Welcome to the MacNN Forums.

NeXTLoop · Jan 22, 2003, 04:36 PM

Evidently, since acquiring SCO, Caldera thinks it should be compensated for all the different uses of technology it feels is its intellectual property.

This would include Apple, M$, BSD/OS, FreeBSD, NetBSD, and OpenBSD.

Take a look...

Pepi Picklefoot · Jan 22, 2003, 04:43 PM

It almost feels like a joke...except it's not funny.

cwasko · Jan 22, 2003, 04:50 PM

I read the article, and I still can't figure out what SCO actualy owns, and why they think compaines owe them 149$ per server.

MacGorilla · Jan 22, 2003, 05:38 PM

First they are going to have figure out the twisted history of Unix, wade through past lawsuits and agreements. Good luck!

Graymalkin · Jan 22, 2003, 06:25 PM

I think SCO is going to be on a slippery slope with this entire process. Caldera doesn't have much room to complain considering any of their supposed IP was implemented openly and thus not bound by some kind of NDA or exclusive license. SCO is an absolute failure at this point and they're trying to sue for some money to pay off debts and maybe last another year. If anything it will get their name in industry rags at the cost of all their community street cred. Not only do they lose any community status they may have had but they look like buffoons to their potential customers. A smart manager is not going to contract with a company that is reduced to suing other companies to make money. Bye SCO, see you on F'd Company pretty soon.

NeXTLoop · Jan 22, 2003, 06:45 PM

Even forgetting the problems they would have proving their case legally, let's take a look at the ethics involved...

Caldera/SCO is looking for a way to make money off of, among others, Linux vendors. Let's see....... what was Caldera before buying SCO???

Oh yeah!!! THEY WERE A LINUX VENDOR!!!! This whole investigation is a crock of crap. If they try to make a case of this, they should have to pay somebody something, just like they're considering trying to make other companies pay. After all, if they try to make a case of this, then by their own standards, they've been just as guilty of the same conduct as the companies they would be trying to force to pay.

bzImage · May 17, 2003, 12:23 AM

I use to work @ SCO, but then Caldera came and lay off the entire company, now its managed by a bunch of greedy guys in khakis.

So bad, it was a really nice place to work.

ngrundy · May 17, 2003, 01:02 AM

SCO Is on some really good crack.

They currently own the rights to the original AT&T UNIX Source code. AT&T sold it off to Novell at one point. It then became Calderas mantle and now SCO is raping it for every cent it's worth and then some.

When Linux was being wrote and such Novell (or AT&T) was hauling Berkley (BSD) through the courts over the use of AT&T code in their software. If this law suit wasn't going on the free software landscape today could look very different, but i digress.

After the court battle you ended up with 'two tapes' Net1 and Net2. Net1 was a tape that contained AT&T UNIX including all the proprietory code. Net2 was a copy of the source code with the AT&T stuff ripped out. This constituted about 9 major components. Net2 was unbootable and unusable.

Some random guy took the Net2 source and wrote the missing sections. after a while you got things like 386BSD, 386BSD Patch Kit. and then onto FreeBSD. Later on NetBSD forked from FreeBSD and then some time after that OpenBSD forked from NetBSD.

Darwin is in core based on the BSD4.4-Lite code which is the Net2 tape with the vital code back in it (wrote freely under the BSD license for anyone to use).

Linux was developed independantly and on the other side of the world at the time the legal rubbish was going on.

If anything SCO should be pointing fingers at the BSD family not the Linux distros.

Money grabs love them. I swear SCO is the only UNIX mob out there that isn't rolling silly in the Linux/UNIX hype at the moment. Might have something to do with THE INSANE LICENSE COSTS. But I dunno

geekwagon · May 17, 2003, 03:17 AM

It's almost as if SCO doesn't realize the split between SYSV and BSD happened and the circumstances surrounding it. Reading this article it seems fairly obvious that they have just decided that anything that was owned by AT&T at one point is theirs and theirs alone, even if it was later decided to be the code actually written at Berkley. It seems to me that this will end very badly for them, as I can't see how any AT&T code could possibly have made it back into BSD after all the bad blood from the legal precedings. I was but a strapling sysadmin at the time, but I remember all the developers being very particular about the code going into BSD having absolutely no ties to SYSV.

