Re:Now, there's the right message

[Curtman]

Embarassing to who?

People who disagree with me are by definition crazy. (Until I change my mind, when they can suddenly become upstanding citizens. I'm flexible, and not black-and-white.)
- Linus Torvalds

There you have it.. Clearly he's in the right.

Re:Now, there's the right message

[Anonymous Coward]

"I'm a card-carrying member of the FSF, a Linux user, a supporter, and didn't mean to HURT anybody."

Slashdot's changed from the sub-20k ID days. It appears to have unintentional but the tone of your first post will attract moderation from anti-Linux fanatics and paid astroturfers like flies to Joe's.

Re:Now, there's the right message

[TorKlingberg]

Still better than "We won't tell you what this 20 MB binary patch does, but install it anyway. Trust us."

Re:Now, there's the right message

[Curtman]

the core linux developers switched their cms?

Don't mean to beat a dead horse here or anything (maybe I do actually)... But the decision to use Bitkeeper in the first place was a bad one. So was the decision to reinvent the wheel. SVN works fine.

Re:Now, there's the right message

[Jugalator]

Oh come on, what does Linux maturity for applications have to do with easily complete and clearly formatted changelogs, really? It helps if you're a developer, but that has nothing to do with maturity for application usage... with the end users.

Re:Now, there's the right message

[Anonymous Coward]

SVN is centralised. Thats a big no-no for real work of that scale.

Re:Now, there's the right message

[Minna Kirai]

SVN works fine.

SVN is designed to replace CVS, but Bitkeeper and git are entirely different software categories (just like Java and gcc are both programming tools). The system topology during execution of the CMS is completely different, because BK and git work without a connected network. For comparison, run subversion without any connection to the repository at all, and test how useful it is. And if that works, delete the repository and try again.

Re:Now, there's the right message

[NitsujTPU]

Yes, Slashdot is a different place :-(

Re:Now, there's the right message

[NitsujTPU]

Try telling that to a comporate entity. With the explanation in place (Linus' explanation as to why) it's fine, however, the blank statement would leave managers somewhat concerned.

Ironically, even ones who wouldn't have been concerned if it had never been mentioned at all.

Quite a few businesses put a greater value on the paper trail than the quality of the system.

Re:Now, there's the right message

[petermgreen]

using bitkeeper was only a mistake if viewed in a narrow light. if it wasn't for bitkeeper i suspect linux would still be struggling on with no version control at all.

afaict the basic story goes as follows.

linus fought off version control and in doing so painted himself into a corner.

bitkeeper came along and exploited that corner to get linus to use it.

linus was forciblly stopped from using bitkeeper and made a crude but workable version control system that suited his way of working based on his experiance using bitkeeper.

some people bend thier way of working arround the tools they have. others try to make or find tools that match thier way of working. Different approaches which suit different personalities but i wouldn't say either is wrong.

Re:Now, there's the right message

[Curtman]

because BK and git work without a connected network.

So git allows people to commit to and checkout from the repository with no access to it. That is indeed impressive technology. We need a web to git gateway then so I can cancel my ISP connection.

There's no reason you can't mirror an SVN repository locally, then generate a diff for submission later.

Re:Now, there's the right message

[Curtman]

linus was forciblly stopped from using bitkeeper

That was exactly the argument against using it in the first place. With proprietary closed source software that is the risk. It's unfortuate that one of the most respected GPL'd projects out there went down this path. On the other hand, it gives us something to point to the next time..

X11 Binary only video drivers are a similar situation. They seem okay now (to some people), but later when nVidia decides they aren't supporting older chipsets to get people to buy new cards this will be a huge problem.

Re:Now, there's the right message

[man_of_mr_e]

Strange then that even the Subversion developers disagree with you that SVN "works fine" for Linus's needs.

You would think that if SVN did in fact work fine, that the SVN developers would be the first to be championing it.

Re:Now, there's the right message

[Curtman]

even the Subversion developers disagree with you that SVN "works fine" for Linus's needs

Links? Linus got by with basically nothing but tar, gzip, diff, and patch for a long time. I think the fact that many of the developers do use SVN is proof that it can work.

Re:Now, there's the right message

[man_of_mr_e]

have you been living under a rock? Google "Linus Subversion" and it's the first link.

http://subversion.tigris.org/subversion-linus.html

Re:Now, there's the right message

[Anonymous Coward]

There is no "the repository" with git.

There isn't any reason you can't mirror an SVN repository locally then generate a diff unless you actually want useful information and changesets.

Quick, better tell IBM and Intel and other companies spending 10K per seat on bitkeeper that they're dumb and you're smart and they should be using SVN.

Re:Now, there's the right message

[DShard]

The key difference isn't the "disconnected from the network" bit. The thing that a distributed CMS does is fundementally different than a central repository, since there is none. There is no single source of truth. Everybody is on equal footing with their own source tree and you are free to merge from everybody, nobody or just a select few. This isn't a model that works well for a single user application that only 3 people develop for... but it works wonders for a 6 million line count kernel developed by hundreds of people (granted only a fraction of that are regulars). Using SVN on the kernel would make revision control useless.

Re:Now, there's the right message

[Curtman]

have you been living under a rock?

Worse. I'm back in University finishing my degree. Thanks.

This quote from that page is more along the lines of what I would consider an ideal:

Someday, Subversion may have the features Linus needs, but they're just vaporware until then

I've been using SVN myself for a while now, so please excuse me if I'm coming across as a Subversive zealot, I'm really not. I just think its wasteful to write git from scratch, while the current projects just need a bit of tweaking to do what is needed. Not invented here isn't any better in open/free software than it is elsewhere IMHO.

Re:Now, there's the right message

[man_of_mr_e]

Unfortunately, Linus didn't have the luxury of waiting for Subversion to mature to his needs. He needed something, and needed it now, so he cobbled together a simple solution that worked for him. Who's to say that Git will be a lonter term thing or not, but for now it works for him.

Re:Now, there's the right message

[Curtman]

There is no single source of truth.

I don't see how this is an advantage. In the good old days everyone would just sync with Linus' tree and thats the way we liked it. :)

Re:Now, there's the right message

[Trelane]

That was actually precisely the reason I chose an ATI card despite its poorer performance (and support) under Linux. When the card gets Old Enough, ATI works with the X devs to ensure that they can get the information they need to write drivers.

Re:Now, there's the right message

[Curtman]

Unfortunately, Linus didn't have the luxury of waiting for Subversion to mature to his needs.

So... SVN would have rejected his patches based on what? His agreement with Bitkeeper not to work on any competing products?

Re:Now, there's the right message

[Curtman]

That was actually precisely the reason I chose an ATI card despite its poorer performance (and support) under Linux.

Same here, and its the reason I still use my old Radeon All In Wonder 7200 in my desktop, and a Rage128 in my media box while the newer one still sits in its box.

Re:Now, there's the right message

[timeOday]

Compared to what? Windows and MacOS don't give you any detail remotely comparable to a changelog. Compared to BSD, I suppose you may have a point.

Re:Now, there's the right message

[man_of_mr_e]

Rejected what patches? You think he could turn SVN into a distributed CMS with a few patches?

Re:Now, there's the right message

[Curtman]

You think he could turn SVN into a distributed CMS with a few patches?

No more so than starting from scratch would require. This seems to be the approach svk is taking, and seems to be pretty far along.

Re:Now, there's the right message

[man_of_mr_e]

You really don't understand the "immediate need" concept do you? Linus threw git together in a few hours, it would take him days just to become familiar enough with the SVN code to even attempt a single patch, much less everything he'd need.

Re:Now, there's the right message

[Curtman]

Linus threw git together in a few hours

He did? I could have sworn that Git development began in early April with many people helping, and the first merge wasn't made until a couple weeks later, and why Git developent is ongoing. That was most of the reason why 2.6.12 took 3 months to get out the door.

Re:Now, there's the right message

[Anonymous Coward]

You're a loser, I though you were told to face the corner and not turn aroun.
As for the rest of us, we are doing well with our NVidia cards.

Re:Now, there's the right message

[Curtman]

As for the rest of us, we are doing well with our NVidia cards.

Poor bastard. You only think you are.

Re:Now, there's the right message

[Vlad_the_Inhaler]

Summing that up: ATI cards are better when they have been around a while, nVidia cards are only much use when new.

Looks like ATI should think about matching nVidia in offering binary drivers for their new cards, in addition to the open drivers later. They do want to sell new hardware, right?.

Re:Now, there's the right message

[joel48]

Exactly how is that any better?

I'm a linux fan, user, and admin - but I'm not for cheap shots simply for their own sake. Perhaps I'm missing the difference, but with no [full] changelog, that's essentially the same thing anyway.

Re:Now, there's the right message

[po8]

Good idea. In fact, such a good idea that they already do.

Re:Now, there's the right message

[toadlife]

Oh come on! It's time to take off the tinfoil hats guys.

Nvidia dropping support for old cards like you suggest is silly, as it would be extremely bad publicity for them, and would surely cost them sales (people would switch to ATI) from linux users.

NVidia still supports jsut about every card they've ever made, in both linux and FreeBSD.

Nvidia just dropped support for a few legacy cards in their unified driver, but say they will continue to support them in a special legacy driver release. Here are the offending cards:

From the readme:

"Below are the legacy GPUs that are no longer supported in the unified driver.
These GPUs will continue to be maintained through the special legacy NVIDIA
GPU driver releases."

RIVA TNT
RIVA TNT2/TNT2 Pro
RIVA TNT2 Ultra
Vanta/Vanta LT
RIVA TNT2 Model 64/Model 64 Pro
Aladdin TNT2
GeForce 256
GeForce DDR
Quadro
GeForce2 GTS/GeForce2 Pro
GeForce2 Ti
GeForce2 Ultra
Quadro2 Pro

There are Open source drivers that work with nvidia chips anyway, so if some day Nvidia completely drops support for some old-ass graphics card, what's the big deal? Do you really want to play the lastest version of tuxracer with a seven year old graphics card?

Besides, this is commodity X86 hardware we're talking about. It's not even really designed to last more than 5 years.

Re:Now, there's the right message

[Curtman]

It's not even really designed to last more than 5 years.

I haven't bought very much new hardware in the past 5 years. The stuff I've got works great. Tuxracer runs just fine.

What pisses me off most is the clowns who mail dri-devel or come in #xorg asking for help with their proprietary nVidia or ATI cards every hour or two when it says right in the topic "Closed driver questions in #ati or #nvidia". Don't people realize that is insulting to Xorg developers? People get upset when their drivers don't work with Xorg CVS. It's rediculous.

I don't understand why someone like you would get defensive about nVidia though. Do they send you a cheque every month? All those drivers do is subtract motivation for people to get involved in Xorg development, or Gatos, or whatever. I'm in favour of purposely breaking compatibility on a monthly basis if need be.

Re:Now, there's the right message

[Minna Kirai]

everyone would just sync with Linus' tree and thats the way we liked it.

Which forced Linus to spend 110% of his time sorting through all the patches he got sent, leaving him no chance to do programming of his own. And that wasn't the way he liked it.

An unconnected CMS allows that manual labor to be handled by software. Maybe it'll help if you think of BK and git as superior diff/patch utils that include version-history in the metadata. They're like patches that can be applied more cleanly because they know which versions the changes were made on, and how your files have changed in the meantime.

Re:Now, there's the right message

[ccp]

Disclaimer: strong language directed not to you, but to your post.

Try telling that to a corporate entity.... the blank statement would leave managers somewhat concerned.

Who the fuck cares about what a corporate entity likes?

Why should I (or anybody else) care about concerned managers?

Linux is free and Free, is a GIFT to them! They should be grateful.

Or is it that as soon a little money appeared in the horizon all of us turned whores?

