What is SCO's legal theory? What exactly is SCO claiming that IBM did wrong? It's fairly simple.
SCO claims that IBM once wrote some original code, which was placed next to some SCO code. Later, IBM took that same original code and put it in Linux. Although no SCO code wound up in Linux, the IBM code was infected with SCO's intellectual property rights due to sheer proximity, and the infection spread through Linux. SCO is claiming that Linux has SCO Disease.
This really is SCO's legal theory. My only contribution is the name "SCO Disease". Mark Heise of SCO's lawfirm Boies, Schiller, and Flexner explained it at length in his opening statement to the judge on February 6, 2004, although it was already apparent from many of SCO's previous statements. SCO has invented a new type of intellectual property for itself, one that trumps copyrights, patents, trademarks, contract law, and trade secrets.
This is why SCO can't show specific lines of System V code in the Linux kernel. There aren't any. What SCO is claiming is that it has an entirely new type of intellectual property, one that rubbed off from the System V code onto the AIX code without leaving any directly observable traces, and then spread from that AIX code throughout the Linux code base. Even SCO can't see SCO Disease by looking at the Linux code, or the AIX code. SCO can only deduce the presence of SCO Disease by seeing AIX code and System V code together, in historical proximity in the long-forgotten AIX releases they've requested. This is why they've demanded copies of every AIX version ever produced, and complain that IBM must produce them before SCO can properly present its claims.
It is literally SCO's position that adding IBM original work to Linux is an infringement of SCO's intellectual property rights, due to the IBM work's contamination with SCO Disease. Every time IBM insists that SCO show which parts of System V wound up in Linux, SCO counters that the infringement SCO is complaining about is that AIX code wound up in Linux, although SCO can't prove it because they don't have copies of AIX. IBM wrote code which IBM never showed to SCO before contributing it to Linux, and SCO considers that to be an infringment. This is SCO Disease.
SCO Disease is so virulent that SCO claims it has spread out from IBM's contributions to saturate the Linux kernel, so that now even removing IBM's contributions wouldn't remove the SCO Disease. For example, the four significant technologies SCO claims IBM contributed to Linux (SMP, RCU, JFS, and NUMA) aren't even used on most of the systems SCO is demanding licenses for. SMP, RCU, and JFS only apply to multi-processor systems, and JFS is a filesystem that is virtually unused outside of IBM (everyone else uses EXT3, ReiserFS, or occasionally SGI's XFS). Therefore, the vast majority of Linux installations do not actually benefit from any of these "enterprise" technologies, yet SCO is demanding $699 single-processor licenses from those users anyway.
Hold tight, it gets even stranger.
SCO Disease originated from an obvious misinterpretation of an old AT&T contract. The first hint of the SCO Disease theory came in the 2003 SCO Forum presentation, page 6, where SCO quoted the following paragraph of an old AT&T license from 1985:
AT&T grants to LICENSEE a personal, nontransferable and nonexclusive right to use in the United States each SOFTWARE PRODUCT solely for LICENSEE's own internal business purposes... Such right to use includes the right to modify such SOFTWARE PRODUCT and to prepare derivative works based on such SOFTWARE PRODUCT, provided the resulting materials are treated hereunder as part of the original SOFTWARE PRODUCT. (Para 2.01)
At first glance, this seems to mean that AT&T was trying to make a copyright license that acted a little like the GPL, requiring that derived works be distributed in their entirety under the original unmodified AT&T license. This raises several points.
First, SCO has attacked the GPL as being unconstitutional. For SCO to base their entire case on a similar "viral" clause of an old AT&T license would tend to undermine that argument. (And the GPL's language is much clearer as to its intent; if AT&T was indeed trying for a viral clause they didn't do nearly as good a job as Richard Stallman did.)
Second, this talks about the licensing of derived works, which only applies to works created by combining IBM's code with System V code. Works created solely from IBM's code, which do not contain any SCO code, are not derived works of System V. In fact, Amendment X explicitly shows that AT&T agreed this was the case, even though it was obvious anyway. IBM was being extra careful to get stuff down in black and white, in case of future unpleasantness. (Amendment X wasn't the only time AT&T clarified the meaning of its license, as Novell has recently pointed out.)