theolein · May 17, 2003, 07:00 AM

This whole SCO is worrying and funny at the same time. It is very hard to believe any of their claims because they are so incredibly inconsistent. The timeline goes something like this:

In the beginning SCO, which has been on the road to bankruptcy for a while now, decided to sue IBM for $1 Billion for alledged infingements that IBM made when IBM and SCO were in some kind of partnership. SCO basically claimed that IBM used AIX (IBM's Unix) to move Linux from being able to run on more than 4 processors to be able to run on 32 processors. SCO claimed at the time that "we aren't going to go after other Linux companies". SCO made the statement that "no independant Linux developers would have had the resources to be able to develop on large enterprise machines". Alan Cox, a unix Kernel maintainer and developer, wrote a disclaimer that SCO/Caldera itself loaned him a large machine to do exactly that development.

Straight after this diclaimer, SCO suddenly started sending out the warning letters to all Linux distributors that they might be using SCO's IP and would be liable, breaking their own word on their course of action.

They then started claiming that "there were infringments prior to and after IBM's involvement, and that basically there are infringements all over the place in Linux". They also stated that they would release the actual names and line numbers of the offending code "in a couple of weeks". Linux dvelopers on /. and elsewhere then stated that if SCO were to actually make concrete claims on the files and line numbers that were in offence that they would delete those files and replace them in a week.

SCO now is trying to claim that BSD derivatives are in infringment of the SYSV code. Someone further up has pointed out that this battle has been fouht before and BSD has no relation to the UNIX sources anymore and Linux was developed across the planet.

SCO made a small statement in the CNET article that they would charge $149 per processor running Linux.

And this is obviously what it boils down to. SCO has no real idea if there are any real IP infringements in the Linux or BSD code, but they are going bankrupt and are clutching at straws in order to find more money somewhere. This is what this suit is. They are playing a very high risk game, and are opening themselves up to being counter sued by, amongst other, IBM, Linus himself, RedHat, Microsoft, SuSE, Apple and HP.

I struggle to think why they are doing this as it is so obviously risky. Two things come to mind:

One is that they know they are going bankrupt (their losses over the past few years have been huge) and assume that even if they lose and get counter sued out of existence, they have nothing to lose. They know full well that no one is going to pay them $149/cpu after the fact, but they think they'll give it a try anyway. They also know that they will lose all their current customers with this piece of sh1t that they are doing.

The second thought is more sinister and paranoid. This whole campaign sounds so much like Microsoft's smear and FUD campaigns of recent years, that I ask myself if it isn't indeed something that was thought of in Redmond. Watching SCO's bank account payments would be very interesting. The fact that older Windows code uses the BSD TCP/IP stack (actually I thought XP still uses it) is interesting because a legal decision against all these Unix clones and derivatives (although I sincerely doubt that this would ever happen) would serve to boost Microsoft's newest OS's, and would badly hinder competition against Microsoft.

I wonder where this will lead to?

Graymalkin · May 17, 2003, 07:51 AM

I've been following the SCO crap for a while now but this is the first I've heard about them going after the various BSD Unicies. Every Unix based of BSD 4.4 Lite is free from any and all legal qualms AT&T ever had. Ngrundy is correct about the two tapes, the Net2 tape became the basis of the entirely BSD license Unix release which contained zero code from AT&T. Someone even mentioning BSD 4.4 Lite derivitives containing proprietary AT&T code needs to be beaten with a stick.

Then again this is Caldera we're talking about, the adled handicapped Linux vendor. The only AT&T Unix references I've ever found in OSX code are program history summaries of the sort you find in man files. I'd really like SCO to just show someone some real code infringing their IP. They never will and likely can't do this so they're just trying to scare people into giving them money to shut them up.

I think they think they will end up with a fat settlement like when they bought DR-DOS and sued Microsoft. Microsoft had to settle out of court because they HAD tried to muscle Digital Research out of the DOS market using their monopoly. They couldn't have an antitrust ruling against them because they were in the middle of another antitrust case brought by the DOJ which you might remember.

