http://www.cic.unb.br/~rezende/sd.htm > english: SCO x IBM

The SCO x IBM + Linux case, 
under historical view

Published by LinuxToday on Jul 4, 2003

Talk presented at the VI Forum Internacional Software Livre

PUC- RS, Porto Alegre, Brazil, June 2003
[original in portuguese]

Professor Pedro A. D. Rezende
Departament of Computer Science
University of Brasilia
June 20, 2003



 

Index

  • Copyright or patent?
  • Software patents and emperor's new clothes
  • New levels of cluelessness
  • Unix origins and first strife
  • Legal framing for some cross-fertilization
  • Motives, faiths and spins
  • What could've Enderle learned from Andersen?
  • Conclusion

  • To understand what lurks behind the legal case SCO brought against IBM, and its implications for the future of our societies, we have to know two things: a little of Unix history and what has been construed as intellectual property in software.
     

    Copyright or patent?

    Starting with the latter, one begins by distinguishing copyrights from patents. Copyrights are commercial rights for the protection of intellectual work against unauthorized use. In this case, a programmer's source code. And patents are legal rights for the protection of inventions, granting the inventor exclusivity rights to explore his or her invention.

    The key point here is that the spirit of these laws have become irrelevant. The fact that granted rights are limited in time or scope for the purpose of striking some social balance, has been obfuscated by sophistry in repeated extensions gained through appropriate lobbies. (i.e. Diamond vs. Deihr, the test case for Sonny Bonno law, etc.)

    SCO's CEO was quoted in news services on may 29 as having said: "we have the contractual right to block improper use of code, methods and concepts from Unix into Linux". To make sense of what he says, one has to know what are the things he claims and names "methods and concepts" from Unix, and put them in the proper context.

    Methods and concepts for computer programs became patenteable only after the first such submission was accepted by U.S. Patent and Trademark Office (USPTO). In 1981, after several prior attempts were refused. These refusals can be accounted by the fact that U.S. law forbids the patenting of "laws of nature", and as such one uses to consider, as well as common goods, the logical and linguistic constructs. Furthermore, patent laws require that submissions describe something new, useful and non-obvious.

    This happened after a 5-4 U.S. supreme court decision broadening the scope of what is patenteable. The floodgates of lawyers' madhouse were then opened. Backed by a 30 year old congressional report suggesting that "anything under the sun made by man" qualifies for patent protection, this vote gave USPTO  the legal green light to run a frantic alchemical gold rush through cyberspace (http://www.forbes.com/asap/2002/0624/044.html).

    There has been, since then, more than one hundred thousand of such patents granted, by bureaucrats whose notion of ineditedness, usefulness and obviousness in the world of bits are, to say the least, weird, albeit turning USPTO into a government cash cow. As a result, social damage from the 2000 internet bubble burst may look like a simple hiccup, depending on how one looks at the skyrocketing costs of cyberlitigations, past and to come, today on the seven digit dolar value figure and rising, in value and number.
     

    Software patents and emperor's new clothes

    These things which became known as software patents, are legal grants for industrial monopolies issued through hermetic, broad and abstract descriptions written in legalese. They are confusing enough to drive deep pocketed lawyers witch hunting programmers and software companies -- especially the autonomous and small ones --, framing just about any practical idea of what can be done through software as someone else's property. Before these opinions get the label of far fetched, a look at some details about patent submission and their dispute resolution procedures is indicated.

    We wish to recall an emblematic case, protagonized by Leon Stambler against RSADSA, commented by Bruce Schneier in his e-zine Chrypto-gram of march 15 (http://www.counterpane.com). Armed with two patents on authentication protocols, Stambler was blackmailing companies in the computer security software development business. The techno-legal analysis of his patent documents would cost the victims more than the several thousand dollars he was asking to settle.

    When Stambler knocked at RSA, threatening the company with legal action for patent infringement in their SSL protocol, there would have been no reason for surprise, since this type of shakedown is, in the words of a lawyer involved with them, "repeated weekly in corporate America". But the victim decided, against economic sense, to call the bluff. After all, RSA had submitted six patent claims for SSL in 1994 and obtained them in 1997, whereas Stambler had submitted his in 1992 and gotten them in 1999.

    Patents are retroactive to the date of submission, despite the secrecy of the granting process. The submitter is obliged to publicize only a short abstract of his claim, which can not be changed during the process. But here's the catch: submission itself can be changed during the granting process. Which in turn require new -- and still secret -- analysis after each modification. Such are the infamous "submarine" patents.