Cheers,

Carlos Cesar

Re:Now, there's the right message

[FooBarWidget]

Concerned corporate managers are supposed to sign a $70.000 contract with RedHat or other commercial Linux distributors, not downloading random tarballs from Slashdot-linked sites.

Re:Now, there's the right message

[Trelane]

Actually, I'd summarize it thusly: new ati hardware is alright (thanks to the drivers they provide). Older ati hardware is excellent (thanks to the information they provide).

Nvidia's hardware is generally excellent because their drivers seem to be very high-performance. Caveat is that they don't support the X devs nearly as well as ATI does, so you're much more stuck if they stop supporting you than if you were using ati.

Re:Now, there's the right message

[Minna Kirai]

This seems to be the approach svk is taking, and seems to be pretty far along.

Sure, it's far enough along for their needs... which so far don't overlap with Linus's needs at all. SVK allows a developer to continue to operate when her repository connection has been temporarily lost, which is a good feature that CVS should've had a decade ago.

But it still requires that some formally centralized repository exist. CVS, SVN, and SVK all operate on the concept of checkout/checkin. What Linus needed was the feeling of pull/push operation, which is what BK and git provide.

Re:Now, there's the right message

[Minna Kirai]

Perhaps I'm missing the difference, but with no [full] changelog, that's essentially the same thing anyway.

You HAVE the info- it's just a question of converting it into a managable format.

With Linux, anyone who cares can get not only the exact lines of code that changed since the prior release, but also the submitted patches that produced those changes. (Just subscribe to the ML). Condensing that info down into a concise, readable changelog is just a matter of desire and effort. A maximally paranoid person would prefer to diff the code in addition to reading the changelog, in case human error had caused any omissions.

A closed source binary patch, on the other hand, doesn't give the recipients any path to generating the changelog on their own. If the publishers didn't include one, you are absolutely out of luck. And if it is incomplete/wrong/censored, there is no recourse. When Alan Cox had to censor a few entries in his Linux changelog, the changes themselves were still out in public view.

Re:Now, there's the right message

[NitsujTPU]

Lets say that you're a programmer, working at a company. You develop a solution under Linux, and want the company to use that.

Instead, you are told to redo it under Windows. You can't convince your program manager to use the Linux solution. They won't budge.

Now, you see the issue.

Re:Now, there's the right message

[toadlife]

"I don't understand why someone like you would get defensive about nVidia though. Do they send you a cheque every month?

No, no check from them, but they do make great drivers for FreeBSD, the OS that I use for my desktop. That's why I am partial to them.

"All those drivers do is subtract motivation for people to get involved in Xorg development, or Gatos, or whatever. I'm in favour of purposely breaking compatibility on a monthly basis if need be."

Well you go ahead and encourage your friends at Xorg to do that, and we'll all watch helplessly as yet another fork of X takes off. ;)

Re:Now, there's the right message

[majoun]

From NVIDIA's latest Linux driver changelog:

Removed support for legacy GPUs; please see "Chapter 1. Selecting and Downloading the NVIDIA Packages for Your System" in the README for details.

http://www.nvidia.com/object/linux_display_ia32_1. 0-7664.html

Re:Now, there's the right message

[Curtman]

SVK allows a developer to continue to operate when her repository connection has been temporarily lost

What kind of system are you thinking of where that does happen? When my network connection goes down, the files that I have checked out of the repository don't get deleted, I keep right on working.

Re:Now, there's the right message

[Curtman]

"That's why I am partial to them."

Well I guess its no different than why people cheer for sports teams. I find the whole thing silly, but I'm in the minority there too.

"Well you go ahead and encourage your friends at Xorg to do that"

You and I both know that's not going to happen. I was just being dramatic.

Re:Erm

[Anonymous Coward]

No

One thing I'm a bit confused about...

[Sheetrock]

When and why did they stop the system of releasing stable versions on the even minor releases (2.4.x, 2.6.x, etc.) and unstable/development versions on the odd minor releases (2.5.x, 2.7.x, etc.)?

Re:One thing I'm a bit confused about...

[0racle]

When 2.6 was released. It was pretty well publicized around here, have you been under a rock since 2.6.0?

Re:One thing I'm a bit confused about...

[Anonymous Coward]

That troll sucked.

Re:One thing I'm a bit confused about...

[ArbitraryConstant]

Because development was going quickly and they didn't want to lose momentum. We're getting new features much sooner than we otherwise would have.

The downside is that 2.6 kernels are now a regression-fest that makes Windows look positively stable. They claim distros are able to stabalize their own kernels, which is a theory I have yet to see put into practice. The idea now is to find a kernel version that doesn't have any show-stopper regressions for your hardware.

Re:One thing I'm a bit confused about...

[The_Wilschon]

Yes, but if you wanted new features sooner under the old system, all you had to do was use the 2.odd.x series of kernels. I don't see that we gained anything by throwing out the idea of having a stable series of kernels. How can we convince/ask Linus et al to return to the old system?

Re:One thing I'm a bit confused about...

[slavemowgli]

A year ago, and because it works better (the -mm kernels are the testbed for newer features now). It's been on Slashdot, Kerneltrap, lwn etc., too.

Re:One thing I'm a bit confused about...

[Anonymous Coward]

Not enough testers were using the 2.odd series, and no distributions were shipping it.

With 2.6, bugs are being ironed out MUCH faster, and lots of features are stabilizing quickly, which means more features can be added.

Re:One thing I'm a bit confused about...

[ArbitraryConstant]

"With 2.6, bugs are being ironed out MUCH faster"

Unfortunately you can't update to a new release to fix the bugs you have, because every new version is so riddled with regressions that it's just as bad as the last one. Bugs are replaced, not fixed.

Re:One thing I'm a bit confused about...

[The_Wilschon]

Yes, but one of Linux's strengths, and selling points, has traditionally been its stability. If we throw that out the window, then people will start drifting away to other systems, and then we'll have even *fewer* people using the buggy kernels. Calling buggy code "production" is a M$ tactic. And one of the reasons I moved away from that platform.

Re:One thing I'm a bit confused about...

[slavemowgli]

Actually... the regular 2.6 kernels are pretty stable, or at least not more unstable than past stable kernel series like 2.4 were. This is especially true if you run the 2.6.x.y kernels (like 2.6.11.12, currently) instead of staying on the bleeding (mainline) edge by constantly upgrading to the latest -rc or even git snapshots.

I for one think that it's all a good idea: instead of backporting features (from devel to stable, or stable to old stable), we're simply back-porting bugfixes applied to mainline to a relatively short-lived stable off-shot so that people can get fixes without new features that also introduce new bugs. That should help stability.

Outside of that... well, yeah, I guess one could complain about the lack of proper regression testing, QA prior to releases, release schedules and all that. But the truth is that these things have not only hindered but actually *helped* a great deal when it comes to Linux development, so I don't see why they should be abolished. In the end, it's the developers that decide how things should work - and that's good, because it's the developers, not management (this includes everyone who's not involved with development), who know best.

That being said, the statement that distros don't roll their own kernels isn't true, anyway. Take Fedora, for example - if you read Dave Jones' blog, you'll notice that a lot of effort *is* going into stabilizing these and making them regression-free. How many distros do you know that ship with vanilla kernels? I know exactly one - Slackware. Every other (big) one uses patched kernels, no matter whether it's Debian, SuSE, RedHat, or whatever else there is.

Re:One thing I'm a bit confused about...

[MooseGuy529]

I don't know when, but the reason was this: Instead of having a 2.6 and 2.7 branch, they wanted to be able to accelerate integration of new developments. So the 2.6 series is under heavier development, and sometimes requires very minor bugfixes and such. Those go in the fourth part of the version number. I personally like this strategy since it means you can still rely on stable versions when you need them, but you also get to use nifty new features without waiting years for the development branch to become stable.

Re:One thing I'm a bit confused about...

[ArbitraryConstant]

"That being said, the statement that distros don't roll their own kernels isn't true, anyway. Take Fedora, for example - if you read Dave Jones' blog, you'll notice that a lot of effort *is* going into stabilizing these and making them regression-free."

Oh they try. Valiantly I'm sure. But they fail. For example, Fedora Core 3 can't see my CD drive unless I put it on a Promise IDE card because of that libata business. I'm sure they'd be able to do better if they weren't starting out with such a disadvantage.

"Outside of that... well, yeah, I guess one could complain about the lack of proper regression testing, QA prior to releases, release schedules and all that. But the truth is that these things have not only hindered but actually *helped* a great deal when it comes to Linux development, so I don't see why they should be abolished. In the end, it's the developers that decide how things should work - and that's good, because it's the developers, not management (this includes everyone who's not involved with development), who know best."

Your pronouns imply you're a kernel developer. Is that the case? You may have a development model that results in features being added faster, but you are not producing kernels that I can risk using for anything important. Judging by the other posts here, I'm not alone. I agree that non-developers shouldn't be in charge of the project, but you should agree that users are free to use something else if what you're making don't meet our needs.

Kernel 2.6 doesn't meet my needs anywhere but my workstation system. It's worth it for the responsiveness, but my workstation is my problem. I can install something else over lunch if I have to. But I will get yelled at if the other machines break, and I don't want 2.6 anywhere near them.

Ultimately I hear a lot of justifications that involve things other than kernels that work robustly from one version from the next, but I don't care about those other things when my system isn't running at all. If you do, that's fine, but I think you should care a little when a Linux update becomes more likely to break stuff than a Windows update.

Re:One thing I'm a bit confused about...

[Anonymous Coward]

Are you actualy aware of how many times you used the word developer/developement in one paragraph? Still not as bad as it could be but this is definitely not a good sign. When people start to talk like Steve Ballmer it usualy means you have to find a new OS.

Re:One thing I'm a bit confused about...

[slavemowgli]

Which words would you use? I dislike buzzwords and empty marketing-speak just as much as anyone on Slashdot, but "developer" isn't one of those buzzwords, I think. Quite the opposite: it's an accurate description of the job done. The same goes for "development".

Correct me if I'm wrong, but I don't see the problem.

Re:One thing I'm a bit confused about...

[Anonymous Coward]

When and why did they stop the system of releasing stable versions on the even minor releases (2.4.x, 2.6.x, etc.) and unstable/development versions on the odd minor releases (2.5.x, 2.7.x, etc.)?
According to the maintainers, the scheme hasn't changed. From the linux kernel faq (http://www.kernel.org/pub/linux/docs/lkml/):

What is an experimental kernel version?

* (ADB)) Linux kernel versions are divided in two series: experimental (odd series e.g. 1.3.xx or 2.1.x) and production (even series e.g. 1.2.xx, 2.0.xx, 2.2.x, 2.4.x and so on). The experimental series are fast moving versions which are used to test new features, algorithms, device drivers, etc. By their own nature the experimental kernels may behave in unpredictable ways, so one may experience data losses, random machine lockups, etc.

Furthermore, if you bothered to look at Kernel.org, you'd see:

"The latest stable version of the Linux kernel is: 2.6.12"

Re:One thing I'm a bit confused about...

[iabervon]

The problem was that starting an unstable series was exactly the kind of fork that causes problems with development. There was a substantial period when the functionality needed to run a system with recent software and on recent hardware was only in the unstable one of the official series, and was backported to the nominally stable series by each distro individually separately and to different degrees. Furthermore, there was constant pressure to add new features to the officially stable series, rather than to the unstable series (or backported from the unstable series).

The central problem was that series progressed from unstable to stable to obsolescent to non-functioning. The solution is to always have a place for unstable features (the -mm series) and a place for stable features (the mainline series), and features move into the stable series as they become stable, rather than requiring a major upheaval and years of fussing. Then there needs to be something that corresponds to the period where mistakes in a release are fixed in a release that doesn't include anything else, even new features which are extensively tested; this is only useful until there are no known regressions in the next version with added features.

