Slashdot Mirror

The article says Linux (pronounced "LINN-ix"). That is incorrect. In the samples you can download from kernel.org you can hear Linus pronouncing Linux.

It's none of their business to decide who can use their stuff and who can't. If this were allowed wouldn't we have companies deciding not to sell their goods and services to homosexuals and black people?

I don't think it's right what the Chinese government is doing but I also don't think it's right to pick and choose who to sell to. I mean if they didn't buy Microsoft they'd download something better and it would make life even harder for those poor Chinese.

mass-distribution before we even hit 2.6.2 might be a BIG mistake

FC2 test1 should be out next week with a post-2.6.1 kernel as default. With SELinux to boot, though it's recommended to disable it at boot time for test1. Mandrake 10.0 beta1 has 2.6.1 too.

http://gcc.gnu.org/projects/tree-ssa/
http://people.redhat.com/dnovillo/pub/tree-ssa/sna pshot/
http://people.redhat.com/dnovillo/pub/tree-ssa/sna pshot/gcc-ssa-3.5ssa-0.20040129.snapshot.src.rpm (24.4M)
http://people.redhat.com/dnovillo/pub/tree-ssa/pap ers/
http://people.redhat.com/dnovillo/pub/tree-ssa/pap ers/gccsummit-2003-proceedings.pdf (1'383'108 bytes)
== http://www.linux.org.uk/~ajh/gcc/gccsummit-2003-pr oceedings.pdf (1'383'108 bytes)
http://people.redhat.com/dnovillo/pub/tree-ssa/pap ers/nordu2003-slides.pdf (102'895 bytes)
http://people.redhat.com/dnovillo/pub/tree-ssa/pap ers/nordu2003.pdf (153'101 bytes)
http://people.redhat.com/dnovillo/pub/tree-ssa/pap ers/tree-ssa-gccs03-slides.pdf (90'925 bytes)
http://people.redhat.com/dnovillo/pub/tree-ssa/pap ers/tree-ssa-gccs03.pdf (87'507 bytes)

ftp://ftp.kernel.org/pub/linux/devel/binutils/binu tils-2.14.90.0.8.tar.bz2 (11'015'696 bytes)

open4free

One thing's for sure. The won't be any software flame wars over which crappy OS is running the Mars show. Hint: it's from neither Washington (state) nor Finland.

>And from experience, if you're looking for a stable Linux distribution, Red Hat ought to be one of your last choices.

umm...Red Hat has always worked just fine on my end (RH6x,RH7x,RH8x). I would take you seriously on this concept of distribution-stableness if this were BSD, but this is linux country where distributions are modular! Packages can be downloaded and recompiled on a whim. To judge a linux distro by its kernel release (or compile) is naive at best, if not shameful. Most of the code included in any distro is pretty much outdated within weeks and there are several kernels to download if you dont want to recompile the stock one.

If your getting core dumps or kernel panics from a RH distro, then you should check redhat's errata, download fresh code, or check to make sure your hardware is both supported and functional. Not to start the flamewar, but you should consider the needs that slackware, gentoo, knoppix, RH, etc. fulfill rather than the kernel they stock before you start bashing linux distros. For example, RH can serve as a good home install because of the number of packages it packs. It also provides the novice with a good introduction to this concept we call "choice" -something most locked-in windows users do not have, at least not without a $price$. Or if you need something slim and fast for older machines whose use will probably only be a private server or router, then there's slackware. Or if you need an emergency boot disk or merely want to test how linux runs on another machine (ie. BestBuy display models, laptops, etc.) there's knoppix.

>most crash-prone Windows machines have been "crashy" due to poor hardware setup/faulty hardware.

Actually, most "crashy" windows machines, especially win9x, were due to unchecked memory leaks throughout the entire system which could have been prevented if not for that whole 1995, 1998 deadlines they imposed on themselves. And then there's all the crap they stuff into their OS. I proved this to myself a little while back. You see, I wanted to reduce the help calls I got from family and friends, so I decided to build a windows install (actually a bootable system disk with a ghost image) that everyone could use and wouldn't contain so many bugs and security vulnerabilities. So I edited the install .inf files to a windows ME installation in order to prevent the installation of useless .dll's, exe's (based on google searches to figure out what each one did), windows media player crap, and internet explorer. Basically, I just opened precopy1.cab, perused layout*.inf to see which cab files had the .inf's I was looking for, and deleted everything under the file header. Sometimes, minor edits were only possible to remove things like "winpopup" since microsoft included it in the tcpip .inf's. (Why?) After recabbing, installing (using an install script, msbatch.inf, to skip being bothered by user prompts and a few "secret" setup.exe commandline switches), and a little tweeking, I had the install down to 150mb (not including the backup cab files)! The shell didnt work, but it was crap anyway so I replaced it with LiteStep (smaller, faster, configurable) and used an alternate file browser (AB Commander) instead. After installing everything I would need for a bare bones install, I ghosted the machine onto ONE CD. Try doing that with XP! Anyway, the machine has been tested for over a year now under several hardware variations and user patterns, and can run for several weeks without a restart. Which just goes to show that the extra CRAP does have a significant impact on your systems performance. Take for example "active desktop", SFP, "system restore", all the failed "innovations" that didnt make it to win2k and XP, and all the excess registry keys that are associated to other MS products you may never use. While some of this crap may work for some, it is rather presumptious of Microsoft to assum