    Thus, Leo Stambler have had seven years to cook up the hundreds of pages of legalese describing his "invention", while able to learn about the ideas RSA and other successful companies were being granted patents for, especially those that could fit under the short abstract of his claim, with the help of some legal semantic grease. These, the world should know, are the patents the US government wants every country in the planet to recognize under "free trade" agreements.

    To show that the emperor's new selectively transparent clothes are indeed none, an australian lawyer recently submitted a patent claim for his invention which, after granted, was announced by him as a patent for the wheel. (http://edition.cnn.com/2001/WORLD/ asiapcf/auspac/07/02/australia.wheel). "But this would invalidate the patent!", complained an australian patent official, while accusing him of bad faith.
     

    New levels of cluelessness

    Yeah, but what about legalese descriptions of "methods and concepts" for computer programs, including submarines waiting to surface? And pirate claims from 2000+ year old classics, such as Euclid's "Element" -- the foundation of Geometry and mathematics, as the "fat line patent" held by IBM proper?  No wonder only about half the patents which face litigation escape revocation.

    As for unix's methods and concepts, what's new under the sun regarding the management of queues, files and resources, with the advent of computer operating systems, besides languages and parameters? And for that matter, with regard to all software patents, what could bits in chips do that cannot be done by pencil in paper, regardless of how more slowly?

    To those who are not clueless, it was just a matter of time before one of these insane imbroglios, and its unfoldings from the courts, would hit some software that is free (in the sense of freedom). Richard Stallman, from the Free Software Foundation, has been warning us for some time. Social movements for freedom of knowledge in the information age, like the GNU project and the open source initiative, have neither the stakes nor the vocation for this maddening gamble.

    On the contrary, they actually reach out to neutralize the social erosion caused by it. This is why Linux rhetoric is different from that of mainstream IT business, and also why Mr. Enderle complains about but don't get it (http://www.newsfactor.com/perl/story/21764.html).

    The SCO versus IBM (read Linux) case is the first shot from a new escalation level in the digital revolution, where a struggle for legal supremacy is being fought. On one side, the freedom of economic power to self-organize. On the other, the freedom of humans to pursue knowledge for themselves. To dismiss such claim as paranoid, one has first to think of a revolution without victim (if one is found, the "r" would have to be dropped from the word).

    This first shot has been called: who, after all, is the owner of the "methods and concepts" of the Unix operating system? Time for some unix history.
     

    Unix origins and first strife

    Unix originated with MULTICS. MULTICS' home page (http://www.multicians.org/history.html) describe its beginning "as a research project [that] was an  important influence on operating system development". This research project, started by MIT, AT&T and GE in 1965, at a time of computer science's infancy, promised the revolutionary idea of "interactive programming", with terminals substituting decks of punctured paper cards.

    MULTICS' first operating system, derived from MIT's CTSS (Compatible Timesharing System, in use by 1961), is known to have been deployed in 1969. Still incomplete and unstable, after four years of inflating specifications, according to participant's accounts. In that same year, AT&T decided to pull out of the project, and its participants began developing an alternative and simpler system at the company's research lab (Bell Labs). This alternative system, which later gained the trademark Unix, was deployed in 1974 after being rewritten in the language born with it -- C --, by authors Ken Thompson and Dennis Ritchie.

    At that time, AT&T was the monopolistic concessionary for long distance telephone business in the U.S., and the rules for that concession did not allow the company to commercialize software. (http://student.lssu.edu/~gbharadw/unix_history.htm). Because of the great need and interest from academic and research institutions in an innovative operating system for teaching computer programming, AT&T began, in 1975, making agreements with several universities, including Berkeley, by which the Unix source code was released. Through a nominal fee, without any commitment to support, with the universities allowed to modify, correct and extend the code on their own, for their needs.

    Snippets of source code emerging from this allowance were exchanged among participants, some of which ended up at later versions of AT&T's unix code. In 1978 one of Unix's authors -- Thompson -- spent one year as a professor in Berkeley, where he and his students developed the first version of a similar system, known as Berkeley System Distribution, or BSD (at that time I was a graduate student at Berkeley, but did not participate in this project). BSD later evolved to the free BSD systems which, having gained an open source license from the regents of the University of California at Berkeley, entered into free distribution (Twenty years of Berkeley unix: http://www.oreilly.com/catalog/ opensources/book/kirkmck.html).