The reason that the numbering changed is that, were the numbering maintained, the recent releases would now be 2.16.11, 2.18.0, and 2.19.0 (assuming that the new system was adopted at 2.6.6, the old numbering would make what was 2.6.7 into 2.8.0, and so forth, replacing one point release with two minors, which would just be silly). Also, it would be confusing if 2.17.x included stuff that wasn't in 2.18.0, was in 2.19.0, and was in 2.20.0; this is what happens to anything that's still in testing when a release is made and then stabilizes. Since there's always stuff in testing, no stable release could ever be made without having the existance of features be confusingly non-continguous.

In any case, 2.6.x.y is now about as stable as 2.4.x was during the period before distros started moving to 2.6.

Re:One thing I'm a bit confused about...

[oliverthered]

' Not enough testers were using the 2.odd series,'
That was because it was buggy, now everyone gets to test the the 2.odd series because it's the the 2.even series

Re:One thing I'm a bit confused about...

[Anonymous Coward]

In the end, it's the developers that decide how things should work - and that's good, because it's the developers, not management (this includes everyone who's not involved with development), who know best.

Bullshit. Developers by and large have no frickin' clue how to build products -- product is more than just a random tar ball of source code. They do not know the "best" how to cater to customers or users. They only know how to cater to themselves.

Now if your pet project is only intended for develoeprs to ever use, that's fine. But if you want to build a little more than a hobby project that pretty quickly becomes an unrealistic expectation.

Theo says it's crap

[Adult film producer]

So I'd stick with Windows if I were you.

Re:Theo says it's crap

[DroopyStonx]

Theo Huxtable?

Linux+OpenSolaris

[Anonymous Coward]

Random question: Does anyone think there will be anything benificial to linux to borrow from solaris now that the source is out, or does their license even allow this?

Re:Linux+OpenSolaris

[Anonymous Coward]

OT Answer: When OpenSolaris was announced, a lot of the idiots round here started posting excitedly about how all the features from OpenSolaris could be ripped out and put into Linux (NOT the point of open source). However, Sun licenced OpenSolaris under the CDDL, which is incompatible with the GPL. This made a lot of people very upset.

So the answer is no.

Re:Linux+OpenSolaris

[Arker]

Does anyone think there will be anything benificial to linux to borrow from solaris now that the source is out, or does their license even allow this?

Short answer: No, and no.

Longer answer, while there are a few places Solaris still has an advantage, you can't just rip code out of one and stick it another. The structure of the code is quite different, so an implementation in one codebase just won't transfer to another cleanly.

And two, the CDDL, besides being horridly written, is clearly and intentionally not GPL compatible, so even if you could transplant code like that technically, it wouldn't work legally.

Re:Linux+OpenSolaris

[zymano]

functions can't be used ?

Re:Linux+OpenSolaris

[njcoder]

Looks like it started on day one

Re:Linux+OpenSolaris

[Anonymous Coward]

Ummm, he already answered that, dumbass. See above.

Re:Linux+OpenSolaris

[Anonymous Coward]

Simple functions with little or no dependencies can obviously be transfered. That's what he was referring to, and you are the dumbass for failing to understand that.

Re:Linux+OpenSolaris

[Anonymous Coward]

What about just looking at the code? Doesn't that contaminate you if you are a kernel developer? "Transplant" of code can happen in more subtle ways.

Re:Linux+OpenSolaris

[rpozz]

Before OpenSolaris was released, there was a load of complaining about CDDL and GPL incompatibility, and some people were even trying to push the insane idea that Sun would pull a SCO. Everyone seemed to eventually come to the sensible conclusion that people who contribute to the Linux kernel should stay away from the Solaris kernel.

On the first day of release, people who contribute to the Linux kernel start looking through the Solaris kernel. What the FUCK are they thinking?

Re:Linux+OpenSolaris

[cortana]

Except that the COPYRIGHT LICENSE PREVENTS YOU FROM DOING SO.

If you're going to be a rude asshole, at least make sure that what you say is actually correct.

Re:Linux+OpenSolaris

[petermgreen]

it depends. copyright isn't supposed to protect the idea itself just expressions of it. With some high profile projects (think java class libraries) paranoid "clean room" methods are justified because the code is going to end up looking fairly similar even if written by totally seperate coding teams as the interface constrains it so much.

otoh with say kernel code the kernels are often so different in structure that its not such a big issue as only the general ideas and not any of the details will be able to come over anyway.

Re:Linux+OpenSolaris

[Anonymous Coward]

How many simple functions with few dependencies are there in an OS? And how many of those contain the cool functionality that you're after anyway?

Stop being a shit.

Re:Linux+OpenSolaris

[Anonymous Coward]

I was referring to it being technically possible to transfer functions, and look at the comment I was replying to and work out why I was being 'rude'. Unless you are the AC I was replying to?

Re:Linux+OpenSolaris

[Anonymous Coward]

GPL incompatiblity is annoying... but more annoying is the Yet Another Open Source License (YAOSL) factor. Open Source isn't very meaningful if each project's code is walled up behind its own license -- *and* protected by patents and carefully worded sabotage clauses in the license... the CDDL is rather dodgy license in many ways, designed to pass muster with the OSI while still being obnoxious (not difficult, they rubber-stamp all kinds of shit).

x86_64 ctl32 removed

[etymxris]

It's some compatibility thing that allows 32 bit apps to run on a 64-bit OS. Shouldn't be a problem for GPL drivers, but will break older proprietary drivers. I believe nvidia just updated their drivers to be compatible with 2.6.12. But VMWare still won't work, last I checked.

Re:x86_64 ctl32 removed

[salimma]

That better be 32-bit drivers not apps, or we won't be able to run OpenOffice...

Re:x86_64 ctl32 removed

[Jethro]

Oh, is that why VMware isn't working now? I just got it working with 2.6.11.12. Oh well, I'll wait for the next any-any patch.

Stupidest thing I've ever read, right there

[Anonymous Coward]

People who lack organizational skills do not write great code.

Re:Stupidest thing I've ever read, right there

[syynnapse]

nobody said "lacking in orginizational skills".
asshat.

Priorities

[etymxris]

Sure, a full changelog would be nice. But Linus, I imagine, isn't too worried about appearing here isn't worth the effort. His time's better spent on actual kernel code.

This is the type of thing that happens when engineers manage projects rather than business people. That's not a criticism.

Re:Priorities

[etymxris]

Accidentally deleted some text there. Should read as follows:

Sure, a full changelog would be nice. But Linus, I imagine, isn't too worried about appearing "professional" here. A full changelog just isn't worth the effort. His time's better spent on actual kernel code.

This is the type of thing that happens when engineers manage projects rather than business people. That's not a criticism.

Re:Priorities

[Speare]

This is the type of thing that happens when engineers manage projects rather than business people.

Yeah, I hate it when engineers manage the business people.

Making it stable...

[fprog]

They should really try to froze features,
at least subsystem by subsystem,
for a couple of months and perform
a deep code review (ala OpenBSD)
for bug hunting and security analysis.

I can understand that some part of the kernel
still needs heavy development
(ReiserFS, VM, some specific broken drivers),
but other parts should be revised
and certified gold bug free.

At least that would give us a roadmap,
on what is to be fixed before
they jump to 2.7.x series.

I mean what's the point to break stuff
at every .x release in a stable series,
that doesn't make any sense to me.

How are 3rd party drivers people, applications
are supposed to "trust" a 2.6 kernel,
if it break stuff continuously.

"You're Nvidia or ATI card works in 2.6.x but not in 2.6.x+2,
and VMware doesn't work in 2.6.y but only in 2.6.y+1"

As long as they keep breaking stuff,
I'm keeping my "stable" linux servers
on the 2.4.x series.

Re:Making it stable...

[Ambassador Kosh]

If you handn't noticed that all of those things that break are binary only drivers. vmware, ati and nvidia all depend on a binary kernel module. I have been using 2.6.x for a long time on desktops, servers and laptops and had no issues however the only proprietary apps that are even installed are acrobat and opera and those are only for testing stuff.

For servers none of those should be an issue at all.

Re:Making it stable...

[njcoder]

Just make it easy
for us to read and do not
try and write haikus. :)

Re:Making it stable...

[ArbitraryConstant]

"If you handn't noticed that all of those things that break are binary only drivers. vmware, ati and nvidia all depend on a binary kernel module."

If that were true, I wouldn't have to have a Promise IDE card to be able to use my CD drive. It used to work... but now using the SATA and PATA ports at the same time is impossible with Linux.

Re:Making it stable...

[Anonymous Coward]

"a deep code review (ala OpenBSD)"

In what way can a code review be deep when the release cycle is 6 months in an absolute fashion? Subtract the time for freezing when a new release is prepared. So much for the 'deep code review'.

Don't believe everything Theo tells you.

Re:Making it stable...

[jadavis]

I can understand that some part of the kernel
still needs heavy development
(ReiserFS, VM, ...

Speaking of VMs, I'm a little confused about the topic. Can anyone direct me to some good material that explains the differences between various VM systems? Specifically, I'm concerned about overcommitting memory and the OOM killer in linux. Do any other OSes have an OOM killer? Why or why not? If an OS overcommits memory, how can it not have an OOM killer? Does setting "vm.overcommit_memory = 2" disable the OOM killer, or just make it less likely?

Re:Making it stable...

[Minna Kirai]

If you handn't noticed that all of those things that break are binary only drivers. vmware, ati and nvidia all depend on a binary kernel module.

Please install the complete ov511 camera drivers (from source) into Linux 2.6.12, then say that again.

Re:Making it stable...

[Bachus9000]

Is this a new 2.6.12 thing? I'm running a sata and a pata drive in my Debian system running 2.6.11 without any such problems.

Re:Making it stable...

[ArbitraryConstant]

Seems to be okay if they live on different chips, which is probably why the Promise card works. Your motherboard might have a separate SATA chip.

I don't remember which kernel versions are affected, but it's a relatively new thing. It's caused by libata, if I recall my debugging correctly. I gave up trying different versions long ago (which is why I've forgotten most of the specifics), I just use the Debian images and try to work around the bugs they have.

Re:Making it stable...

[Anonymous Coward]

be gone silly man
do you live under a bridge?
please shower today

Re:Making it stable...

[Ambassador Kosh]

I am using 2.6.11 right now I have a a via kt880 chipset and I have the SATA and PATA on the same chip and my SATA hard drive and both cdroms that are PATA are working without issue.

The motherboard is SY-KT880 Dragon 2 v2.0

Yes I know that board has two SATA controller chips but I am using the VIA one for both the SATA and PATA functions.

On the plus side this has also been a great board so far. Everything seems to work fine under linux with free drivers but I have not tested the other SATA controller yet.

Re:Making it stable...

[Ambassador Kosh]

Hmm it looks like someone abandoned that software. However drivers do break under windows also. I have an older video card and have run into printers, scanners etc that are only a year or two older then windows xp that won't work with it because the drivers where never made.

Re:Making it stable...

[ArbitraryConstant]

I'm happy for you that your computer works, but it doesn't refute my point.

2.6 kernels have a much higher probability of something going wrong. That'll cost you over time if you use them enough. You can't apply updates as much because they're forever breaking things, etc.

Re:Making it stable...

[Ambassador Kosh]

I have used them since 2.6.1 or so. When is it going to start costing me more over time? I pick hardware carefully to not have issues and I have done that since I have been using computers. I pick software even more carefully and nothing the business relies on can be proprietary.

Re:Making it stable...

[ArbitraryConstant]

Again, I'm happy for you that you haven't had problems but that experience doesn't generalize.

Maybe?

[Saeed al-Sahaf]

Could this be part of the reason hardware manufacturers don't put a high priority for Linux drivers?

