Open Source Community Approaches SCO

An anonymous reader writes "eWeek has an article about the open source community approaching SCO. SCO now says there are over a million lines of offending code in Linux and they still won't show them to anybody."

Comments

Over 1,000

[Marx_Mrvelous]

Over 1 mil? Does anyone know how many lines of code there are in the linux kernel?

Re:Over 1,000

[NoUse]

The funny thing is when all this started there was supposedly no code in the kernel.

Then two weeks later the same guy claims that UnixWare source "is all over the place"

While these quotes probably wouldn't be too useful as evidence in court. It does well to show that these guys are throwing darts to choose how to go forth with this legal campaign.

Re:Over 1,000

[CmdrPinkTaco]

I noticed on a site that the kernel 2.4.20 is indeed ~4.4 mil lines of code. However, what is to say that the lines SCO finds offensive doesn't extend past just the kernel. There are some that consider Linux just an operating system (kernel, drivers, file system, tools, etc) and others that consider it the whole shebang (the aforementioned plus applications designed to run on top of the operating system layer). I admittedly haven't been following the SCO drama with the greatest of detail, but my question is - what does SCO consider "Linux" - is it just the OS or is it an entire distribution (OS and apps)?

A million lines of offending code?

[Zog The Undeniable]

Wasn't it 70 lines yesterday?

Heise News shows a code:

[Anonymous Coward]

Heise News shows the code:
http://www.heise.de/newsticker/data/jk-19.0 8.03-00 0/imh1.jpg

The code seems to come from arch/ia64/sn/io/ate_utils.c, copyright by SGI:
http://www.funet.fi/pub/Linux/PEOPLE/Linus/v 2.4/pa tch-html/patch-2.4.19/linux-2.4.19_arch_ia64_sn_io _ate_utils.c.html

Does this code come from:
http://minnie.tuhs.org/UnixTree/V5/usr/sys/ ken/mal loc.c.html
http://minnie.tuhs.org/UnixTree/V7/usr /sys/sys/mal loc.c.html

Re:Heise News shows a code:

[hotair]

Am I missing something or are there syntax errors in the code in the jpeg image pointed to by the parent?
if (size==0)
return) ((ulong_t NULL);
...

In which language does this compile?

Re:Heise News shows a code:

[msgmonkey]

You can download the source tar file from http://public.planetmirror.com/pub/ancient-unix/an cient/sysIII/sys3.tar.gz

The file you want is usr/src/uts/pdp11/os/malloc.c

first few lines are:

#include "sys/param.h"
#include "sys/systm.h"
#include "sys/map.h"

/*
* Allocate 'size' units from the given map.
* Return the base of the allocated space.
* In a map, the addresses are increasing and the
* list is terminated by a 0 size.
* The core map unit is 64 bytes; the swap map unit
* is 512 bytes.
* Algorithm is first-fit.
*/
malloc(mp, size)
struct map *mp;
{
register unsigned int a;
register struct map *bp;

The rest is identical.. the only question is what is the legal status of this code? Is it in the public domain?

Heise News shows code

[Anonymous Coward]

Heise News shows code:
http://www.heise.de/newsticker/data/jk-19.0 8.03-00 0/imh1.jpg

The code seems to come from arch/ia64/sn/io/ate_utils.c, copyright by SGI:
http://www.funet.fi/pub/Linux/PEOPLE/Linus/v 2.4/pa tch-html/patch-2.4.19/linux-2.4.19_arch_ia64_sn_io _ate_utils.c.html

Does this code come from:
http://minnie.tuhs.org/UnixTree/V5/usr/sys/ ken/mal loc.c.html
http://minnie.tuhs.org/UnixTree/V7/usr /sys/sys/mal loc.c.html

Re:Heise News shows code

[mocm]

And did you read the "encrypted" text below the commentary (greek letters) on the Unix side:
As part of the kernel evolution towards modular naming, the functions malloc and mfree are being renamed to rmalloc and rmfree. Compatibility will be maintained by the following asembler code: (also see mfree/rmfree below)

There is not even C code following. So SCO copied the commentary and than put something else in the file.