Caldera has bitten off a bite more than it can chew with IBM however. IBM doesn't have the anti-IBM sentiment in the industry that Microsoft had when Caldera went after them. IBM also quickly removed the case to federal court where disclosure procedings will go much quicker than they would in a state case. This means Caldera doesn't have time to drag out the FUD they're currently throwing out. The crap about $149/processor licensing and going after BSD derivitive Unix systems is just FUD to get hits on their websites and make PHBs worry that the Linux installation they're rolling out in the office might get them into legal trouble.

alex_kac · May 17, 2003, 02:05 PM

Also, remember that they are talking about the KERNEL only - and here OS X has nothing to do with BSD.

OS X runs on the MACH kernel. OS X is NOT a BSD variant. It has a GNU toolset that runs on top of Mach as well as BSD APIs making it compatible with BSD - but it no way IS BSD or uses any kernel elements of BSD or any other UNIX for that matter.

In fact, OS X isn't even UNIX. Its just UNIX - like, just like Linux is, except its TECHNICALLY farther away from Linux, BSD, Sys5 due to the Mach kernel.

So Apple is 100% free from any of SCO's claims - unless they start going after the GNU toolset, which would be like Microsoft going after AutoCAD for drawing boxes on the screen, i.e. nothing even remotely similar.

macmike42 · May 17, 2003, 03:07 PM

Netcraft conrifms: SCO is dying

oh nevermind...

someone_else · May 17, 2003, 08:38 PM

Originally posted by alex_kac:
Also, remember that they are talking about the KERNEL only - and here OS X has nothing to do with BSD.

OS X runs on the MACH kernel. OS X is NOT a BSD variant. It has a GNU toolset that runs on top of Mach as well as BSD APIs making it compatible with BSD - but it no way IS BSD or uses any kernel elements of BSD or any other UNIX for that matter.

In fact, OS X isn't even UNIX. Its just UNIX - like, just like Linux is, except its TECHNICALLY farther away from Linux, BSD, Sys5 due to the Mach kernel.

This is wrong. The Darwin kernel is made up of the Mach kernel and a BSD based kernel that sits "on top" of Mach. The BSD/Mach portions share the same address space (ie kernel space) and are included in the same binary file (mach_kernel). The BSD portion of the kernel handles networking, filesystems, user security, processes, IPC, and lots of other things. The Mach portion deals with the hardware, VM, thread scheduling, and IPC primitives (mach ports).

So Apple is 100% free from any of SCO's claims - unless they start going after the GNU toolset, which would be like Microsoft going after AutoCAD for drawing boxes on the screen, i.e. nothing even remotely similar.

This is true. Apple's BSD portion of the kernel is mainly based off of NetBSD which is a direct descendent of BSD 4.4 Lite, which is totally free of AT&T SYSV code.

Apple doesn't have anything to worry about from SCO. Although that doesn't prevent SCO from filing a frivalous lawsuit just to try to get a settlement.

alex_kac · May 17, 2003, 09:41 PM

Originally posted by someone_else:
This is wrong. The Darwin kernel is made up of the Mach kernel and a BSD based kernel that sits "on top" of Mach. The BSD/Mach portions share the same address space (ie kernel space) and are included in the same binary file (mach_kernel). The BSD portion of the kernel handles networking, filesystems, user security, processes, IPC, and lots of other things. The Mach portion deals with the hardware, VM, thread scheduling, and IPC primitives (mach ports).

This is true. Apple's BSD portion of the kernel is mainly based off of NetBSD which is a direct descendent of BSD 4.4 Lite, which is totally free of AT&T SYSV code.

Apple doesn't have anything to worry about from SCO. Although that doesn't prevent SCO from filing a frivalous lawsuit just to try to get a settlement.

Tis true, I may be wrong since I don't involve myself with the kernel, but from what I read from Jordan Hubbard as well as other Apple developers at other forums regarding the discussion of BSD vs Mach - I believe I'm right and that it is NOT a Mach kernel with a BSD kernel layer. It is a BSD personality that is compatible with BSD, but is not derived from BSD.

But as I said, I'm not Mr. Hubbard - I've only read his words and that was my understanding.

someone_else · May 17, 2003, 11:33 PM

Originally posted by alex_kac:
Tis true, I may be wrong since I don't involve myself with the kernel, but from what I read from Jordan Hubbard as well as other Apple developers at other forums regarding the discussion of BSD vs Mach - I believe I'm right and that it is NOT a Mach kernel with a BSD kernel layer. It is a BSD personality that is compatible with BSD, but is not derived from BSD.