OK you got 1 of 6 correct. You fail.

Linux and FreeBSD kernels, which
1. is the most SMP scalable (parallel)

Linux. (Linux and FreeBSD both started with a BKL about 5 years ago. Linux is now being used on 512 processor machines, FreeBSD doesn't scale past 4 CPUs for all their research).

2. is the most algorithmically scalable
Linux

3. has fastest single threaded performance
Linux (from the horse's proverbial mouth).

4. runs on more architectures
Linux. See here and here (The Linux Kernel supports more architectures than the NetBSD kernel, idiot).

5. supports the most hardware
Linux. See here and here (NetBSD is actually the one that prides themselves as running on toasters. And nobody cares about your shitty DEC Alphas. IA64, POWER are where its at now).

6. has the fastest TCP/IP stack
Linux

See you again next year.

no one here really. ;)

• Free, but worth thousands more: FreeBSD, Linux, and more...
  
• Pricey, but worth every penny: Mac OS X
  

Nothing I described in any way limited what you could do with your computer.

Maybe I misread something, but that's not how I read your words. You are talking about putting legal limitations in place so that only properly signed content can get distributed through peer-to-peer networks and then outlaw the distribution (and thereby effectively also the use) of clients that do not include such checks. That would seem to be a limitation on what one can do with their computer to me at least. Maybe I am missing something?

By the way, peer-to-peer networks do not "broadcast" content. They allow someone who has a piece of digital information to share it with someone who doesn't, but wants it. That's more like having someone ask a question and then you answer it. If you ask a question you have to expect an answer. Pretty simple.

As for the copyrighted == non-distributable section, that was not directed at you specifically. My apologizes if I did not make that as clear as I should have or intended.

This is all beside the fact that another good thing about peer-to-peer networking is that content can be untraceable. Once a few people have downloaded and made available for download a file, regardless of its content, it's pretty hard to find out who first distributed it unless the originator or author steps forth. Which in itself helps guarantee free speech.

(I also put up both OpenOffice 1.1 and the Linux kernel for download from my Gnutella node. Within days several people had downloaded copies of both (not necessarily the same people, obviously). It certainly proved my point, that peer-to-peer networks have perfectly legal uses, and can help take some of the load off the official distribution servers. Before anyone comments that the files may have been altered, that's when the cryptographic hashes come in handy. Downloading a file of a few hundred bytes certainly puts less load on a host's Internet connection than downloading tens of megabytes. Given appropriate checks of public key authenticity by the recipient, the hash files can easily get distributed over the same P2P network as the data they are used to verify.)

Yes, but until the release of kernel linux-2.192.367-test54283.tar.bz2, only the 'F', 'A' and 'G' keys work.

To enable the other keys, you have to compile and install a kernel module Neko64, but this will also break the framebuffer support.

We're talking about Linux vs FreeBSD here, clown, not to mention that you provided zero evidence. You got failed.

Linux and FreeBSD kernels, which
1. is the most SMP scalable (parallel)
Linux

2. is the most algorithmically scalable
Linux

3. has fastest single threaded performance
Linux

4. runs on more architectures
Linux. See here and here

5. supports the most hardware
Linux. See here and here

6. has the fastest TCP/IP stack
Linux

Hope this has been helpful. See you again next year.

i mean, a one man OS is impressive, but i can't see anyone actually using it...

I'm actually refering to this 2.0.x.

that's an old email. 2.6.0 kernel changelog torvalds is at OSDL (open source development labs).

or goto slashdot archives on the topic.

or visit the lkml archives and read the email addy from the already done search.

Users of their distribution's kernel should install upgrades when their distribution releases an update (should be soonish). Those who prefer stock kernels can grab patches from your local mirror or use my torrent (includes linux-2.4.24.tar.bz, pgp signature and 23->24 changelog).