    It was only in 1983 that AT&T was given permission to commercialize software licenses. Its Unix version at the time gained the "System V" label, distinguishing it from other versions, including those evolved in other companies from other licenses. In 1992 AT&T sold System V to Novell and, one year later, Novell began litigation against the University of California at Berkeley, for the alleged reuse of System V code in BSD. The case was settled in court, with a good part of BSD rewritten by 1994, before BSD distribution could resume, with Novell insisting the exact terms of the agreement kept sealed (http://www.opensource.org/sco-vs-ibm.htm).

    Berkeley rewrote a major part of BSD source code which happened to be, in 1993, replicated from (or into) System V, but not all. Why not all of it? The detail to notice here was Berkeley's legal council's insistence while pursuing settlement with Novell. Why did Berkeley insist, and heeded was it in this regard, that some source code be kept identical in BSD and System V?  The answer is important the relation it may hold to the quality of faith in SCO's motives to litigate, as well as to the quality of faith in the motives of people with vested interests in IT's mainstream industry, to spin SCO's actions and motives toward psychological warfare.
     

    Legal framing for some cross fertilization

    In the Novell versus Berkeley case, no patent regarding System V's "methods and concepts" was brought to the table. If such patents existed, and were used to fuel that litigation, they could have been questioned with regard to the ineditism of their protected "inventions", due Unix's origin: a multi-institutional research project taking place long before any such thing as software patents were ever considered by patent offices, including USPTO.

    The project's main publications -- six articles at the 1965's Fall Joint Computer Conference and more than three thousand pages of the MSPM document -- were enough to establish the public domain under which such methods and concepts would fit in. Thus, the Novell versus Berkeley case revolved solely around copyright matters. That is to say, around authorship rights over System V's source code.

    In the out-of-court settlement for that case, Berkeley insisted that some code remained identical in both systems so as to formalize the two-way character of the contributions fructified from the agreement at BSD's origin. The agreements made between AT&T and universities permitted and promoted multilateral contribution for the evolution of the several versions of operating systems derived from unix's released source code, including AT&T's unix proper, versions that became known as "unix-like" systems.

    Being the adaptation and installation of an operating system in varied and new hardware platforms the main advantage for its source code availability, as well as the main challenge facing the agreed participants -- including AT&T --, all ended up benefiting from the crossed fertilization of code development, allowed by those agreements and disseminated through virtual groups such as AUUG and USENIX, (http://minnie.cs.adfa.oz.au/Seminars/AUUG96/pdppaper.html) .

    System V being no exception, the Novell versus Berkeley settlement was signed in 1994 as Berkeley insisted it should. With some code remaining identical in the kernel's process control module on both systems, and with BSD declared unincumbent on System V's copyright.

    Shortly before the dispute between Novell and Berkeley began, having learned about operating systems in other MULTICS-derived systems at the University of Helsinki (Ultrix and Minix), Linus Tovalds licensed, in 1991 and under GPL (General Public License), his Linux kernel. On the other hand, this kernel fitted into the void remaining in the GNU project aiming at developing a free operating system.
     

    Motives, faiths and spins

    Some people relate the success of the GNU/Linux system, in detriment of free BSD among free operating systems, to the stagnation and uncertainty the year of 1994 represented for BSD fans, contributors and customers (its community). Nonetheless surpassed, for those were different times, when the freedom humans enjoyed to pursue knowledge for themselves was not yet threatening what was to become the biggest business in history.

    Today, based on fuzzy logic and ethereal, uncertain and undisclosed intellectual property titles, SCO wishes to blackmail those who defeated fears of pack sentiment violations and invested in free software. After threatening the top 1500 of them for alleged violations of its property over unix's "methods and concepts", yielded from a contract (with Novell) which SCO's own CEO later admitted to be "confusing", the company came to admit that evidence of such violations may amount to similarities in some lines of source code present in System V and some Linux distributions. That is, amounting to copyright disputes.
    .
    But if that is so, what code, exactly, would have been "pirated" from System V? According to Bill Craybook, a computer science professor and programmer of operating system kernels, hired by SCO to analyze and compare code from System V and some Linux distributions, what he could find in a first look was only similarities in some eighty lines of code, within a function of the kernel. Now comes the intriguing and spectacular part of the story.