SCO showing portions of code at "SCO Forum"

[Markos]

According to heise SCO is showing portions of offending source code. Here is an image of the offending code, and here is another image. From the Gentoo Forums:

There seems to be nothing of value here. As pointed out by Starborn this comment goes at least as far back as BSD 2.11, which, according to the file, is from subr_rmap.c 1.2 (2.11BSD GTE) 12/24/92

...

The linux version is slightly modified with some differing variable names, but the algorithim seems to be nearly identical.

Basicly, the code they've showed goes as far back as 1992 from BSD 2.11, perhaps even further?

Hypothetical Scenario

[Anonymous Coward]

to quote this piece from the article:

"The current NDA is completely unreasonable and will prejudice the ability of any of us to do our job going forward if we sign it. We feel that SCO needs to be reasonable about granting access to the offending code."

Now imagine SCO actually makes few people sign NDA and view the code, but the viewed code in question is safe. One of these renegades decides to take a bullet for the team and makes the information public. Days later, OSS coders rewrite the "offending" parts in the linux kernel to rid it from alleged IP. They release the code, all hell breaks loose... you get the point.

Is this something that's a possibility? SCO isn't playing fair. You come to expect the most evil acts from these guys.

Isn't samba GPL

[nuggz]

How can they distribute samba?
It is GPL, and they are arguing it is an invalid license.

Unless the agree to the GPL, they can't distribute samba. Isn't this a stupid strategy?

Here is the linke to the story, look at the bottom of it.
http://news.com.com/2100-1016_3-5065286.html? tag=f d_lede1_hed

BSD?

[hughk]

If this code was in BSD, wasn't it explicitly covered by the settlement with AT&T?

Note that the AT&T version of this code is also probably old enough to appear in the Lyons book. Whilst that doesn't do anything for copyright, it sure nukes the idea of a "Trade Secret".

C'mon, SCO, Show SOME lines

[Esion Modnar]

I mean if you have a million lines of your code in there, what's a mere hundred lines? Or ten?

That way, even if everybody ran off and fixed those lines, you still have well over 900,000 lines of evidence (according to you) in your back pocket.

And you would gain (maybe) some credibility. Not to mention what it'd do to your stock price.

Contract issue lines , not copyright lines

[LightSail]

Darl is using a smoke screen: the million lines are not SCOX copyrighted, They were Sequent and IBM copyright. SCOX has a small chance of proving contract violation, but cannot prove copyright claim to Sequent/IBM code. Even winning the lawsuit does not give SCOX the copyright status to the donated code. The SCOX copyrighted code that may be in the Linux kernel is most likely either BSD, previously published algorisms, an open standard or hardware vendor release. These are not exclusive to SCOX. If SCOX had actual hard evidence, they would be using it to prove the point with a small public display. Any public display of code would easily be connected to the actual source outside of SCOX, so no show without nda. My feeling is that none of Linux is an actual copyright violation of SCOX copyright. This entire situation is a Hail Mary to increase market cap of SCOX, then use that value to buy profitable companies. Once SCOX has converted the inflated stock into additional sources of income, they will settle and rename the company to reflect its new identity. They may even end up by donating UNIX to open source to make amends, once they have milked the lawsuit and publicity for every dollar possible.

Slides code won't even compile...

[Spam.B.gone]

But the ')' after the return statement (2nd example, 2nd line) will prevent compilation of this code...

SCO code shown = BSD

[Anonymous Coward]

SCO did a big dog and pony show yesterday at their conference. They also raised their claims, and according to them, there are "millions of lines" of offending code which they have identified by "rocket scientists" using "spectral recognition" and "pattern analysis".

To convince SCOforum attendees of their case, SCO showed obscured slides which supposedly proved copying.

Research reveals that the code fragment SCO showed in one of their slides, doesn't even belong to SCO - it's from BSD. See for yourself, the code originated from, and is Copyright 1986 Regents of the University of California! And, while they might have more up their sleeve, it's is revealing that the most compelling example they can show at their forum, doesn't even belong to them!

License change

[coyote-san]

I'm not sure how many contributers there are to the Samba project, but it's almost certainly far fewer than to the Linux kernel, GNU tool chain, etc.