When was the last time a 2.0 series kernel was released with a new patch?

Would OSS have to be writen entirely by licensed developers to be considered secure?

I'm sure glad the DHS steps in and prevents all those 1ee7 uncontrolled hackers from creating evil unlicensed, software that aren't secure.

Why do I always picture half-drunken bar patrons reinventing the world in front of a beer when I hear about the DHS talking about things they don't have much of a clue about?

go to kernel.org, get the kernel source, and write your own. i'm sure linus would be up for a better scheduler.

Linus commented that he himself remembers writing those files. Well, thanks to Kernel.org and a little too much time on my hands, I did a little exploring. Kernel 2.6.0 has errno.h in two files. To make my life a little easier, I combined the two files, errno.h and errno-base.h. In Kernel 2.3.50 it is one file.

Well, as we know, SCO is claiming that 2.4.21 is the kernel that started with the problems. If that is the case, assuming that SCO actually has a case then we have a problem.

But the thing is that the errno.h and errno-base.h in 2.6.0 and the errno.h in 2.3.50 have only one difference other than being split up and the appropriate location indicators. The only difference is:

#define E2BIG 7 /*Argument list too long*/ -- 2.6.0
#define E2BIG 7 /*Arg list too long*/ -- 2.3.50

I obtained this by using diff. So a simple utility disproves SCO's claim on that ground. Also, you will notice that the Kernel v. 0.01 has only 39 error numbers. They are also included, with the same error numbers in the current 2.3.50 and the 2.6.0 files. A cursory look revealed that 2.4.23 has the same errno.h err codes.

So when Linus says that he wrote them there is proof. Further, since 2.4.21 is the infected one, what is the difference between 2.3.50 and 2.4.23 and the comments. Surely SCO can not be so stupid as to say that comments are a cause for action -- the end user does not even see nor are they accessable to the end user unless they have the source.

Linus commented that he himself remembers writing those files. Well, thanks to Kernel.org and a little too much time on my hands, I did a little exploring.

Kernel 2.6.0 has errno.h in two files. To make my life a little easier, I combined the two files, errno.h and errno-base.h. In Kernel 2.3.50 it is one file.

Well, as we know, SCO is claiming that 2.4.21 is the kernel that started with the problems. If that is the case, assuming that SCO actually has a case then we have a problem.

But the thing is that the errno.h and errno-base.h in 2.6.0 and the errno.h in 2.3.50 have only one difference other than being split up and the appropriate location indicators. THe only difference is:

#define E2BIG 7 /*Argument list too long*/
#define E2BIG 7 /*Arg list too long*/

So if you buy SCO's argument they are saying that a comment is to blame on this. Again, this is an SCO FUD campaign, but come on.

Thanks to diff for the comparision.

Acutally this is a much discussed bug, it has already been fixed (As of last night) in Andrew Morton's patch set 2.6.0-test11-mm1.

Andrew is in charge of 2.6 now and he'd probably include this patch in 2.6.1.

Specially the patch that would fix this problem would be this and it could be applied to the vanilla 2.6.0 kernel without any problem.

Acutally this is a much discussed bug, it has already been fixed (As of last night) in Andrew Morton's patch set 2.6.0-test11-mm1.

Andrew is in charge of 2.6 now and he'd probably include this patch in 2.6.1.

Specially the patch that would fix this problem would be this and it could be applied to the vanilla 2.6.0 kernel without any problem.

Got my speedy BitTorrent download, thanks. But how do I check it's validity? No MD5 to download from kernel.org? Use the .sign - read http://www.kernel.org/signature.html

From the "should-fix " list that was ignored for 2.6 release:

drivers/net/wireless/
~~~~~~~~~~~~~~~~~~~~~

(Jean Tourrilhes )

o get latest orinoco changes from David.

PRIORITY 1

2.6 is still using sucky wireless drivers that lock up the system when the wireless hub goes down!

I guess I'll wait for 2.6.5 for this to be fixed :(

Be sure to check the PGP signature if you download from an unofficial mirror.

Instructions here.

Be sure to check the PGP signature if you download from an unofficial mirror.

Instructions here.

How does this benefit me?

RTFM ChangeLog for a detailed explaination. Or go back to this slashdot story on the linux 2.6 kernel.

The 2.6.0 kernel is out (within the last 20 or 30 minutes). Go grab it here

Or are there actually some decent altruistic people out there who want to make the world a better place?

Uhh... I think you're missing the point. SCO's contention is that IBM has placed, in Linux, code which is actually SCO's, or at least derived from SCO's code. And according to the article, SCO tried to "[pressure] the courts to force IBM to reveal its Unix and Linux source code". But that doesn't make any sense, as SCO can *already* see it, what with that whole "open source" thing. Hell, if you want, you can see it too! Thus, clearly, the author of the article has no fscking clue what he/she is talking about.

Fun for the whole family with guess where the patch was applied From the snapshot directory

bk6 - 424K

bk7 - 964k

bk8 - 1.2M

Well thats increased the kernel by about another 5-10%. However I would say I do like xfs and its proven quite stable now.

Rus

"people don't have to use it. They can tell Marco and Linus to fuck off for stranding them after they told their execs that Linux was better and more stable - which might be the case for the server uptime, but perhaps not for the development model."

Stranding them? You mean like providing both a kernel for new features (2.6) and a kernel thats more reliable(2.4)?

"And the companies that make money off Linux, especially the ones like Red Hat who have yet to use any sort of 2.6 for anything, who EMPLOY people who make major contributions, don't have to continue to contribute the time of their employees."

Then a redhat developer like Alan Cox should apply the patches he wants and maintain his own kernel fork..oh wait.

I think SCO will either give the information...

Personally, I use a FAT32 partition to transfer data from one OS to another I know I'm not the only one.

Guess we need to get the ext2 driver working a little better, though. Just in case.

It was introduced in 2.4.23-pre7, disguised as "Add TASK_SIZE check to do_brk()" in the changelog.

If you aren't averse to compiling your own kernel, the fix is a really easy two line patch. Just add the following to mm/mmap.c at line 1047 (immediately after "if (!len) return addr;")

I'm enjoying the thrill of compiling patched kernels on two different machines as I write this. Thank goodness for Debian's make-kpkg.

You need to apply the cryptoloop-jari patch on top of the 2.4.23 kernel.

Some people reported that you need to use updated userspace tools and the "hashalot" tool as well, but for me applying the patch above did the trick.

I agree that it's disappointing that the cryptoloop support is only partially integrated, since the correct instructions on how to get it working are hidden among a lot of no-longer-accurate descriptions :-(

-Klaus

A final test11 version is expected before they sign off on the production version next month.

Actually, according to kernel.org, linux-2.6.0-test11 was released today.

and you may want to check bugzilla for the 300+ bugs reported for 2.6 ...

Why would you want to pay through nose for a proprietary, no-support, closed source *nix when you can have a modern, high-performance version for free?

For example, Python has evolved into an extremely intuitive yet powerful programming language.
Perl was also fairly new in its time.
There's GNU Emacs which is one of the most powerful text editors in existence.
There's the Apache Webserver. Although webservers aren't new, I would hardly call Apache a copy of anything.
I'm not sure whether the first publicly-released blog software was open source, but I think it might have been.
OpenBSD was, AFAIK, the first secure-by-default modern Unix system.
Linux (the kernel) has also done (or been modified to do) several things not done before.
X11 started as a project out of MIT (which I would guess was open-source, even though the phrase hadn't been coined yet.)
GNU readline is also something that is exclusive to open source
I'd guess that ls --color was something new to free software, as well, just because I douby anyone with a pure profit motive would consider it worth the time to implement. :-)
The Debian Project has made several innovations in operating system integration.

Anyway, there are plenty of examples. You just have to look.

at first sight, I'd say here

But maybe some people believe that innovation is what you can only see.

Regards,
jdif

BTW, I'm not faking Linus, here is the signature :)

OK, this might be redundant by now (there wasn't any replies when I started writing this), but I post it anyhow in case it has additional information:

Debian GNU/Linux always has at least three releases in active maintenance: "stable", "testing" and "unstable".

There is different ways of getting Debian. A new way of installing is prepared for Sarge, but works for other versions too I think.

/Spam .

This POSIX kernel is different to that POSIX kernel which is different to that POSIX kernel is totally different to that POSIX kernel.

POSIX; it's what's for dinner!

The architecture is 10 years old. Why would I want to develop on an old technology when I can get it done 10x as fast on the latest technology

Karma to burn so,
PUT DOWN SCO's CRACK PIPE