Third, copyright holders may issue more than one license to a given copyright. The dual-licensed QT library from Trolltech is one of many modern examples. AT&T itself issued thousands of different licenses to the System V code. Two such licenses are at issue in the SCO case: SCO is trying to claim IBM's Amendment X doesn't apply to Sequent's System V license (even though Sequent's System V license was not transferrable, thus probably terminated when the company was sold to IBM; See Para 7.09 of the same contract on page 7 of the SCO Forum presentation). So SCO's claim to possibly have a license to some of IBM's code doesn't prevent IBM from issuing additional licenses to that code, with different terms.
Fourth, this contract section is about AT&T putting conditions on the license it granted IBM to use the System V code, when that System V code is used to create derived works. It says nothing about IBM granting any kind of rights to AT&T. IBM did not grant AT&T a license to use any IBM code. During the USL vs BSD lawsuit, AT&T's successors got in trouble due to the unauthorized use of Berkeley code in AT&T's Unix products. The judge did not believe that AT&T's contracts granted it blanket licenses to use code from other companies derived works in System V. The transfer of intellectual property went from AT&T to IBM. There may have been some minor restrictions on the use of that IP, but there was no grant of IP rights in the other direction, from IBM to AT&T.
Therefore, SCO does not have a license to any IBM code. The most obvious indication that the System V contract was never meant to grant AT&T or its successors any rights to IBM code is that the contract provides no delivery mechanism whereby IBM would deliver the source code to derived works back to AT&T. SCO literally claims control over code they've never even seen, and that the most generous possible reading of their contracts never require IBM to even show them, under any circumstances. This is why their frantic discovery requests for AIX and Dynix code, and their insistence that they can't state their case until IBM produces its own source code for them to look over. They haven't got it, have never had it, and IBM has never been required to give it to them.
If the contract meant for SCO to have even a simple license to use the IBM code in its own propreitary Unix products, let alone more sweeping unspecified intellectual property rights, the contract would have granted them access to it. It did not. How can it possibly have meant to give them any kind of rights that it gave them no mechanism to enforce? Can any contract that requires a lawsuit and discovery requests to figure out whether or not the plaintiff's rights have actually been infringed possibly be a valid contract?
Obviously, this paragraph doesn't mean what SCO would like it to mean, but reality has never been on SCO's side in this. Nevertheless, on this foundation of wishful thinking, SCO has built an enormous edifice of delusion, and is pressing it in court. They've promoted their newly invented category of intellectual property, SCO Disease, as more powerful than any existing type of intellectual property; including copyright, patent, trademark, contract, and trade secrets. Let's go over each one in turn.
SCO has admitted that IBM still owns the copyrights to all the code IBM developed, including the AIX code and the Dynix code, and IBM's contributions to Linux. As mentioned in the previous section, SCO does not even have a valid license to any of IBM's AIX or Dynix code, nor do they have copies of it or any agreement with IBM that requires IBM to give them copies (except as part of the discovery process in the current lawsuit). Yet SCO claims that it can control IBM's copyrighted materials, preventing IBM from issuing new licenses to this material that SCO does not have a license to either. (Apparently, SCO believes that nobody can validly have a license to IBM's original AIX technology, that IBM's original work must be kept off the market entirely.)
SCO is also claiming not to be bound by another copyright license, the Free Software Foundation's General Public License (GPL), despite SCO's history as the Linux distributor Caldera, and the basis of their own case on a similar "viral" clause in the AT&T contract that makes the AT&T license act like a pale immitation of the GPL.
For most of a decade SCO's core business was developing and selling Linux, under the GPL. Their Linux-based IPO even included a copy of the GPL as Exhibit 10.14 attached to their S-1 filing submitted to the Securities and Exchange Comission, and the proceeds from that IPO were what funded their purchase of the Unix business of the Santa Cruz Operation. The fact that Caldera was dumb enough to buy what the original SCO wisely wanted to sell was their own fault. Appendix A of the Halloween 9 document traces The Decline and Fall of Caldera, along with Caldera's original reasoning for buying SCO, and the changes it underwent as its business model failed.
Caldera distributed Linux under the terms of the GPL (selling it as a retail product with the GPL as the license between themselves and their customers), and developed and contributed a significant amount of original code to Linux under the terms of the GPL. Caldera co-founded the UnitedLinux consortium, and contributed extensively to the Linux Standards Base, but now it wishes not to be bound by the terms of the Linux license, the GPL. (Despite this, it continues to distribute Samba as part of its own proprietary Unix offerings, and Samba is licensed exclusively under the GPL.)