The bottom line is that they may be able to take direct action: change the license to "GPL-SCO." That's a stock GPL license with an extra clause superceding all others and explicitly prohibiting the use of the software on SCO systems, on any system owned by SCO regardless of the OS used, or distribution in any form by SCO or its successors. Finally, since SCO is claiming that none of these licenses are valid anyway there would be a final clause inserted by the lawyers that basically say that if the rest of the license is invalidated then SCO owes a licensing fee of US$1,000,000,000 per CPU, payable immediately. A billion dollars/CPU to the people who actually wrote the code is no less unreasonable than SCO trying to collect a kilobuck/CPU from Linux users who never invited SCO to the table.

In short, if they want to support MS products but refuse to accept the standard license, they can damn well write the code themselves. The same applies to any other application they use.

This is a bit more direct that what the GCC group is supposedly considering - dropping SCO hardware from the list of supported hardware - but it's clear that SCO isn't going to stop until the feds get off their ass and start prosecuting these clowns.

SCO has become "those guys"

[raw-sewage]

What baffles me is that there seems to be some folks who are taking SCO seriously (not the open source community, of course). Why can't the general public see SCO for the ridiculous group they are?

If they really wanted to be compensated for this "stolen code" or "contract infringement" or whatever complaint-of-the-week they may have, wouldn't they be pushing this harder through the official channels? Anyone who is wronged wants a quick resolution. SCO seems content to just sit around and release a lot of anti-open source press.

The overwhelming majority of their claims has been debunked, except that which can't be seen (the "offending code" and the contract with IBM). I doubt an entire million lines of code is stolen. No way! Sometimes I think that SCO may actually believe that there is stolen code in Linux. Somewhere, someone made some assumptions, and used these assumptions to demonstrate that Linux uses stolen code. And now they seem to have people working off of these assumptions, finding even more illicit code! I can see where there might be a few shady or questionable areas of the kernel, but a million lines? Pshaw!

Every week or so, SCO seems to come out with a more outrageous claim. Sometimes I think that maybe they are playing the open source community against itself: there's so much talk about the legality of the GPL, SCO's claims, intellectual property, etc. SCO at least knows how passionate the Linux community is about its operating system; they are using this fact to keep their smear/FUD campaign alive. SCO is just a school yard bully. This whole affair has become the reality TV of the open source world.

I don't see it happening (as I myself am just as guilty of wanting to know "what's next" with this ordeal), but we need to just ignore SCO. And by we I mean everybody who is not SCO. Until they are willing to actually work with anyone (outside of an NDA or the courts), there's nothing we can do except keep making better software.

Of course, the other side is that there's always an argument for "any publicity is good publicity". If a few more people know about Linux/open source than who are turned off by the propaganda, it works to our benefit.

Comment lines could be damning

[pcause]

A number of articles on this issue point out that many copied lines are comments and that the comments have the identical spelling errors that lines of *nix code do. If this is true, and if the lines of code are truly owned by SCO this is very damning.

Someone might make the case that given a task to do the code to accomplish the task could look very similar or nearly identical. Bit comments? And spelling errors? Not a chance. Comments can be rare enough and programmers idiosyncratic enough that it stretches credulity to think that multiple programmers would write the same comment with the same spelling error.

The issue is whether or not these lines came into UNIX from another source, such as from BSD. If the code came from BSD, perhaps TCP stack or utilities, than SCO really has no claim. The other possibility is that it is in driver code. This gets murkier.

I don't think AT&T ever made the APIs for drivers public. You had to have a non-disclosure agreement with them or a license. But it is possible that you could replace the AT&T interfaces with Linux interfaces and had the code look identical in 90% of its content and not be a copy, since the device parts would be identical, but they would be owned by the driver writer. One exception is if the driver writer started with an AT&T driver and modified it. In this case, SCO wins.

There was a 386 reference port of UNIX done for AT&T by Intel and Interactive Systems. As part of that port there were a number of driver provided to AT&T. They are all owned by AT&T and drivers that were built starting with those drivers would be a violation of the license. One itneresting fact is that Interactive went into the packaged UNIX business and their x86 UNIX was eventually bought by Sun and was the base of Solaris for the x86.

Again, it all comes down to the details: which parts of the code are we talkign about.