Re:Maybe?

[cuerty]

If they just desing right the drivers or release the spects, ...

But the problem is, nothing in the arquitecture of how the drivers works has change, as long as I see in the changelog, so why should that driver not work? It seems like a "if kernel_version 2.6.11 then wait another six months".

Re:Maybe?

[big_groo]

This is precisely why I choose Nvidia hardware - and I always will. Nvidia took the time to ensure that I can have accelerated graphics on my choice of OS, so I will reward them with my pocketbook. You should do the same.

Re:Maybe?

[Dolda2000]

At the risk of being modded off topic -- have you tried that Slashdot W3C validation link in your sig recently?

I can't believe Slashdot would be so embarrased by their ugly HTML to actually ban the W3C validator...

Re:Maybe?

[flithm]

Could this be part of the reason hardware manufacturers don't put a high priority for Linux drivers?

No.

You don't need to release a new driver with every kernel release. ATI just has no idea what they're doing that's all.

A good example: When 2.6 first came out, the nvidia 2.4 kernel module could be made to compile with only minor modifications! I upgrade kernels with every release, and I've never had to go out and find a new driver release.

But to answer your question, the reason why most manufacturers don't put a high priority on Linux support is because the amount of effort it takes is simply not worth the trouble. Linux just doesn't have the install base.

On the other hand, if you're a cool company, and a lot of your engineers, devs, etc already use Linux, or *BSD then the effort it takes is much smaller, because they already know the system.

Ever compared the install between ATI and Nvidia. ATI's is a freakin' joke. It's horrible.

You know they basically took some of their people, forced them to slap something resemembling driver support together and released it on the public.

I agree with the grandparent post. Reward nvidia with your money. Quit your bitching, stop all the silly petitions to try to get ATI to release decent drivers. All that does is let them know they can do it and people will still buy them!

Re:Maybe?

[Mr. Underbridge]

I can't believe Slashdot would be so embarrased by their ugly HTML to actually ban the W3C validator...

Wow. That's freaking incredible.

Re:Maybe?

[gnarlin]

Then perhaps you should take a look at this working free r300 driver It may not be fully stable yet (although is has worked smoothly for me for about a week) but people seem to have forgotten the fight for getting free software drivers for their graphics cards indstead if kissing the asses of nvidia and ati for not releasing any source code or much specs. I will always choose the card which has a free driver over any that don't. Even if the perfomance is not on par with the best.

Also worth mentioning is the open graphics project.

Re:Maybe?

[Anonymous Coward]

Ever compared the install between ATI and Nvidia. ATI's is a freakin' joke. It's horrible.

Standard Gentoo comment (valid): emerge ati-drivers && opengl-update ati

Re:Maybe?

[smittyoneeach]

Absolutely. Let the market be the driver. Not the courts. Not some ideology.

Re:Maybe?

[nathanh]

I agree with the grandparent post. Reward nvidia with your money.

Reward nvidia for releasing *binary* drivers? Anybody with an interest in the long-term success of Linux should be *punishing* nvidia by NOT buying their hardware.

Reward the companies that produce open source drivers, or publish specs, or help the developers of open-source drivers. Don't reward companies who are destroying the core value of Linux.

Re:Maybe?

[flithm]

Open source isn't the be-all end all. Nvidia puts out a great product and a good driver to match. I fail to see how doing this "destroys the core value of Linux."

Without a proper nvidia driver, Linux would be basically useless for any real modern desktop use, as all we'd have is driver support comparable to say that of any of the other open sourced ones, SiS, via, intel, matrox, the open source nvidia, or ATI drivers. Which by the way all tremendously suck. Sure basic 2d operations are supported, but that's it.

nvidia is under no obligation to release the internals of their product, and why would they in such a competetive market!

Punishing nvidia for running a good, smart business, and support free and alternative operating systems with quality products is an absolutely ridiculous thing to say.

Before you spout your mouth off like that, I'd like to see you create and maintain the number one graphics card company in the world, then release the source code to your driver which would give your competitors a HUGE leg up on understanding the internals of your product.

Don't be such an idealistic ass. It's people like YOU that destroy the core value of Linux.

Re:Maybe?

[drsmithy]

But the problem is, nothing in the arquitecture of how the drivers works has change, as long as I see in the changelog, so why should that driver not work?

Because unlike every other remotely mainstream OS out there, Linux makes no attempt to preserve binary compatibility across kernel releases.

Re:Maybe?

[Saeed al-Sahaf]

A good example: When 2.6 first came out, the nvidia 2.4 kernel module could be made to compile with only minor modifications!

I'm sorry, I'm not savvy to the mechanics of writing drivers. I, and the VAST majority of people, are not going to fuck around under the hood of a video card driver to make it work, we will simply use a different card, or a different OS.

Re:Maybe?

[Frequanaut]

As if the market isn't an ideology.

Re:Maybe?

[nathanh]

Open source isn't the be-all end all.

With Linux, yes it is. If you don't care about open-source then use something else.

Don't be such an idealistic ass.

Idealism is not a dirty word. Idealism means seeing the bigger picture and foregoing fleeting fancies in the pursuit of long-term success. If you're too short-sighted to understand that then perhaps Linux is not for you.

Re:Maybe?

[flithm]

No No, you misunderstood what I said.

I didn't mean you have to change the driver itself. This would be impossible as it's a closed source binary only driver!

I just meant minor modifications to the system. And I was talking about a fringe case: the case where a whole new version of the kernel is released (not just a point release). If you think about this it speaks very well on nvidia's behalf.

Also, in their favor, they released a 2.6 version soon after the release of 2.6. (No modifications necessary).

My point, which you completely missed, was that you do NOT need to upgrade your driver every time a kernel point release is made, and I was talking about a very extreme case where nvidia really shone.

Re:Maybe?

[flithm]

Nice way of skirting around the real issue. Please tell me how having a closed source driver destroys the core value of linux. But, in doing so, you must explain how not having this driver at all (which is the only other alternative in todays society) would be a better option.

It's not as if I don't see how having an open source driver would be a really good thing. Of course I do!

In your case, idealism is a dirty word. Your idealistic views keep you from seeing the reality of the situation, which makes you say such nonsense as "destroying the core value."

It's zealots like you that keep a lot of people from adopting Linux, and give the whole community a bad name.

Nvidia is doing everyone a favor by releasing quality drivers.

People like you do a lot of complaining, but do nothing to make things better. And in the process you piss people off enough to turn them away.

When you talk like that, it is you who destroys the future of Linux. If you're too short-sighted to understand then perhaps being alive is not for you.

Re:Maybe?

[flithm]

Just a clarification... when I said having no driver is the only other alternative, I meant because the alternative would be the current open source driver... which is basically unusable beyond the most basic tasks. (ie no OpenGL, or advanced accelerated 2d operations).

Re:Maybe?

[Nasarius]

Idealism is not a dirty word. Idealism means seeing the bigger picture and foregoing fleeting fancies in the pursuit of long-term success. If you're too short-sighted to understand that then perhaps Linux is not for you.

I agree. Unfortunately, the desktop video card market is largely binary (NVIDIA/ATI), and NVIDIA is clearly superior in their support for Linux. There are many theoretical and practical problems with binary drivers, but NVIDIA has gotten it mostly right. ATI's Linux drivers are awful, which is why I'll never buy their cards.

Re:Maybe?

[Saeed al-Sahaf]

Gotcha.

Re:Maybe?

[nathanh]

Nice way of skirting around the real issue.

With Linux, open source *is* the real issue.

Please tell me how having a closed source driver destroys the core value of linux.

The core value of Linux is that it is open source.

It's zealots like you that keep a lot of people from adopting Linux,

Good. Linux will be better off without them if the mere mention of "open source" is enough to scare them away.

Re:Maybe?

[Uerige]

I, and the VAST majority of people, are not going to fuck around under the hood of a video card driver to make it work

That's good, because your'e not supposed to. You, and the VAST majority of people, are supposed to use the kernel that comes with your Distribution.

Re:Maybe?

[nathanh]

I agree. Unfortunately, the desktop video card market is largely binary (NVIDIA/ATI),

All the nvidia and ati cards have open source drivers. The open source drivers sometimes lack features, or don't perform as well, but they are open source. It is not correct to say that the "video card market is largely binary". That is a choice that the end-user makes.

Re:Maybe?

[Anonymous Coward]

Yeah, real big of nVidia to release closed, binary drivers that may or may not run on different configurations and most assuredly will not run on anything other than Linux/i386, and to make no mention whatsoever of hardware specifications or documentation. Really nice, how they've ensured that there will be no non-trivial open development of their drivers by providing a buggy, impractical, but "good enough" solution.

nVidia is ensuring that F/OSS graphics will continue to be criplled by establishing an inadequate and, in the long run, harmful status quo. Until some company comes along that releases specifications---not even drivers, just specifications---for its graphics cards, I will not spend a single penny on them. Closed drivers don't help me run Blender or play games under *BSD, or do much of anything at all on a non-i386 architecture, anywhere, so I'm not going to bother with them. Frankly, there's something to be said for the hardline, "we won't take it unless it's Free" approach that, e.g., the OpenBSD guys have taken.

Re:Maybe?

[toadlife]

"the core value of Linux"

I thought the core value of linux was to have a free, stable, fast, powerfull UNIX-like OS?

Perhaps you confusing 'the core value of linux' with 'the core value of Richard Stallman and his army of anti-capitalist minions'?

Go ahead and mod me a troll. I've got Karma to burn.

Re:Maybe?

[nathanh]

I thought the core value of linux was to have a free, stable, fast, powerfull UNIX-like OS?

A single non-free binary driver means the system as a whole is not free. Haven't you ever wondered why many Linux distros won't bundle the nvidia driver, instead requiring the users to download it as a separate step? It's because the nvidia driver isn't free.

Re:Maybe?

[toadlife]

Well I wouldn't know...I'm a BSD user and to be honest I havn't used linux in over two years. Are there any linux distros that distribute Nvidia's binary driver?

Assuming NVidia allows redistribution of their drivers, the entire OS could still at least be "free as in beer", and "free" software advocates such as yourself, could chose to use the OS driver instead.

Re:Maybe?

[Anonymous Coward]

This is a good point that gets brought up a lot but i don't think either side really understands. You see it all comes down to why you have and use a computer. If your computer is an expression of your beliefs (as yours appears to be) then you should not compromise those beliefs for anything and you should support the card that has the open driver support. However if your computer is used as a tool to get things done (as it appears the parent who uses nvidia drivers because they work) then you should use what gets the job done best.

The main thing to remember is to have an open mind and accept that people use their computers for different reasons than you and thus place their priorities in different areas. That's not a bad thing (even if it means they use Microsoft products).

Re:Maybe?

[Anonymous Coward]

But in order to make use of the product you need to use the binary driver as using the open source driver is not an option due to the fact that it is lacking the feature that I need to use 3d accelerated graphics. If I did not need this feature I would get some crappy old PCI video card.

If you need to get something done which requires 3D accel you do not really have an option to bitch simply because you will not get the task done. Thus there really is no choice on the user side.

I think it would be better for everyone if NVIDIA just provided COMPLETE specs for their videocard regarding interfacing and some kernel maintainers would write open source driver from those specs.

Regarding someone's comment above that giving out the specs would give a leg up for the competitors this is utterly false. This is not something like smart cards where a competitor reverse engineered the technology and leaked it so people could pirate it or wireless chipset where you have access to VCO (voltage controlled oscillator) to change the frequency outside the FCC accepted range. No, here we only need interfacing. I can think of Java APIs as a good example, think OOP, you do not know the implementation, however, you know how to interface the device. And that is all that is needed. I do not see why NVIDIA doesn't want to save themselves their engineers time as well as increase their customer base by hiding by excuses such as it gives competitors a chance to catch up. Even if that were the case, NVIDIA would already have released a card and the competitor would only start developing and testing chips and then complete product. Rellevant example I can think of is when USSR stole 8088 [or was it 8080]. By the time they finished reverse engineering, building and testing it to make it ready for production, Intel made 80386 chips.