Comparing SCO Disease with the GPL specifically is worth a closer look. The GPL doesn't prevent dual licensing by the copyright holder. The GPL doesn't apply to code that isn't a derived work. You can't GPL code you don't own; license violations by licensees mean the resulting derived work cannot be legally distributed, not that third parties can change the license terms of snippets they don't own the copyrights on. Code never "becomes GPLed" by the actions of someone other than they copyright holder; copyright law doesn't work that way. And removing the offending code is widely accepted as sufficient remedy for unintentional infringement of the GPL, since the result is no longer infringing and distribution of offending infringing works ceases. There are many limitations which apply to the GPL, but none of these limitations seem to apply to SCO Disease.
More generally, copyright lawsuits require showing specific infringment. Although SCO has claimed "millions of lines" of infringing code, it has been unable to show examples of its copyrighted material in the Linux tree. Eric Raymond's comparator checked several versions of the Linux kernel against the System V kernel, and found no matches (other than one widely-publicized historical instance of code System V shared with BSD, which that had already been removed as unnecessary before it was even identified by SCO, and was traced back to SGI, not IBM). If SCO can't show misappropriation of code it owns the copyrights to, SCO can't pursue a copyright case. SCO repeatedly points to AIX and Dynix code as "infringing", but since SCO is not claiming to own the copyrights to AIX or Dynix code, those cannot be copyright infringements.
To top it off, Novell claims to own the System V copyrights. The original 1995 relationship between the first SCO (from which Caldera purchased the already failing Unix business in 2000) and Novell had SCO acting as an agent for Novell, collecting Unix royalties from existing System V licensees and passing 95% of those royalties on to Novell, keeping only 5% as an agent's fee. Many of these old customers (including IBM) long ago chose to buy out their license obligations for a one time fee, hence IBM's current "irrevocable, perpetual" license to the obsolete rubbish from the 1980's that is System V.
The original SCO never asked for the System V copyrights to be transferred to it while it was acting as Novell's agent, and the current one moved to claim these copyrights only after filing suit against IBM. (Oddly, by asking Novell to transfer them and then deciding it didn't need Novell's permission and unilaterally registering them.) Novell is contesting the matter in court, and among other things has pointed out that on a major change of ownership of SCO (such as its acquisition by the Linux distributor Caldera) many of the rights that SCO acquired from Novell would revert back to Novell.
But the most damaging things to SCO's case remain its own actions. As a Linux distributor, one of Caldera's first acts after acquiring SCO's Unix business was to release older versions of propreitary Unix under an open source license. (Their reasons for doing so are interesting as well.) In addition, Caldera repeatedly announced the intention to "unify Unix and Linux" as its rationale for purchasing Unix, and went on to release new Linux products (under the GPL) even after filing suit against IBM, such as the April 19, 2003 release of SCO Linux 4.0, using the 2.4.19 Linux kernel. It also partnered widely with other Linux companies right up to the end, including IBM, leaving itself open to questions of promissory estoppel.
By imposing additional restrictions under the GPL, SCO has terminated its own right to use GPL code. This might affect SCO's right to ship other GPL products, such as Samba, but this has no effect on other people's use of GPL code. Third parties who agreed to accept additional restrictions from SCO, such as accepting a SCO license to use Linux, might similarly place themselves in conflict with the GPL, but it would be the third parties' own actions that did so, not the actions of SCO. Only the copyright holder can set the license terms, what licensees do cannot change the terms of the license because they don't own the copyrights. All they can do is comply or infringe, not unilaterally relicense the code without the copyright holder's permission.
To pursue any sort of copyright strategy, SCO would have to invalidate its own development and distribution (under the GPL) of versions of Linux that already contained the code it objected to; not just as updates, but as the core of entire new product releases whose now-disputed "enterprise scalability" features they explicitly bragged about at the time. SCO would also have to prove that its earlier releases of "Archival" Unix didn't contain and grant permission to the disputed material. And SCO would have to diprove Novell's claims to the Unix copyrights. Then SCO would have to undermine SuSE's blanket cross-licensing agreement from UnitedLinux. It would have to rule out independent development of similar code, and would have to show that the BSD settlement neither left them without an enforceable copyright on their current derived works (unlike in the Ashton-Tate case), nor provided a valid alternate route for any disputed Unix technologies to show up in Linux. And finally, it would have to show the presence of actual infringing code, which many other System V licensees (including IBM and SGI) have been unable to locate.
Clearly, SCO Disease would have to trump copyright entirely in order to exist.