Ok, Steve, Linux has some of the most advanced features in the world like Kernel level exectuion checking, and True 100% fully functional streemlined , low latency memory management. Linux, and for that matter GNU, have grown at least ten times more than windows has over the same period of time. Microsoft has been using the win32 api for almost 7 years and the individual systems are so similar that a security hole in WinNT 4.0 will allow a virus to be created that destroys windows server 2003, while the kernel (22 stable kernels in 2 years) GNOME and GTK+ and Qt/KDE have not gone a year without major improvement, and is there anything in windows that can compete with the speed of GLADE's GUI development?

Linux is constantly being improved upon and features added, like a Kernel Level cypto API and the futex system, which makes 2.6 incredibly fast, and they took out khttpd because apache was fast enough on a Linux system to kick IIS ass. Microsoft cannot win the argument on technical superiority, because Linux is constantly scrutinized and improved, 2.4.0 was released in Jan of 2001 while 2.6.0 is going to be relaced almost two years later, when is the next major installment of windows comming? How long between XP and Longhorn again?

Dont forget major Kernel improvements in Windows only happen with the upgrade cycle.

Amazing, something that actually made me de-lurk on Slashdot...

Here is my letter to this guy:

From: Kevin B. McCarty <kmccarty@nospam.Princeton.EDU>
To: howard@princeton.edu
Subject: Your article in Syllabus (perspective from a Princeton graduate student)

Sir:

I am a graduate student in the Princeton University Physics Department. I came across your article regarding open source software on Syllabus Magazine's web site, in which you do a grave disservice to Princeton University's reputation of technical excellence. Allow me to elaborate.

You say, with a tad of sarcasm:

"These folks [open source software developers] are some of the same great people who are supposed to be working for you anyway, plus a smattering of teenagers too young to work at Redmond, hackers, virus creators, and a menagerie of others with whom you will feel great pride in entrusting your IT infrastructure."

I am interested, then, in how you feel about the Princeton University web servers at www.princeton.edu running Apache, the most well-known open source web server. Apparently [1], Apache has more than 2/3 of the web server market share on the Internet, so someone must trust these people. Of course, the fact that source code is available for open source projects may have something to do with this trust. By the way, how many open source viruses have you seen? (Microsoft Word macros don't count.)

[1] http://www.netcraft.com/

You say:

"We may have to give up project planning, quality control, coding standards, accountability, version control, and support, but it's FREE and we get the ability to modify the source code ourselves, something that is extremely dangerous to do, was discredited decades ago, and few people do anyway."

Really? Who discredited the ability to modify source code? Did I miss a Congressional report or something? I apologize for calling you dead wrong, but in fact the Linux kernel [2], one of the most successful open source projects in existence, has been continually updated and improved since its first release in 1991, all by people with an interest in changing source code. These "dangerous" modifications have strangely made Linux and its BSD Unix cousins more stable than any release of Windows. The open source software development process is self-regulating: stable, good software survives, while low-quality efforts are ignored and drop from the face of the Internet. It is too bad that mediocre commercial software does not do the same, since it is too well-supported by people who will not consider using anything they are not required to pay for.

[2] http://www.kernel.org/

You say:

"We either pay commercial software developers, pay to build it ourselves, or pay the even higher price to manage and maintain FREE open source software."

I don't suppose you are aware of the existence of companies who provide support for open source software. Believe it or not, it is possible to buy a support contract from most major Linux distributors, e.g., [3]. It is even possible to ask (politely) for FREE support on open source message boards, such as [4], where you will usually get far more helpful responses than the standard Microsoft "Have you tried rebooting? Reinstalling?".

[3] http://www.redhat.com/apps/commerce/
[4] http://lists.debian.org/debian-user/

You say:

"Another way to get free software is to have students develop our critical systems," and "You can also get free software developed by having your users develop it for you."

These are ridiculous straw man arguments. No sane system administrator would tell his/her students or users to develop their own softwa

• does run Linux
  
  
• proofs that Mirosoft is making bad software
  
  
• shows how stealing intellectual property simply cannot be wrong ("Actually, it is not stealing") because every fucking idiot does it
  
  
• assumes that any law preventing child molesters from distributing their filthy shit is an attack on Free Speech
  
  
• promotes homosexual, interracial or interspecial cohibitation
  

• does run Linux
  
• proofs that Mirosoft is making bad software
  
• shows how stealing intellectual property simply cannot be wrong ("Actually, it is not stealing") because every fucking idiot does it
  
• assumes that any law preventing child molesters from distributing their filthy shit is an attack on Free Speech
  
• promotes homosexual, interracial or interspecial cohibitation

• does run Linux
  
  
• proofs that Mirosoft is making bad software
  
  
• shows how stealing intellectual property simply cannot be wrong ("Actually, it is not stealing") because every fucking idiot does it
  
  
• assumes that any law preventing child molesters from distributing their filthy shit is an attack on Free Speech
  
  
• promotes homosexual, interracial or interspecial cohibitation