When viewed from perspective of an end user, I have things to do and I have deadlines; of course, there's room for optimism.

Best regards,

Oleg M

Re:Maybe?

[Mornelithe]

The open source nvidia drivers are only good if you want a $600 video card to perform like a $10 video card. The ati drivers are better than that on old cards, but not much on the new ones.

If you want to play 3D games on Linux today, you need to use binary drivers. Another alternative is to use Windows for gaming, and Linux only for desktop applications. In that case, nvidia still has no incentive to release any specifications or open source drivers for Linux.

A third alternative is to forgo playing any 3D games at all, in which case ATI/nVidia lose perhaps 1% of their sales, and the user has to be unhappy. And ATI/nvidia still have no incentive to release open source drivers/specifications, so these people will be unhappy forever (or at least until way in the future when Linux has a large enough market share that a successful boycott can actually be established).

So, essentially, your position is that you should make significant sacrifices for no conceivable gain, whereas the people you're arguing against are suggesting that people reap the benefits of an almost completely open-source desktop in addition to being able to enjoy their games, by making a concession on a point that they have no hope of winning.

Does that sum things up pretty well?

Re:Maybe?

[nathanh]

If you want to play 3D games on Linux today, you need to use binary drivers.

That is incorrect. There are 100% open source drivers for several 3D video cards. See DRI.

So, essentially, your position is that you should make significant sacrifices for no conceivable gain,

No, my position is that the long-term success of Linux depends on the distribution remaining open source. Your desire to play 3D video games using $600 video cards is relatively irrelevant.

Re:Maybe?

[Anonymous Coward]

You don't need to release a new driver with every kernel release. ATI just has no idea what they're doing that's all.You don't need to release a new driver with every kernel release. ATI just has no idea what they're doing that's all.

Oh, I thought it had something to do with Linux not having a stable driver interface for reasons mostly philosophical and partly technical. I guess Linus and the kernel developers were lying to me all this time. Good to know that Slashdot Poster #756019 knows better than the people who write the code, though! Good job!

But to answer your question, the reason why most manufacturers don't put a high priority on Linux support is because the amount of effort it takes is simply not worth the trouble. Linux just doesn't have the install base.

Yup, and this problem is not magnified at all by the lack of a stable driver interface. Manufacturers don't have to recompile and test and quality control drivers for every new minor release, every new distro release with patches, and so on - all this crap they have to do dozens and dozens of times a year, they actually don't have to. Why? Because Slashdot Poster #756019 says so, that's why! And wouldn't we all prefer to live in his fantasy world?

Quit your bitching, stop all the silly petitions to try to get ATI to release decent drivers. All that does is let them know they can do it and people will still buy them!

Yup, because companies really tend to listen to people who aren't just not their customers, but actively and purposely avoid their products. Better yet, let's hope they go out of business - then they'll have to give us better driver support!

You, my naive friend, are a fucking moron. There are lots of reasons for poor driver support on Linux. But the lack of a stable driver interface is probably #1.

That's a separate issue from whether Linux should or should not present stable driver APIs - maybe the hard-line approach will win out in the end, or maybe the GPL demands such an approach anyway (this is subject to interpretation). But to ignore this huge, incredibly important factor is the height of stupidity. And for Slashdot to reward you with a +3 insightful moderation for this ignorance is... well, sadly, it's par for the course.

I strongly suggest that you have the slightest fucking idea what you're talking about before opening your mouth next time. But hey, take that with a grain of salt. Apparently lots of folks around here are more than happy to reward ignorance and intellectual dishonesty.

Re:Maybe?

[Mishura]

From the r300 website:

The source code on this website may damage your hardware.
It is *UNTESTED* and *BROKEN* !

Okay, I'm all for free drivers, don't get me wrong. But I am not about to use drivers that may damage my expensive piece of hardware, just for the sake of "Freedom". After all, I am not a programmer, and source code is alien to me.

I am primarily a Linux user, and I love to play 3D video games. I want them to "just work". This is where nVidia comes to play. They make drivers that work, and are easy to install (In my opinion, milage varies).

So, in this case, nVidia gets my money, not ATi. Let me know when there is a video card on the market with free drivers that are fast and stable, and supports common OpenGL so that all my games continue to work, then I'll switch. Until then, I will continue to recommend nVidia to anyone who asks about graphics cards.

P.S.: Not a flame, I am just not an idealist. I wish the r300 project the best of luck. We do need free drivers, after all.

-- Mishura

Re:Maybe?

[Bloater]

> It's zealots like you that keep a lot of people from adopting Linux, and give the whole community a bad name.

It's zealots like him that made Linux worth having (Richard Stallman and the Free Software Foundation).

Without them, the BSDs would also not be worth having.

> Nvidia is doing everyone a favor by releasing quality drivers.

Also nVidia's drivers are not quality, they don't work with Ingo Molnar's realtime patches and they don't work with the less used kernel options for several months after after a kernel comes out. If they were Open Source, people would be able to something about the quality.

Re:Maybe?

[JRIsidore]

Free drivers are surely a good thing and I generally prefer them, too. But on the other hand after paying several hundred euros for a new graphics card and then not beeing able to do with it what it was bought for in the first place is ridiculous. If buy a mid-range to high-end card and actually want to use it (i.e. do more than displaying my desktop) there's sadly not much choice with regard to drivers.

Re:Maybe?

[cowbutt]

It's a shame, then, that nVidia seem to have decided that geforce2go-based notebooks running kernels newer than 2.6.10 or newer are not worth maintaining their drivers for.

Rely on binary drivers, and sooner or later you get screwed.

Re:Maybe?

[Anonymous Coward]

Well, xscreensaver crashes my desktop and my laptop when using the open source rage 128 and radeon driver. According to your theory, the bugs should be fixed by now. Unfortunately they haven't. The rage 128 did actually work reliably in Red Hat 7.3 or 9. The driver actually appears to have become more buggy, not less. They have been broken for quite a while and nobody seems to care. While I haven't tried any closed source video drivers, Nvidia's drivers have a reputation for stability. If nvidia's drivers are occasionally break when newer kernels are released that is much better than the open source radeon driver that has been broken for years.

Re:Maybe?

[Anonymous Coward]

Unfortunately, lots of modern games don't run properly on a radeon 9200. Neverwinter Nights runs about 13fps on my radeon 9000 under windows. I haven't tried it with the open source drivers under linux, but I'm guessing it would run at 10fps or less. It just wouldn't be playable.

Re:Maybe?

[glitch23]

ATI is the entity making the dependency on the kernel version, not the kernel developers making the dependency on the ATI driver version. If they are not making Linux drivers a high priority it is their own doing.

Re:Maybe?

[Anonymous Coward]

Shhhhhhhhh. You can't talk about these sorts of realities. Get back to the real non-issues man! Open source!

Re:Maybe?

[m50d]

How about get one of the 3dfx cards with good open drivers? It's getting to the stage that they're not good enough for the latest things, alas, but the opengl is certainly good enough for genuine 3d gaming.

Re:Maybe?

[kubrick]

The Nvidia drivers crash for me within 30 minutes using NVAGP, or 5 minutes using AGPGART. This is using an NForce2-based chipset and a 6600, so it's not like they can't replicate the setup.

People should be careful what they buy; if it's not open, there are no guarantees that it will work or that it will continue to work in the future.

Re:Maybe?

[cfsmp3]

If you're too short-sighted to understand that then perhaps Linux is not for you.
Ok, so now linux is not a piece of software people use if it suits their needs, but rather it's something they are allowed to use if if they share YOUR ideas.

Re:Maybe?

[Curtman]

It may not be fully stable yet (although is has worked smoothly for me for about a week)

Which card have you got? That project looks like a lot of fun.

I have filled this goddamn form out a dozen or so times, phoned, and even sent snail mail asking for their hardware docs, each time kissing ass and telling them how wonderful they are. I'm listed as a developer for Gatos, and they still won't talk to me. I think the "SCO Linux" option on their target platform list is a good enough indication of how much attention that section of the web page gets. WTF does that even mean, SCO Linux.. RedHat, SuSE, and Debian aren't on there, is that some kind of jab?

I think its highly doubtful I'll ever see any docs, so its probably time I just get a card and start hacking.

Re:Maybe?

[Esine]

Look, not every single person who uses Linux, uses it only because it's Open Source, but because it's a good operating system. I like Linux, I like GNU tools, I like UNIX. I want my system to function fast and the way I want it to, and that's exactly what Linux, nVidia (binary) drivers, XFCE 4 and Opera gives me.

Re:Maybe?

[ccp]

I've posted before, so accept my virtual moderation:

+1 Well said!

Cheers,

Carlos Cesar

Re:Maybe?

[Curtman]

Please tell me how having a closed source driver destroys the core value of linux. But, in doing so, you must explain how not having this driver at all (which is the only other alternative in todays society) would be a better option.

Because it takes away incentive for people to get involved in the free projects. Not only that, but the people who do get involved are constantly bombarded with fanboys telling them how their work is useless because the proprietary drivers are superior. Of course they are, they're the only ones who know how the damn cards work. In the past we saw great contributions from 3rd parties like when The Weather Channel paid out of their own pocket to port the Radeon driver to the 8500 cards. This kind of thing is much less likely today.

Thats enough of the Z word too. The community is zealous. Thats a big part of what makes it so much fun. There's nothing wrong with being excited about free software at all. Don't tell people they should be ashamed of it.

Re:Maybe?

[drsquare]

In reply to "Open source isn't the be-all end all.":

>With Linux, yes it is. If you don't care about open-source then use something else.

What? I don't recall Linus Torvalds (the creator of Linux), saying that the whole point of Linux is to run entirely open source software. Perhaps you can find a quote, either from Torvalds or from the kernal copyright licence that says you're not supposed to use any non-open-source software on Linux.

The way I see it, Linux was created to be a decent Unix-alternative, the fact that it's open source is a bonus. Who are you to tell people what operating system to use because they don't follow your ideology. I could equally say that if you don't want binary-only Nvidia drivers then use another graphics card.

Idealism means seeing the bigger picture and foregoing fleeting fancies in the pursuit of long-term success.

What bigger picture is this? Are you saying that using your hardware is a 'fleeting fancy'? Where is the long-term success in having your hardware non-operative? Perhaps you think that Nvidia will open the source of their drivers if enough open-source zealots use 640x480 16 colours because they won't install the closed-source drivers.

If you're too short-sighted to understand that then perhaps Linux is not for you.

Repeat after me: (This also goes for the zealots who modded your hilariously close-minded rant to +4)

LINUX IS AN OPERATING SYSTEM, NOT A RELIGION. IT'S A MEANS TO AN END THAT MAKES YOUR WEB BROWSER AND EMAIL WORK, THE WORLD DOES NOT REVOLVE AROUND IT.

Re:Maybe?

[Anonymous Coward]

This is precisely why I choose Nvidia hardware - and I always will. Nvidia took the time to ensure that I can have accelerated graphics on my choice of OS, so I will reward them with my pocketbook. You should do the same.

ATI has probably realized that the extra $1000 they'd get per year from the 4 or 5 people who would buy a new game accelerator for a linux box isn't worth the development effort. That's the market for ya.

Re:Maybe?

[Anonymous Coward]

Interesting, I've worked on and released some pretty popular OS stuff for Linux. But I don't think OSS is the be-all end all, so would the world be better off without my contributions?