This one's easy: SCO simply hasn't got any patents. Although SCO threatened to use Patents early on, it turns out that Novell retained the Unix related patents when it delegated its Unix business to SCO.
IBM, on the other hand, has tens of thousands of patents. At the start of its countersuit, IBM pulled out four patents that SCO's entire product line infringed upon, just as an opening salvo. SCO has claimed that the four IBM patents were fraudulently gained and therefore invalid, but has been unable to specify exactly what fraud it meant when it said that. In any case, IBM has thousands more patents to choose from in future, gains thousands more every year, and can continue to countersue SCO (on patent grounds alone) forever.
IBM even has patents on one of the technologies SCO explicitly disputes: RCU. SCO never objected to the filing of this patent (it's highly unlikely they were even aware of the existence of the technology back then), and obviously SCO could not have developed that sort of technology themselves. (IBM has designed and manufactured its own supercomputers for half a century now, while the best selling SCO Unix product, in terms of gross revenue is OpenServer -- an ultra-low-end unix primarily used to run fast food cash registers.)
SCO is trying to tell IBM that IBM can't use its own patented RCU technology without SCO's permission. Usually, patents work the other way around.
Patents also have limits that SCO disease doesn't suffer from. Patents apply to a process, not a concept. If you can think of a different process to accomplish the same goal, you're not infringing upon the patent. This is why you can't have a patent on something as vague as "SMP" or "NUMA", a patent would read more like "a method for performing SMP", and there could be many other non-infringing ways of accomplishing the same goal.
SCO is claiming that SCO Disease allows it to ignore its own infringement of the patents of others, control others' ability to use their own patents, without suffering from any of the limitations of patents. Clearly, SCO Disease is a much more potent form of intellectual property than mere patents.
Trademark law is only peripheral to the case, but clearly SCO thinks SCO Disease beats it. To start with, Novell never gave SCO the Unix trademark, granting it instead to The Open Group. SCO itself can only use the name "Unix" for its products with the permission of The Open Group. SCO claims carte blanche ownership of Unix, yet doesn't even own the Unix name.
Meanwhile, Linux is a trademark of Linus Torvalds. SCO's own products, such as the "Linux Kernel Personality" of UnixWare 8, do not acknowledge Linus's ownership of the Linux trademark. UnixWare's Linux kernel Personality is in an interesting position: if it doesn't contain significant portions of the Linux kernel it's in violation of the Linux trademark. If it does contains any of the Linux kernel's code, it's in violation of the GPL. Either way for SCO, this is a lose-lose scenario.
SCO is completely ignoring the trademark issue, which is subsumed in passing by their SCO Disease theory.
If SCO's complaint really was a contractual dispute with IBM, its only recourse would be against IBM. Of course SCO has chosen to ignore the text of the contracts it inherited, such as IBM's Amendment X, the Sequent termination upon transfer of ownership clause, and the "perpetual", "irrevocable" nature of IBM's later buy-out of that contract.
SCO Disease itself derives from an obviously nonsensical misreading of a 20 year old contract, which probably wouldn't survive scrutiny AS a contract even if it did clearly state what they seem to think it does. Among other things, it would then be effectively a convenant not to compete, and open to questions of reasonableness. But that's a moot point since it clearly doesn't say what they think it says; if it did it would have to have provisions for delivery of code from IBM to SCO, and it doesn't. SCO is interpreting its contract to give it control over code that the contract does not even grant it a copy of. That CAN'T be the original intent, or there would have to be some kind of auditing mechanism, and there isn't. The contract would have specified delivery of copies so SCO knew what it was that it had rights to, if what SCO thought it meant was what it actually meant.
Meanwhile, there are alternative interpretations of this contract that DO make sense, such as the obvious one that the non-disclosure provisions on the System V code apply to the portions of System V code included in derived works, and thus the same care should be taken with the source code of derived works that contain any System V code. SCO's interpretation granting it rights to the moon and the stars is by no means the only interpretation of this contract.
But SCO has targeted a much wider group than IBM: the entire Linux using world. The genesis of its case was really sour grapes: it had failed to make money or market share as a Linux distributor where others had succeeded, and if it couldn't profit from Linux no one would. IBM simply had conveniently deep pockets; the SCO Source initiative started by demanding (and receiving) tribute from companies like Sun and Microsoft that had no (positive) interest in the future of Linux.