But as I said, I'm not Mr. Hubbard - I've only read his words and that was my understanding.

They can call it a "personality" and be technically correct since the BSD portion won't work w/o the Mach portion; of course Mach by itself is useless too (you couldn't run any applications w/o the BSD layer). But the BSD code in the kernel is directly derived from NetBSD (and some FreeBSD 4.x) and it is therefore BSD.

Here's a quote from : http://www.apple.com/macosx/technologies/darwin.html

Darwin is a complete BSD UNIX implementation, derived from the original 4.4BSD-Lite2 Open Source distribution. Darwin uses a monolithic kernel based on FreeBSD 4.4 and the OSF/mk Mach 3...

MacGorilla · May 18, 2003, 09:39 AM

The BSD Unixes were 'cleared" in the At&t lawsuits years ago. SCO is going after Linux, not the BSDs.

In any case, SCO has produced little more than hot air as of now. Someone translated a German court paper filed by SCO a few weeks ago and appears their case rests on a single obscure Red Hat How-To.

In this How-To Red Hat described how to get WordPerfect 6.1 running on Linux using some shared libraries fromn SCO Unixware. It didn't say where to get them and those libraries were never included in Linux but thats what SCO has. They'll be laughed out of court.

Coxy · May 18, 2003, 11:46 PM

Not to mention the fact that Apple owns a whole OS of System V code.

moonmonkey · May 19, 2003, 12:18 AM

Originally posted by NeXTLoop:
Evidently, since acquiring SCO, Caldera thinks it should be compensated for all the different uses of technology it feels is its intellectual property.

This would include Apple, M$, BSD/OS, FreeBSD, NetBSD, and OpenBSD.

Take a look...

British Telecom patented hypertext links 20 years ago, every website ever made us using BT technology illegaly.

Nobody gives a monkeys.

Stratus Fear · May 19, 2003, 12:50 AM

Originally posted by alex_kac:
Tis true, I may be wrong since I don't involve myself with the kernel, but from what I read from Jordan Hubbard as well as other Apple developers at other forums regarding the discussion of BSD vs Mach - I believe I'm right and that it is NOT a Mach kernel with a BSD kernel layer. It is a BSD personality that is compatible with BSD, but is not derived from BSD.

But as I said, I'm not Mr. Hubbard - I've only read his words and that was my understanding.

Mach is technically what they call a microkernel. It can't run an OS by itself. That's where BSD comes in.

alex_kac · May 19, 2003, 01:01 AM

Originally posted by Stratus Fear:
Mach is technically what they call a microkernel. It can't run an OS by itself. That's where BSD comes in.

And I understand that. But my point was that typically BSD IS a kernel all to itself. I've run FreeBSD 2.1 all the way through 4.4. I've installed the actual iBSD 4.4 system and compiled its kernel, etc...

My point is that OS X is NOT built off those kernels. I admit it may be based off part of those kernels as a personality, but its Mach that's the star underneath as far as I'm concerned as it relates to the SCO stuff. Not only that, but as its been mentioned here before, SCO's issue is with Linux, since BSD was freed from AT&T licensing a long time ago.

macmad · May 19, 2003, 03:19 AM

Don't worry, M$ has just bought a SCO license. Longhorn Unix anyone?

Developer · May 19, 2003, 04:18 AM

Originally posted by macmad:
Don't worry, M$ has just bought a SCO license. Longhorn Unix anyone?

I guess just something like that was what SCO really wanted.

slider · May 19, 2003, 06:58 AM

Hey great topic and very informative, thanks. How does a company like Sun Microsystems fit into all of this?

geekwagon · May 19, 2003, 02:26 PM

Originally posted by slider:
Hey great topic and very informative, thanks. How does a company like Sun Microsystems fit into all of this?

Sun has a SYSV license from way back. SCO can't really say anything to them. Same with HP for HPUX and IBM for AIX. It's only the free software versions that they have any grounds to say anything about. Everyone else has already paid their money.

redboy · 03:11 PM

What's this I hear about Microsoft going to license SCO's Unix patent! Or maybe Microsoft plans to buy the license?

http://news.com.com/2100-1016_3-1007528.html

-Q- · May 19, 2003, 02:55 PM

Originally posted by moonmonkey:
British Telecom patented hypertext links 20 years ago, every website ever made us using BT technology illegaly.