Linux is a tool, nothing more, nothing less. Don't try to make it into a device for some political agenda. That approach hasn't and never will work for Linux.

Re:Maybe?

[sgtrock]

The core value of Linux is that it is open source.

Agreed. However, I don't think things are as black and white as far as Nvidia is concerned. What you may not be aware of is that Nvidia uses technology from other companies. Those companies have chosen to keep their technology closed source. Some of it is patented, some copyrighted. In either case, as long as Nvidia wishes to use that technology, they are legally and morally bound to honor their wishes.

The copyright issues are pretty trivial to work around, and I gather from what I've read in various places that Nvidia has worked to minimize the impact of that material. The patented technology, however, presents them with a much larger headache. There really is no safe way to code around known patent issues. You can try, and face the probability of a $20M lawsuit, or fork over some licensing fee and move on. Nvidia chose the latter solution.

Do we have fast, completely unencumbered, open source 2D drivers for Nvidia's cards? Absolutely!

Do I want to see completely unencumbered, fast, open source 3D drivers? Absolutely! Can I program them? Absolutely not! Can Nvidia or anyone else? Apparently not! At least, not for anything resembling a reasonable cost. :(

Re:Maybe?

[Skjellifetti]

There is a practical reason for not purchasing closed source hardware. Sooner or later, the manufacturer will quit supporting the driver as they move to newer generations of their product. Not all of us care to own the latest and greatest hardware. We just want our stuff to work and keep on working long after the manufacturer has declared our system obsolete.

Don't be such an idealistic ass. It's people like YOU that destroy the core value of Linux.

Its better than just being a plain ass. Its people like YOU that destroy the core value of respect for other people.

Re:Maybe?

[Mornelithe]

That is incorrect. There are 100% open source drivers for several 3D video cards. See DRI.

Yes, and for cards that can actually play relatively new games, there are no acceptable open source drivers. That solution is growing increasingly unacceptable. Your closest shot might be a Matrox Parhelia (I believe Matrox has 100% open drivers), but that had trouble shaping up to compete within its own generation of cards, let alone two generations later.

No, my position is that the long-term success of Linux depends on the distribution remaining open source.

Which distribution? Just one of them?

What if all the distributions are 100% open source, and people are able to download and install binary nVidia drivers separately? Does that totally fuck up Linux in every which way?

Or are you arguing that the existence of a binary driver will magically make Linux proprietary; that people will suddenly stop writing open source drivers because some proprietary drivers are available? Why then are people still writing open source drivers for nforce components, then, when nvidia offers binary drivers? Just because some people can be contented with a proprietary driver doesn't mean that others won't be motivated to write open source drivers, whether through idealism or because, wonder of wonders, they enjoy doing it.

Your desire to play 3D video games using $600 video cards is relatively irrelevant.

It's irrelevant to you, but it's not irrelevant to a whole bunch of people out there. In fact, the number of people who care about that dwarfs the number who hold the view that if everything isn't open source, the world is going to end.

Being open source is a big boost to Linux, and it would most likely be a big boost to the nvidia and ati drivers. However, that isn't going to happen today, no matter what you do. In the mean time, it's probably much more important for more people to adopt Linux, as that actually spreads open source to more people, freeing them from proprietary software for the most part. When the culture becomes more ubiquitous, it's more likely that other things, like the ati and nvidia drivers, will become open source themselves. In fact, I'd say it's much more likely to happen that way than it is for a thoroughly insignificant portion of the population to try to force the issue.

You've also as yet failed to explain exactly how the use of a binary driver today will destroy the future of Linux. Repeating it over and over isn't a valid argument.

Incidentally, your view that (paraphrasing) 'if you're not going to use exclusively open source, you shouldn't use Linux,' is flawed, and many kernel developers would probably disagree with you. Linus himself is oft noted as a pragmatist; he merely saw the GPL/open source as being the best way to run his project. People work on Linux for a variety of reasons, and Free Software isn't the end-all and be-all of them

If you're concerned with Free Software above all else, you should be running the HURD, not Linux. That's more the mission statement of the former.

Re:Maybe?

[kelnos]

Agreed. However, I don't think things are as black and white as far as Nvidia is concerned. What you may not be aware of is that Nvidia uses technology from other companies. Those companies have chosen to keep their technology closed source. Some of it is patented, some copyrighted. In either case, as long as Nvidia wishes to use that technology, they are legally and morally bound to honor their wishes.

Legally, sure. Morally? That's a bit subjective.

The copyright issues are pretty trivial to work around, and I gather from what I've read in various places that Nvidia has worked to minimize the impact of that material. The patented technology, however, presents them with a much larger headache. There really is no safe way to code around known patent issues. You can try, and face the probability of a $20M lawsuit, or fork over some licensing fee and move on. Nvidia chose the latter solution.

Actually, you have it backwards. If the closed-source driver contains code copyrighted by a third party, nvidia cannot release that code without the third party's consent. Patents in this case are irrelevant. In fact, the *point* of a patent is that it's public. A patent -- a temporary monopoly on an invention -- is granted in return for the materials in that patent being made public knowledge. Now, if nvidia's patent license with a third party states that nvidia cannot implement the patented tech in open source code, that's a problem. It's just as easy (or hard) to "work around" that situation as it is a copyright problem.

Anyway, you're still missing the original poster's point. The point is, the core value of Linux is that it's open source. Having closed-source drivers for it defeats that core value. Would it be better if we didn't have a closed-source driver for nvidia? Perhaps not immediately, but the lack of that driver (perhaps because Linux's license disallows binary-only drivers), coupled with consumer demand, might eventually pressure nvidia to create a comparable open source driver.

Having said all that, I have two nvidia boards in two PCs. They both run Linux, and I have the binary driver running on both. When I think about it though, I don't need it. I don't do anything that requires 3D acceleration, and what 2D acceleration that is present in the open source driver should be enough for my needs. And if not... well, it's not all that big a deal. It's just a video card; I can live without the fancy features.

Re:Maybe?

[runderwo]

No. Matrox does not provide open source drivers for any of their recent cards, nor specifications for any cards anymore. They have completely abandoned supporting third-party open source development. Furthermore, they have placed a license on their previously free software G-series drivers which is no longer usable in any free software project.

Re:Maybe?

[Anonymous Coward]

Without a proper nvidia driver, Linux would be basically useless for any real modern desktop use, as all we'd have is driver support comparable to say that of any of the other open sourced ones, SiS, via, intel, matrox, the open source nvidia, or ATI drivers. Which by the way all tremendously suck. Sure basic 2d operations are supported, but that's it.

I've got a (semi-old) Matrox card. It can't put 3d triangles on the screen as fast as your NVidia card, but it supports high resolutions, high color depths, two displays, and accelerated 2d and 3d.

Despite having what is (these days) a "slow" graphics card, nothing I do would go faster with a new graphics card. I suppose if the fully-3d-accelerated X server ever becomes ubiquitous, I'd be able to move windows around a bit faster with a newer graphics card, but that's about it.

The only reason I can think to call a Matrox card "basically useless for any real modern desktop use" is if you think the average desktop user spends most of his day playing Doom3.

Don't be such an idealistic ass. It's people like YOU that destroy the core value of Linux.

Ah, there it is. The true "core value" of Linux: calling people asses on internet message boards!

Poor Linus

[RickPartin]

Sorry about being off-topic but I've been thinking, since Linus is a normal guy and not some super human CEO, he must go through a "family tech support guy" hell that only exists in only our darkest of nightmares. I pity him.

Re:Poor Linus

[JustOK]

Not an issue since he switched to a Mac

Re:Poor Linus

[aCapitalist]

At least you didn't compare him to Nick Burns.

Re:Poor Linus

[Anonymous Coward]

No way, dude! I heard that guy runs Linux.

Re:Poor Linus

[Feztaa]

Nah, he just tells them to install linux. Then they laugh at him like some kind of kook, and don't bother him anymore. I mean -- who uses linux? Puhleeease!!

Re:Poor Linus

[Anonymous Coward]

No way man.

Like the truly wise man he appears to be, he chose the only option in such situations: Move several thousand kilometres away from home :-)

borked

[Sparr0]

Yet another kernel release without fixed/rolledback highpoint RAID drivers :( Kernel Oopses and Panics abound and they insist on keeping the broken code merged in around 2.6.9. Well, there's always hope for 2.6.13!

Re:borked

[Anonymous Coward]

They should remove support for highpoint RAID drivers altogether, to protect people from making the mistake of using those pieces of crap.

Re:borked

[Anonymous Coward]

yeah, suckers like you that keep coming back for more... why SHOULD they bother to fix it?

Re:borked

[Sparr0]

news flash, they make perfectly adequate IDE controllers if your ignore their shitty "hardware" (lol, can you say winmodem?) RAID offerings. but even that doesnt work in linux right now.

Re:borked

[toadlife]

May I suggest the 3ware RAID series. They handle the RAID array in hardware, which probably makes the driver less complex and less prone to breakage (although we are talking about *linux* here, so you never know) between kernel versions.

I bought one of those babies for my FreeBSD box. I love it.

Some explaination in the changelog...

[Whyte]

commit 1da177e4c3f41524e886b7f1b8a0c1fc7321cac2
Author: Linus Torvalds
Date: Sat Apr 16 15:20:36 2005 -0700

Linux-2.6.12-rc2

Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.

Let it rip!

Re:Some explaination in the changelog...

[NitsujTPU]

Good explanation.

Re:Some explaination in the changelog...

[Anonymous Coward]

Good complement.

Re:Some explaination in the changelog...

[Anonymous Coward]

Good misspelling.

Switch to Linux

[Anonymous Coward]

We were negotiating with the Pentagon.
We had a blue screen of death.
That was the last straw.
When you're holding the moon for ransom, you value stability in an application.
Linux gives us the power we need to crush those who oppose us.
It's compatible with our orbiting brain lasers.
I've got a beowolf cluster of atomic supermen.
I have more friends now.
Genetically engineered cybergoats.
Henchmen with bad teeth.
Georgous fembots with a penchant for evil.
I mean Linux runs on anything.
I'm all about open source.
It's just changed my love life.
You have to uh.. config it.
Uh.. and then you have to write some shell scripts.
Update your RPMs.
You have to partition your drives... and patch your kernel.
Compile your binaries.
Check your version dependencies... probably do that once or twice.
It's just so easy and so simple, I don't see why most people don't run Linux.
Thank god they don't, because they'd all be super villans, wouldn't they?
Huh uh ha!
I'm Steve, and I'm a super villian.

Re:Switch to Linux

[einhverfr]

You have to uh.. config it.
Uh.. and then you have to write some shell scripts.
Update your RPMs.
You have to partition your drives... and patch your kernel.
Compile your binaries.
Check your version dependencies... probably do that once or twice.
It's just so easy and so simple, I don't see why most people don't run Linux.
Thank god they don't, because they'd all be super villans, wouldn't they?
Huh uh ha!
I'm Steve, and I'm a super villian.

So, Ballmer, What's up?

Re:Switch to Linux

[almostinsane]

Too bad all of that is true. Linux could be usuable if you could just click on a setup file and be done with it. what a piece of....

Re:Switch to Linux

[Anonymous Coward]

Actually, its not, unless you want it to be.

Go get Ubuntu... or one of the other very user-friendly distros. Most of them are easier to install, upgrade, and manage than any windows version in existence.

Re:Switch to Linux

[einhverfr]

I think I had to click the mouse four-five times to install Fedora.... Oh, and enter a root password. But that was all....

I am all for shell scripting and manual setup (I once built my own live CD from scratch as a way for studying for the LPIC-2) but these are optional these days....

Re:Switch to Linux

[Ziviyr]

Config... Choosing what programs you want? Free programs? Terrible...

Write shell scripts? Must be some stupid RedHat thing.