SCO has recently sent 6000 letters to entities that once had a contractual relationship with it (or its predecessors), somehow claiming that IBM violated THEIR contracts as well. (So far, none of them seem to have been stupid enough to believe that signing another contract with SCO would somehow improve matters.) Clearly, having a contract with SCO doesn't provide any defense. Having anything to do with that tar baby simply makes one more of a target. When a company has no ongoing business model except litigation, there's no such thing as a "one time payment". Once they've spent it, they'll be back for more, using an excuse just as trumped up and flimsy as their current batch. (This is apart from the fact that accepting such a license might violate the GPL in and of itself, jeopardizing your own right to use the software anyway by accepting the additional license terms the GPL forbids.)
But SCO has threatened a far wider group, the entire Linux community. It started by sending letters to 1500 Fortune 500 companies, most of whom were not Unix distributors. One company it is reported to be in negotations with is Google, a company that has been pure Linux from its founding and never needed or wanted to use AT&T's obsolete proprietary Unix technology, even long-removed derivitives of it licensed through third parties like Sun or SGI. But you don't have to have any contractual relationship with SCO for SCO to come after you, and no contract will protect you from SCO.
SCO even sent one of those 1500 letters to Novell, the company from which it had licensed the Unix code in the first place, a company to which SCO has paid System V related royalties for years. Novell claims that the contract it has with SCO grants it veto power over SCO's entire Unix-related litigation campaign, but like all other inconvenient contract provisions, SCO has ignored this.
SCO is bound by many other contracts dating back to its time as a Linux distributor. UnitedLinux is one example; SuSE (now owned by Novell) claims to have a blanket license to all SCO/Caldera code as part of the UnitedLinux initiative, allowing SuSE to GPL any of it at will. When this kind of thing comes up, SCO just ignores it.
Fundamentally, contracts are based on informed consent. Parties must knowingly agree to the contract for it to be binding. But nobody wants to use SCO's code in Linux, and would happily remove it if they could. SCO has explicitly rejected such removal (quoting SCO CEO Darl McBride, "The Linux community would have me publish it now, (so they can have it) laundered by the time we can get to a court hearing. That's not the way we're going to go.").
Since allowing the community to remove the code is something SCO's CEO explicitly wanted to prevent, and assuming for the sake of argument that SCO might actually have had any code to infringe, this would mean that SCO was forcing people who didn't want to to use its property to do so. SCO's behavior borders on chasing down fleeing non-customers and ramming its product down their throat in order to present them with a bill afterwards. If there's never any opportunity to refuse, how can you have informed consent?
All this means SCO must think that SCO Disease trumps contracts.
This one is more "SCO Disease vs Logic", but that's a persistent theme. SCO cannot diagnose SCO Disease, even under its own twisted theory, unless it can show AT&T code and IBM code side by side, and later show that same IBM code side by side with Linux code. This SCO has been demanding to receive copies of IBM code that its contracts never require IBM to provide it. IBM can rightly claim that that code has indeed been kept secret, and that IBM has the right to continue to keep it secret -- from SCO. Yet SCO keeps demanding that the court hand it over, and claims it can't present its case properly until IBM tells it what its case is.
Of course SCO's trade secrets continue to be secret, even after SCO published the code in question under the GPL, up through Linux 2.4.19. SCO's behavior here is now more extreme than the much-maligned Recording Industry Associaton of America; even the RIAA has finally admitted that DeCSS is not a trade secret. (After placing DeCSS in Hoy Exhibit B, any hopes of that defense evaporated for the RIAA and the DVD Copy Control Authority. They themselves had disclosed DeCSS to the world; whatever its status before that, it was no longer a trade secret.) SCO did more than just "leak" Linux code: SCO shipped the Linux kernel (with full source code, under the GPL) to customers as its core business for many years, and hired a number of full time employees to upgrade and modify the Linux kernel, and issued press releases and web pages bragging about it.
SCO has claimed control of publicly documented technologies (including things in Posix and the Single Unix Specification). Technologies that SCO itself didn't invent, yet somehow are SCO secrets. SCO has even claimed as its secrets things that SCO itself doesn't have access to, yet which it somehow still claims to control, even though it doesn't have copies of them and isn't sure what's in them.
SCO Disease covers things that are secret from SCO, and unlike trade secrets, SCO Disease never expires. The only thing that has to remain a secret to function is SCO's evidence.