Nobody gives a monkeys.

Actually, no. BT got their a** handed to them in court. I'm hoping the same will occur with SCO.

geekwagon · May 19, 2003, 03:10 PM

Originally posted by redboy:
What's this I hear about Microsoft going to license SCO's Unix patent!

http://news.com.com/2100-1016_3-1007528.html

I'm sure it's just a way for them to legally funnel money into SCO to support the lawsuit.

P · May 19, 2003, 05:12 PM

Originally posted by someone_else:
They can call it a "personality" and be technically correct since the BSD portion won't work w/o the Mach portion; of course Mach by itself is useless too (you couldn't run any applications w/o the BSD layer). But the BSD code in the kernel is directly derived from NetBSD (and some FreeBSD 4.x) and it is therefore BSD.

Here's a quote from : http://www.apple.com/macosx/technologies/darwin.html

That exact quote says that it's derived from FreeBSD rather than NetBSD, actually... Not that it really matters, but I was beginning to feel slightly schizo there.

Darwin 6 - aka Mac OS X 10.2 Jaguar - has a BSD layer that is supposed to be more or less in sync with FreeBSD 4.4. There is NetBSD code there as well, but methinks it's more similar to FreeBSD 4 - at least as of Jaguar.

More interesting though: A/UX was derived from SysV, wasn't it? If so, then Apple must have a UNIX license. Let's hope they didn't put any of that code into the opensourced Darwin...

MacGorilla · May 19, 2003, 05:53 PM

Here is a good read:
http://news.com.com/2010-1071_3-1007...ml?tag=fd_nc_1

MacManMikeOSX · May 19, 2003, 05:57 PM

Well, since sco sells linux under gpl they introdued the librarys into the public so hence open sourcing them. So I don't think they'd be to hard to knock down in court.

someone_else · May 19, 2003, 08:55 PM

Originally posted by P:
That exact quote says that it's derived from FreeBSD rather than NetBSD, actually... Not that it really matters, but I was beginning to feel slightly schizo there.

Darwin 6 - aka Mac OS X 10.2 Jaguar - has a BSD layer that is supposed to be more or less in sync with FreeBSD 4.4. There is NetBSD code there as well, but methinks it's more similar to FreeBSD 4 - at least as of Jaguar.

More interesting though: A/UX was derived from SysV, wasn't it? If so, then Apple must have a UNIX license. Let's hope they didn't put any of that code into the opensourced Darwin...

That's true for userland. It is very close to FreeBSD 3.x/4.x. However the kernel BSD is closer to NetBSD than FreeBSD. Although with every rev. they are using more and more FreeBSD based code.

Morenix · May 20, 2003, 08:18 AM

YESSSSSSS! WE WILL BACK TO MAC OS 9.3!

absmiths · May 20, 2003, 10:36 AM

This was my favorite quote - as if customers have been standing in line shouting "Please, license our free linux servers!":

"The more limited Unix library licensing program announced Wednesday at the LinuxWorld Conference and Expo will let companies pay $149 per server processor to use the Unix libraries, McBride said. It's a program that customers have been asking for."

suthercd · 11:50 AM

Apple is selling for a huge earnings multiple plus is in an industry that is re-grouping. Consumer purchases are squat, Apple's stock is up 40%, why would they buy now?????

Craig

voodoo · May 20, 2003, 12:06 PM

SCO and M$ have a long history of cooperating. M$ bought shares of SCO for $25 million in '89 for instance. Don't know how long they had it or if they still do. SCO is the shame of UNIX.

fritzair · 04:49 PM

If anyone would be kind enough....

Reading the thoughts of so many about UNIX,BSD and AIX. I have noticed that this code is 30 years old. How does this code translate into a modern operating system?
Is the entire kernal rewritten to take advantage of the power of modern processors and such? Or is the code just added on?
Why is it so good? The coder couldn't have known about the changes of future processor designs...

Thanks to you all.

Mike

Superchicken · May 20, 2003, 04:05 PM

This is soo dumb, I guess this is M$'s way of freaking people out about linux without doing it directly.

clarkgoble · May 20, 2003, 05:49 PM

Originally posted by moonmonkey:
British Telecom patented hypertext links 20 years ago, every website ever made us using BT technology illegaly.