Update RPMs? RED-HAT...

Partition drives? I thought we already installed, only a true supervillan would partition after installing!

Patching the kernel ... I guess if you're running on cybergoats and fembots you'll need special drivers... Ideally you'd have modules figured out if you're doing a big rollout though.

Compiling, an option in Gentoo, otherwise download and go!

Version dependancies, thoes things that automatically whiz by when I choose what free software I want, so it all works?

... This guy is so evil, he probably stabs himself in the hand if he has nothing better to do...

Re:Switch to Linux

[oliverthered]

I have a home wifi network and a pan bluetooth network at home. Sure setting them up an installing with routing tables ment changing a couple of lines on a file that I took from a howto, but in that was a hell of a lot easier than setting them up under windows.

I think most people can work through a howto, it just a shame that windows doesn't come with a couple of hundred megs of howtos.

Somewhat OT but...

[rsilvergun]

anyone else having trouble with saa7134 based tuners (I've got a compro videomate tv/fm). My image is way off center (towards the top right). I'm currently us Dscaler in Windows because of it (great program, but Windows... ug). The weird thing is, dvds played from my PS2 are centered, but games are way off center (and some like Street Fighter Alpha 3 won't play in tvtime but will in xawtv, go figure). No luck on the v4l list yet (noone there's seen the the problem), so I'm wondering if I should try patching up my kernel. tia.

ah, the apologist is touchy, too

[Anonymous Coward]

Did I hurt your widdle feewings?

No, you didn't say "lacking in organizational skills". You said something which means the exact same thing. Are you desnse?

I like how you added "asshat" afterward. That'll show me!

PS: Eat a dick.

Re:ah, the apologist is touchy, too

[yahwotqa]

I liked how you added "Eat a dick" afterward. That'll show him!

2.6.13

[jd]

There was a discussion on LWN about what was going to happen in the 2.6.13 and 2.6.14 timeframe. Apparently, there is speculation Linus may merge in a lot of the more stable stuff from Andrew Morton's -mm patchset. If the updated RAID drivers are in that patchset, there is a good chance they will be in 2.6.13 or 2.6.14.

In the meantime, there are a lot of valuable, interesting and worthwhile projects that aren't in ANY of the patchsets at this point in time. I e-mailed a few of the maintainers about that, and it appears that they're aware of the problem but want general users to pressure the patch maintainers to publish patches on the kernel mailing list AND that said patches should conform to the kernel programming style.

So, again, if you want updated drivers for RAID, or additional features you know damn well exist and are out there, lobby the maintainers until they publish the stuff in a way the core kernel maintainers like.

There is simply far too much good stuff out there that is not being seen and not being used. It has got to the point where I will be reviving my own FOLK patch series, to start documenting the patches that live out on the fringes of kernelspace. If we want a better Linux, all we have to do is ask in a way that will be heard.

Re:2.6.13

[Anonymous Coward]

guess i don't feel so insane for ditching windows for os x. heh, i messed with this linux crap for a few years on the side, but you know what, it is still a joke, hate to say it.

Re:2.6.13

[Anonymous Coward]

Jesus - what a joke. It sounds like a full on mess. Custom CMS system written in 30 days. Patches being submitted that are ignored because the users don't use the proper kernel coding style. Seperate patches maintained outside of the kernel group that never get integrated in.

What a mess.

Re:2.6.13

[jd]

I hate to say it, but there's a lot to what you're saying, which is exactly why I do try to work on building patch collections. The more people know what is out there, and the more those patches are in actual (as opposed to theoretical) use, the more likely it is that developers will work to sort this out.

For as long as people believe that what they see is all there is, there will be no pressure on anyone to change what they see.

Re:2.6.13

[Nasarius]

Patches being submitted that are ignored because the users don't use the proper kernel coding style.

What's wrong with that? Do you want it to become an even bigger mess?

Re:2.6.13

[stor]

You're not familiar with the Tao of Linux ;)

It sounds like a worse situation than it is: it can depend on your perspective. There are often patches that float around and never get integrated but usually there's a good reason for that (code quality sucks, minimal testing, breaks other stuff, not conforming to kernel coding style).

Speaking of Kernel Coding Style, are you claiming that this is unimportant? Do you understand the concept that once a piece of code makes it into Linux then it's supposed to be maintainable by a group of people? You realise that the original maintainer of a piece of code is free to walk away at any time, so if they've used their own funky style the code will be obfuscated and unmaintainable?

Have a look at the recent discussion wrt inotify! It's an important feature for the desktop IMNSHO but core maintainers are sticking their noses up at it. Is this a bad thing? If you are more interested in the functionality than you are in the opinion that it's "Just Not Right Yet(tm)" then maybe. I'm really keen to see it in but I'm sincerely appreciative that Christopher Hellwig criticises and vetos the work until it conforms to his idea of aesthetics and sanity. CH can be a hard-liner but people like him, Andrew Morton and Al Viro raise the bar and push for quality. This is inarguably A Good Thing.

Anyway, mistakes happen, patches get ignored, people crack the sads, life goes on. Problems occur in commercial software development too. It's not time to ring the alarm bells.

Cheers
Stor

Re:2.6.13

[Tim C]

I agree with you on the CMS thing, and at least partly on the "seperate patches" thing, but not the insistence on obeying the coding guidelines. Any code that is going to be maintained by more than just the author needs to have a coding style. It doesn't really matter what the style is, as long as there is one and it's properly enforced.

Trust me on this - having different parts of a codebase in different styles makes maintenance a lot harder, and increases the risk of introducing subtle, hard to spot bugs. It also makes it look like arse.

human readable summary

[digitalderbs]

Changelogs are great, but does someone have a link to a list of major changes (short point list summary) from 2.6.11? I read the kerneltrap blurb, and that didn't satisfy me. thanks

Re:human readable summary

[BinLadenMyHero]

I would ask just that.
That should be present on every major release.

Re:human readable summary

[smash]

It is, and has been since the mid 90s at least.

If you're bothering to compile your own kernels, etc - you should be on the kernel digest mailing list.

smash

REAL Men..

[Anonymous Coward]

...use the odd numbered revisions...

More about 2.6.12's TPM support

[UnapprovedThought]

There was no mention of TPM in the summary. Only a line about an i945G driver.

However, sourceforge lists a new TPM device driver at http://sourceforge.net/projects/tpmdd. It is a set of patches which add TPM support to pre-2.6.12 kernels.

Also, a TPM-tools package using the trousers library is at http://sourceforge.net/projects/trousers .

An article in heise (translated link) titled "Linux kernel 2,6,12 with TPM support."

Linus is our Family Tech Support Guy!

[Anonymous Coward]

he must go through a "family tech support guy" hell that only exists in only our darkest of nightmares

It seems today
that all we see
is Longhorn delayed
and OS X on PeeCees
but where's the free and open source
on which we used to rely?
Luckily there's our Family Tech Support Guy,
the guy who makes the kernel
that runs on all the hardware
we bought at Fry's.
He's
our
Family
Tech
Support
Guy!

Hmm. Sorry. I got carried away :).
Thanks Linus for all your hard work!

You're Fired

[Doc Ruby]

What kind of ridiculous release management is that? Why doesn't the release manager "puzzle it out" once, so the rest of us can look at it, to tell whether it's worth bothering with? Their job is to ensure the release is packaged, not just brag to their friends that they're the glorious release manager.

Re:You're Fired

[slavemowgli]

There is no release manager. A new kernel gets published when Linus decides it's time; in a way, that makes him the release manager, but it's not really managing as in "creating schedules, specifications, requirements, deadlines and all that". And I at least would rather see him do actual work instead of meeting arbitrary requirements imposed on him by the more marketing-oriented types.

That being said, Linus *has* given a reason why there's no full changelog this *one* time (it's reproduced right above in this very Slashdot discussion, for example); if anyone has issues with that, I assume they're more than welcome to create a full one and post that. If noone does... well, then the itch probably wasn't worth scratching after all.

So there. If it really matters to you, then go and create a full changelog. If it's not worth your time and effort, why do you complain that Linus feels the same way?

Breakage warning

[Anonymous Coward]

Unfortunately this one isn't so good and breaks vmware again, and strangly postfix.

I will be ready for dual kernel machine with my

[LM741N]

Gentoo Sources. Then I can do an eval. I will try to post results to main Gentoo list, so everyone can see it there. I assume that because this doesn't even have a changlog, that only minor changes have been made, although I may be dead wrong. Please correct me if this and true and I will sit back and wait, wait, wait

Sincerfely, Rob

CPU-FREQ changes

[glMatrixMode]

Skimming through the changelogs (link in story), I found many interesting CPUFreq changes, like :

* New governor 'Conservative' based on 'ondemand', except that it increases cpu freq step-by-step, instead of switching directly to the highest freq. This should improve battery time and address latency problems on amd64 systems.
* Improved support for PPC32 and ARM
* Support for dual-core opterons

Actually GNU HURD...

[rekrutacja]

...is fully operational death s^[[3~^[[3~^[[3~ partly working. You can even run GNOME on the top of it:
http://lists.debian.org/debian-hurd/2005/05/msg001 09.html

...GNU/HURD and KDE

[Anonymous Coward]

http://lists.debian.org/debian-hurd/2005/06/msg000 56.html

One Change I Like

[Borealid]

I don't know if anybody cares, but this update supposedly fixes usb-audio so that disconnecting a running sound card won't eliminate your keyboard. Those of you with SB Audigy 2 NX or Extigy cards should probably upgrade.

.. and here's the original animation

[nacs]

The text above is taken from this flash toon for those interested.

Re:.. and here's the original animation

[dabigpaybackski]

Phwew. I thought it was some random "word salad" from a pschizophrenic Slashdotter.

It is not that simple

[Anonymous Coward]

Things aren't as simple as just "according to specs". The kernel currently does not keep a standardized API that binary modules can easilly use. Therefore you need to recompile your driver for the new kernel, even though you do not actually have to change the source code.

This is the very problem with Linux - and they do not want to change this. Instead they want manufacturers to open their code. Of course ATi and nVidia will never do this and we are stuck with having to wait for them to release new drivers.

nVidia has partially solved the problem by providing a wrapper part that you can recompile for your specific kernel. It is a work-around and not a good solution.

Re:It is not that simple

[pAnkRat]

"It is a work-around and not a good solution."

yes, but it works very well.

For linux (opensource) development, it is completly bogus to be binary compatible, just do a recompile when needed.

Hardware vendors don't want to open up their drivers.

So you just meet in the the middle, the hardware vendor build an BAPI stable driver (that's their problem) and an interface to it.
You recompile the intrerface, problem solved.

This is a very good real world example for the "interface pattern".

I think there will be no alternative.

Re:It is not that simple

[Anonymous Coward]

The wrapper is designed to wrap a binary driver to the kernel. The problem is that the biary driver is not always compatible with the kernel.

The wrapper is not a "work-around" to anything other than the fact that they want to glue a binary-only driver to every potential kernel out there. It's not that it's a bad solution. It's the *only* solution for binary-only drivers. Every videocard manufacturer that has ever made binary drivers has done it this way on Linux. nVidia, ATI, PowerVR, Matrox. They all do it this way.

ChangeLog this!

[Anonymous Coward]

Alright, enough bitaching about Linus' changelog that leaves you wanting... I've forked the kernel proper (see, they always said this would never happen but it did!!) to include the full history.

I shall call it the "Jinux Kernel".

Xen

[nighty5]

Doesn't look like it made it into this release :-(

Re:Xen

[Anonymous Coward]

Yeah, the changelog makes several references to code changes that will support Xen, so hopefully it will arrive in .13

Apparently there were a lot of organizational changes that had to be made to the Xen source before it could be allowed in the main kernel tree. I'd rather it be right and not rushed.