Re:Corporate Death Penalty

[Zocalo]

In the US maybe, but not here in the UK where we really do have a corporate death penalty. It's actually called a "Compulsory Winding Up Order" and can be issued by a court when a company if found to have sufficiently dubious trading/financial practices to make it a liabilty to do business with them.

Besides, who do you think can buy the most law makers - IBM or SCO?

Re:Corporate Death Penalty

[Lord_Dweomer]

"The problem is this: there is no downside for SCO... they can say and do anything without fear - and there's the very remote chance that they might win something. It's like buying lottery tickets."

No, there IS a downside. Its called the SEC. Trust me, once they get involved, you will see one of the most frantic backpeddlings in history. Unless of course Darl and his crew think they can pay some big fines and walk away with several million.

Anyone else get the feeling...

[Rogerborg]

That this is starting to read like a The Onion article? You know, one of the recent ones that starts with a moderately amusing title - "SCO claims 'All your code base are belong to us'" - but then just dribbles on and on until you get tired of reading it.

SCO are trolling for dollars. We should stop helping them out by disseminating their bullshit. We shouldn't even bother to refute it, because by doing so, we make it looks as though there's something there that needs refuting. Nuh huh. Until they back up their claims by listing the source, there is no story here. They're simply begging for publicity to sell shares to pointy haired morons. Let's not be a party to that any more.

SCO reminds me of Senator Joe McCarthy:

[Futurepower(R)]

From the Slashdot story: "SCO now says there are over a million lines of offending code in Linux and they still won't show them to anybody."

This reminds me of Senator Joseph McCarthy's 1950 Communists in the State Department speech. See the end of the article for a quote from Senator McCarthy:

"I have in my hand fifty-seven cases of individuals who would appear to be either card-carrying members or certainly loyal to the Communist Party, but who nevertheless are still helping to shape our foreign policy."

Senator Joseph McCarthy said he would show Dean Acheson, then U.S. Secretary of State, the list, but only under special conditions that often changed. McCarthy said: "It would be a waste of effort to give Acheson the names, then have him deny they are Communists and we can not get the records."

The number of Communists McCarthy said were in the U.S. State Department also often changed, too. Soon it was "81 subversives":

The article cited above says, "Senator Lucas of Illinois, Democratic leader, repeatedly tangled with McCarthy, who also said he has case histories of 81 subversives--including what he called a 'big three'--who are working in and with the State Department. Lucas challenged McCarthy to name names. McCarthy refused, saying Lucas or any other interested authorities could get the names at McCarthy's office."

"The Senate voted 67 to 22 to censure McCarthy" (See the end of the article.) "Senator Joseph Raymond McCarthy died soon after the censure, at the age of 48, of hepatitis and liver disease related to alcoholism."

Senator McCarthy gave many people a big Red scare. However, in the end, everyone realized that he was a liar.

SCO's Got Copyright Backwards

[reallocate]

I'm pretty sure SCO is blowing smoke, but keeping their code under wraps until they're in court seems to be a reasonable strategy to me. Their success depends on convincing the court, not the open source community.

That said, I'm also pretty sure that SCO is dead wrong to argue that copyright law prevents others from making copies of your work. Copyright law protects an author's rights, which include establishing the conditions under which others can acquire copies of that author's works. Commercial software sells you a copy; free software gives you a copy.In both cases, rights to own a copy are transferred by the author to someone else. Whether or not payment was received is of secondary importance. In either case, copyright law protects the author.

A license is a different animal entirely, and it will be interesting to see ow the GPL fares in court.

Why no individuals will sue SCO...

[rongage]

To me, the answer to why nobody within the developer community (including myself) is bothering to sue SCO for copyright infringement and GPL licensing violations is quite obvious.

As developers, we have barely enough time in our busy lives to deal with things like writting our next "killer app", let alone things like family and living. On top of this, to sue such a company requires fronting a fairly large amount of cash to a lawyer. Nevermind the level of trust and respect our profession seems to harbor towards lawyers in general.

We all would love to see "someone else" sue (and obviously win) a suit against SCO. The problem is that "we" don't have the time nor the money to defend what is effectively ours. Hence, we will loose our creation(s) because of our lack of willingness to go after the bully.

Like everyone else, if I had the time and the money, I would be all over SCO for violating my copyright and the GPL license. I just can't afford the risk financially right now...