Nobody gives a monkeys.

As someone else mentioned, BT didn't win the lawsuit. Oddly the article linked to didn't mention the obvious problem of numerous papers published on hyperlinks long before the BT patent. Xanadu is the most famous "vaporware" product imaginable but was based upon a lot of academic papers published earlier.

http://www.iath.virginia.edu/elab/hfl0155.html

I still meet lots of people trying to patent various hyperlink technologies who somehow miss all the work done in the 60's and early 70's that invalidate most patents of that sort.

Eug · May 20, 2003, 08:34 PM

Hmmm. Interesting read.

It sounds like SCO is RAMBUS'ing themselves. I hope that means most of the rest of the world will start more actively ignoring them too (with the courts' support) in the very least, or counter-suing them into oblivion at worst (or best).

Rainy Day · May 20, 2003, 11:34 PM

Reading the thoughts of so many about UNIX,BSD and AIX. I have noticed that this code is 30 years old. How does this code translate into a modern operating system?
Is the entire kernal rewritten to take advantage of the power of modern processors and such? Or is the code just added on?
Why is it so good? The coder couldn't have known about the changes of future processor designs...

Well there are two things to keep in mind:[list=1][*]UNIX was first written for a DEC PDP-11. This was a mini-computer, not a micro-computer like today's desktops and notebooks. And UNIX subsequently evolved on even more powerful hardware like DEC VAX's and IBM mainframes. It took the micro's decades to catch up to the instruction set complexity of a VAX or even a PDP-11. While today's micro's are more powerful than these machines of old in terms of speed and resources (e.g. memory and storage), the difference in CPU complexity between a VAX and a PowerPC isn't all that great; certainly nothing nearly as striking as the difference between a 6502 and a PowerPC, for example.
[*]Except for it's very first incarnation, UNIX was written in the C Programming Language and abstracted from the underlying hardware. This meant that in order to port it to a new hardware platform, all that was required is to write a C compiler for the new platform and re-write any hardware specific device drivers.[/list=1]
The maturity of the UNIX code brings with it tremendous advantages:[list=1][*]It is very well tested and debugged (which explains, in part, why the OS is so stable).
[*]Because computers of old lacked the memory and sheer speed of today's hardware, much time was spent writing code which was lean and efficient.
[*]The code has been refined over the years, and in most cases has been made feature-rich.[/list=1]

Brass · May 21, 2003, 12:26 AM

Originally posted by Rainy Day:
Well there are two things to keep in mind:[list=1][*]UNIX was first written for a DEC PDP-11. This was a mini-computer, not a micro-computer like today's desktops and notebooks. And UNIX subsequently evolved on even more powerful hardware like DEC VAX's and IBM mainframes. It took the micro's decades to catch up to the instruction set complexity of a VAX or even a PDP-11. While today's micro's are more powerful than these machines of old in terms of speed and resources (e.g. memory and storage), the difference in CPU complexity between a VAX and a PowerPC isn't all that great; certainly nothing nearly as striking as the difference between a 6502 and a PowerPC, for example.
[*]Except for it's very first incarnation, UNIX was written in the C Programming Language and abstracted from the underlying hardware. This meant that in order to port it to a new hardware platform, all that was required is to write a C compiler for the new platform and re-write any hardware specific device drivers.[/list=1]
The maturity of the UNIX code brings with it tremendous advantages:[list=1][*]It is very well tested and debugged (which explains, in part, why the OS is so stable).
[*]Because computers of old lacked the memory and sheer speed of today's hardware, much time was spent writing code which was lean and efficient.
[*]The code has been refined over the years, and in most cases has been made feature-rich.[/list=1]

Another reason why unix is still a great operating system, despite its age, is because of its modularity. I has been developed in a very modular sense (very well designed in the first place), such that if a particular element or component IS outdated, due to some security implications, or device modernisation, it can very easily be removed/replace/updated without effecting other components.

This, coupled with the open source of most unix flavours bringing a wide developer base, makes sure that even though it has very old origins, unix is and will always be a very modern operating system.

Mac OS X proves that. (Yeah, okay, Mac OS X has some drawbacks, but none that are related to its unix side that I can think of).