Re:Xen

[Anonymous Coward]

I know I was hoping for it. I can't wait until Xen matures into something like ESX Server. We need Xen! :)

Re:Xen

[Monkey]

If you're ignorant like me, and don't know what Xen is... its VM software that allows you to run multiple operating systems concurrently with close to native performance.

The fix works 100%.

[Anonymous Coward]

..which is a big nice help for me. Thanks, Linux and others!

See bug 4495, which I suppose can now be closed.

Yawn so interresting

[vrioux]

Oh no! Not a new Kernel? Now I need to recompile it on all my servers! Oh wait, I figure I don't need because I use BSD only. When is Linux (and BSD too) going to start working on simplifying it's upgrade process? What about rules in it's development preventing anyone from making changes that would hinder backward and forward compatibility? When was the last time you had to update your Windows Kernel? Did it break something major? I had too much problems with new Kernels having issues with drivers and configuration. All that development process is disproving itself after each release.

Re:Yawn so interresting

[Gogo0]

98 to 2k?
2k to XP?

perhaps nothing broke for you, but for many others, a lot of things broke. hell, even MS' own programs broke.
its as easy as a complete reinstall of your OS and all of your applications.

Re:Yawn so interresting

[someone300]

Well, it's quite simple as it is...

cd /usr/src
wget $url
tar -xvzf linux-$release
cd linux-$release
make menuconfig .. enable new stuff ..
make
make modules_install
make install # if your bootloader is configured to boot /boot/vmlinuz
cd ..
rm linux
ln -s linux linux-$release

then reboot.. or kexec (not sure how that's coming along)

There could be a fancy gui for it, and it can be managed in shell scripts, but it's an advanced user's task anyway, like manually replacing dlls on windows or something. Most users would not even know about it as FC or something's upgrade process will probably do it silently.

And why bother upgrading your kernel on your servers if you don't need a bugfix or a new feature?

Last update i had to do to the windows kernel was 2k to XP... the backwards compatibility can't have been hard for that, considering how little was changed, and anyway, it's very rare that something breaks on my system due to the kernel upgrading.. (except ATI drivers.. but I use the default X.org ones now)

I think the kernel devs are currently in the position where they're relying on the disto to monitor what packages break and stuff.

Re:Yawn so interresting

[einhverfr]

Win2k SP4 (and possibly early SP's) upgrage the kernel (ntoskern.exe, hal.dll, etc).

I know this because I had such an install go horribly wrong and give the ntoskern.exe not found error message.

So the kernel is upgraded a whole lot more frequently than one generally assumes in Windows-land.

Better the devil you know...

[Circlotron]

...than the one you don't know. If someone is telling something negative about themself then they are almost certainly telling the truth. That ought to give you more confidence in them in general, not less. Besides, in this particular case it gives you another parameter with which to make your decision on whether to use Linux in general. Can the same be said for commercial OS manufacturers? Would you rather get promised a "rich user experience" :puke: :barf: :rolleyes: and simply swallow the sales babble? If you "can't handle the truth", get out of the kitchen ;-)

Re:Better the devil you know...

[Anonymous Coward]

This truth does NOT instill confidence. In fact, it would have been better to post a slashdot message without them dirty stupid excuses - whatever are the reasons.

Actually, if the message said just one sentence - it would have been more positive than a message that sais "its out. but i'm super lazy today, so whatever it is you might have expected of me, do it yourself".

Its not about the developers, and not about the kernel itself --- its about the MESSAGE!!! the actual PR that linux is promoting on one of the most popular websites on the internet.

And good positive messages can make people love linux even more than a thousand kernel updates. Whoever these people are. Why not make them proud?

Yay!

[ASUSanator]

Im such a kernel whore i always upgrade as soon as a new kernel is released >_>

Reiser4

[azdruid]

Well, I'm a bit disappointed that native Reiser4 support wasn't included in this release. It's one of the features I'm greatly looking forward to...and I'm too noobish to compile a Reiser4 kernel module myself.

Re:Reiser4

[Anonymous Coward]

Reiser4 support is in the Andrew Morton "mm" kernel. Just grab the latest patchset and patch against 2.4.12 and your're good to go.

Re:Reiser4

[Pflipp]

Totally agree. Especially since they were apparently planning it for this release, so I was kind of looking forward to it.

Re:Reiser4

[Anonymous Coward]

My friend.

Emerge Nitro-Sources
Patched to your power hungary ricing satisfaction

or

Emerge mm-sources
The offical test bed for kernel imporvements

Re:Reiser4

[Creechur]

Well, I'm a bit disappointed that native Reiser4 support wasn't included in this release. It's one of the features I'm greatly looking forward to...and I'm too noobish to compile a Reiser4 kernel module myself.

It's not clearly advertised for some reason, but Hans makes available patches that add Reiser4 support to the mainline kernel. This should probably be spelled out more clearly somehwere, since I know alot of people avoid trying Reiser4 because they don't want to run -mm.

You left one out.

[Anonymous Coward]

Does the upgrade add something you want?

Re:You left one out.

[mr_z_beeblebrox]

Does the upgrade add something you want?

That is a good point but, I had a reason. Someone who can understand enough to compile a new kernel could figure out that if your current software does not solve your needs that would be an issue to address with new. ie...Does the upgrade add something you want? I made this list very simple because of the audience...they can add the complexity as you did. But, I concede to you that point probably should have been on :-)

Re:2.6.13 and coding style

[Anonymous Coward]

The difference being that in the past, folks would put a lot of effort into helping to reshape patches into the proper coding style and whatnot before folding them in. Somehow, volunteering a patch for the kernel has become a privilege rather than a service, and "coding style" has too often become a codeword for "we don't want to think about this"---when you can get a core developer to look at your patch at all.

IMNSHO not running concurrent experimental and stable trees any more is a horrible mistake. There ought to be a place for maybe-bad code to be integrated in with little friction while causing little damage and gaining experience. The kernel team is getting pretty full of themselves, and in the long run it's going to hurt Linux, I think.

GSM PCMCIA support

[wikinerd]

Linux 2.6.12 has a driver for GSM PCMCIA cards, which will enable GNU/Linux laptop owners to wirelessly surf on the Web via a cellular Internet connection, without the hassle of manually installing the hardware using modprobe and AT commands.

Come on!

[szo]

The changelog _is_ available. It's just not in one piece. You might as well complain about the lack of changelog file 2.2.21-2.6.12.
If you can understand a single word from the changelog, you must be already intimitely close to the linux development, who knows whats going on. So you probably don't even know what's inside in a changelog like this.

Szo

Re:Come on!

[NitsujTPU]

I'm not actually complaining about the absense of the changelog... or even complaining for that manner.

I merely stated that, I bet that a handful of guys will come in on Monday. They've been promoting Linux to their company.

Some manager who reads /. will have seen that article, and will use it in an argument for Windows.

That's all. Don't read any more into it than that please.

Re:Come on!

[szo]

In this case, if you one of the guys promoting, then simply point it out that the kernel is a just a kernel, and it goes through a number of filters before it reaches the users of the distributions.

Fuck you, moron.

[Anonymous Coward]

"Open source isn't the be-all end all."

With Linux, yes it is. If you don't care about open-source then use something else.

Don't be an idiot. What, people should only use software because they happen to support your ideology? Whatever happened to "use the right tool for the job"? I'm going to continue using Linux with my "binary only NVIDIA drivers" because it works and you can suck my nuts.

Kernel 2.6.12 and GCC-4.0

[Anonymous Coward]

I compiled Linux kernel 2.6.12 under GCC-4.0 (gcc 4.0.1 20050522 -- Debian 4.0.0-9) nicely! Just 'make bzImage && make modules && make modules_install'. :-)

I'd like to see you create and maintain . . WTF?

[Anonymous Coward]

Oh, that's a sweet argument you're putting forth there.
Since nobody without billions of dollars in investment capital can enter the ASIC graphics chip market in a big way those peasants opinions are all nothing more than meaningless chatter. Right? They're all peon idiots and should take what they're given and shut the fuck up. Isn't that right?
That's beautiful. Your logic really shows your insight. You are a wise fellow. That much is obvious.

That's nothing

[rsilvergun]

I hear 2.6.11.12 compils on the Phantom!

Cool

[slicedoranges]

I live Linux, so this is indeed good news.

Wrong word

[phorm]

Yes, because the word he should have used is Zealot. Which in large doses is just as much a pain in the ass as it is for various religions etc. Be happy sitting on your little pedestol, just don't expect all the rest of us to climb up to it too...

Re:Wrong word

[nathanh]

Yes, because the word he should have used is Zealot.

He did use the word zealot. It's very amusing because apparently a lot of you think "zealot" is an insult. I'm not insulted by that word at all. I'm proud to be a zealot.

zealot (n): one who engages warmly in any cause, and pursues his object with earnestness and ardor

I pity all of you people who are so jaded with life that you can only express yourself with anger and cynicism. It's so... teenager. Try being a zealot. It's much more fun.

Be happy sitting on your little pedestol,

PS: the word you should have used is "pedestal". HTH. HAND.

Plain english changelog?

[DroopyStonx]

Where are the BIG fixes, changes, and additions?

They really should have a section in the changelog for this.

Things like "fixed function_t() in blah module" doesn't really shine a whole lot of light on things...

Maybe a section such as, "Fixes in this kernel release will improve [list of things]"

patch collections

[Her0]

you "patch collections" real thing or virtual/in_todo_state ?

do you have url to it (if 1st)?

Yes, but there is more

[pkphilip]

Yes, the fact that the Linux kernel point releases happen quite often does cause real problems for hardware companies.

Linux's driver architecture is flawed - this is because the way the Linux kernel must be installed onto a machine is in itself a flawed process.

Here is why I consider it flawed:

The device driver must be recompiled for each point release of the kernel. That is, when I install a new version of the Linux kernel (even if it is a minor point release), I must necessarily recompile all the device drivers for that point release of the kernel or I must get the precompiled driver modules (for that exact point release) and install them.

This is not how Windows 98/2000/XP works - you can install a kernel update (Service Pack) and still continue to use the same device drivers - you won't have to download new device drivers or recompile existing ones.

Why is this a flaw? It is a flaw because it expects all hardware manufactures / device driver writers to provide either:

The source code for each point release of the kernel.

or

The precompiled driver for each point release / patch of the kernel.

This is not going to work since there are so many different linux kernel versions and a whole range of patches (-ac etc) out there.

This requirement to recompile device drivers for each kernel release can be real problem - I experienced this problem first-hand recently when I tried to install Linux on a machine with a SATA harddisk and a southbridge chipset (SIS965L) which is not supported by Linux (might be fixed now).

During the installation, the harddisk was not being found since the Southbridge chipset was not being recognized - the distribution (CentOS) did give me the option of loading the device driver from a separate CD/floppy, but the problem is that I only have the the device driver source code.. and I cannot install the device driver unless I compiled it for that specific kernel version of the CentOS distribution I was trying to install.. and I cannot compile for that version of the kernel unless I already have a Linux machine with that kernel installed - a catch-22 situation.

Ideally what we should have is something like this on the file system: /kernel (with subdirs /2.6.11 /2.6.12 etc) /drivers (in compiled form) /conf (where we list which drivers are to be loaded with settings for each with subdirectories for different profiles /server /gaming etc). This could also be in the /etc directory.)

One should be able to change the contents of /kernel without updating either /drivers or /conf unless there is a huge change in the structure of the kernel or a change in the device driver interface in the kernel.

This way, I can take the backup of the /drivers and the /conf directory to a disk and if I have to reinstall the Linux on my machine (even if it is a different distribution of Linux with a slightly different kernel), I can just restore the /drivers and /conf directory to setup all my hardware properly.