    This function where similarity of source code was found, seems to have originated from freeBSD. (http://www.osopinion.com/perl/story/21722.html - cached). If this information is confirmed, the entry of such code into System V would have been a result of the early cooperation agreement between AT&T and universities, legitimized by the settlement between Novell and Berkeley in 1994. And the license under which this code was originally distributed -- that of the free BSD systems -- allows for unrestricted free use of such code, without any discrimination or privileges.

    In its insistence, Berkeley's legal council negotiating a settlement with Novell wanted to prevent exactly the kind of confusion SCO seems now eager to exploit against Linux big users. Meanwhile, there has been some worrisome news regarding the case. First, Microsoft decided to pay up for licensing whatever SCO may claim to be its property rights held over Microsoft unix versions or other code, lending credibility to SCO's wild claims and injecting some much needed cash for its legal fight. Then, "influent analysts" (such as Rob Enderle) began spinning off the case the kind of strategic counseling Enron got from its external auditor.
     

    What could've Enderle learned from Andersen?

    If Linux is immature, puts the customer too much in the hands of developers, and speaks the wrong language, as Enderle says (http://www.newsfactor.com/perl/story/21764.html - cached), what is a user to make of alternatives Enderle may have in mind? Systems that have to roll and evolve to the tune of their developer's stagnating cash flow, ever more plagued with undue complexities, planned obsolescences and lock-in proprietary traps, thus ever more bugged and flawed, all held with impunity through censorship and threats imposed by martial-like kafkian laws such as DMCA, UCITA and the like, lobbied by these same developers?

    What to make, for instance, of the latest show of maturity performed by Linux's main "alternative", when a long awaited Windows XP security patch from MS had to be abruptly recalled, after more than half a million downloads on may 28, for it was crippling the connectivity of anxious would-be-safer costumers? (http://www.infoworld.com/article/ 03/05/28/HNmsyanks_1.html?security)

    What to make of its owner's disrespect for a settlement signed in august with U.S. Federal Trade Commission, to quell disputes over false claims regarding its main digital security services? ( http://www.politechbot.com/p-04732.html - cached)

    Or its recent scatological blunder, when a PR stunt announcing its iLoo project in England (yes, wireless PCs w/ XP in portable toilets!) was later dismissed in Redmond as a self-hoax, only to be again unhoaxed so as to, shall we say, be flushed away? (http://www.usatoday.com/tech/news/ 2003-05-14-iloo-hoax-retract_x.htm)

    Gandhi once said that revolutionary movements are first ignored, then ridiculed, then attacked, before their final judgement. Free software -- and its developing model known as extreme programming -- were first ignored: the way the Internet has arisen, simply forgotten. Then they were ridiculed: "just can't work, it's so weird, so un-businesslike". Now they are being attacked. Reminding us, for that matter, of the weapons of mass destruction in Iraq.

    The language and religion of greed are not only harmful to moral values. They cripple the intellect and blind the soul, infecting the subject with terminal cluelessness and unabashed hypocrisy, making everyone else look fool to their eyes and themselves deceiving to mirrors. Just like the Aztecs, with their elaborate rituals of escalating sacrifices of prisoners of wars, whose hearts were yanked out still alive and pulsating, believing that doing so would win them prosperity from the gods of chance.
     

    Conclusion

    Patents on software "methods and concepts" are like junk bond derivatives in a gambling game, a game set up by the lobby of monopolistic proprietary software industries to perpetuate their business. Armouring against the effects of its own escalating inefficiencies, its business logic weaves the same strategy weaved by pre-colombian mesoamerican mythical order and by Roman church's counter-reform ideology.

    If patents are to enter the SCO versus IBM case, they will be like slot machines at the hands of judges and jurors. Worse yet, like the intellectual hearts of prisoners of the digital revolution. The patent and trademark offices, with their ethereal indulgences selling salvation into the afterworld of bits, seem ready to fulfill the duty of registrars for a new Inquisition, throwing us back into a reloaded edition of the Middle Ages.

     We are all being made virtual prisoners, for our own ideas will, indeed, be ever more dangerous. Dangerous in the sense preconized by George Orwell in 1984, if our societies stay under the spell of greed. If they keep going about their business as usual, chasing the mythical carrot of monotonic prosperity, hanging from the market fundamentalist doctrine and waved by its invisible hand, ridiculing human values while paying lip service to them.

    Our future and real challenges rest now at the hearts and minds of judges, legal scholars and international treaty negotiators. May God bless them!


    v.2, revised july 5 2003

    Released under copyleft :  http://www.gnu.org/licenses/fdl.html