So SCO believes that it doesn't have to prove its claims before demanding to be paid, that it can demand money from people who actively do not want to use its intellectual property, from people who do not even use the contributions to Linux that SCO has identified, from entities that SCO has never had any contractual relationship with, that SCO is not bound by its own previous contractual committments or its own past behavior. The only secrets it is protecting are the specifics of its claims of infringement, but pursuing such claims should give it carte blanche access to anyone else's secrets. And due to its virulent ability to spread to any code it comes in contact with, not only does SCO's code not even have to be present to be infringed upon, neither does IBM's.
There are dozens of other things wrong with SCO's case, all solved by SCO Disease. SCO doesn't seem to think it needs to mitigate its damages, and can intentionally force people to infringe, or drag out the process with delays and missed deadlines in an attempt to get more money. (To again quote McBride, "Every day they don't resolve this, the AIX meter is still ticking... We're sort of fine to let the AIX thing tick, because the longer it goes, when we actually end up in courtroom, we can go back to June 13, 2003, and add damages.".) SCO's history as a Linux developer, Caldera's claims that it was buying SCO's Unix division to "unite Unix and Linux", its distribution of the Linux kernel to its customers under the GPL, and its previous claims that backtracing to the 2.2 kernel (which it helped develop) would be a remedy against infringment, might be considered prommissory estoppel if it wasn't for SCO Disease's ability to go backwards in time and cause retroactive infringement. (Backtracking to the 2.2 kernel was once SCO's suggested remedy, now SCO claims the 2.2 kernel is also infringing.) SCO's repeated amendment of its claims to be based on completely different theories, and constant missing of deadlines when asked to show actual evidence, well who needs evidence when you can unilaterally invent an entirely new category of intellectual property?
A tiny little company (Caldera) bought some ancient obsolete software that a series of companies (AT&T, Novell, SCO) all wanted to get rid of it for successively smaller amounts of money, and now they're claiming that the thing they spent $36 million on (claiming that what they were buying was not the software but a distribution channel they squandered and destroyed through sheer incompetence) is worth literally over 100 times that in damages from IBM alone, plus who knows how much from all the other companies they're suing. Caldera has already made back in (gross) revenues from their SCO Unix lines more than they paid for the SCO acquisition. They destroyed their own core Linux business along the way through sheer management incompetence. (Again, see The Decline and Fall of Caldera.) How can Darl (who cheered Linux on when he first took over as Caldera's CEO) now claim to have been harmed by anything but his own profound incompetence?
There's no argument that thousands of independent developers are trying very hard to make an independent, non-infringing linux kernel, and have spent over ten years working to create such a thing from scratch. But SCO is arguing that it's not possible to do so (at least it's not possible if there's no money in it for SCO). SCO is arguing that it owns the idea of Unix, and that the Linux kernel developers' work was contaminated (against the wishes of any developers who do wish for an independently created, non-infringing work) and that removing that contamination is not possible. SCO is arguing that there's a massive conspiracy against it, requiring intentionally infringing actions from thousands of companies, people, and organizations. And that this, not incompetence, is why Caldera lost money every year since it was founded and repurposed its business based on a technology three consecutive companies were willing to sell for less than they acquired it for.
Nobody wants SCO code. The linux developers certainly don't want it in Linux, they've offered to remove it if SCO can ever actually identify any.
What SCO wants is to be paid for every sale of Linux. Specifically, they want to profit from other people's code, since all of SCO's code is available in SCO's own products, which nobody wants to buy. SCO isn't doing this to increase sales of their own products, and it isn't doing it to stop Linux from using SCO intellectual property. SCO wants to consume Linux, and profit from the work of millions of Linux developers who have created a system that long ago rendered SCO's own offerings obsolete and unsellable. This is the reason for SCO Disease: SCO wants money it has failed to earn.
I could go on, about the AT&T vs BSD lawsuit and the possible relevance of the lawsuit that brought down Ashton-Tate (makers of Dbase), about expressions of an idea in a tangible form vs a concept, about SCO claiming to have terminated its contracts yet also claiming they remain in force, the way SCO keeps claiming to own or have invented other companies intellectual property (especially Intel's)... But pointing out things about which SCO is mistaken is a never-ending process. This paper is about SCO Disease, and now you've seen it for yourselves. In various filings SCO has put forth SCO Disease as superior to patents, copyrights, trade secrets, contracts, trademarks, slander, libel, the need to mitgate damages before suing, bad management decisions, consistency, and causality.
Clearly, SCO disease trumps logic.