|HOME||PUBLIC LIBRARY||ANOTHER PERSPECTIVE||INFOPERSPECTIVES||CONTACT|
On January 19, Platform Solutions belatedly responded to IBM's patent infringement lawsuit, which was filed November 29. Platform's response was more than an answer to IBM's complaint. It was also a countersuit alleging IBM had treated Platform unjustly and illegally, invoking a concept lawyers call promissory estoppel; asserting IBM was in violation of antitrust laws; and claiming IBM's behavior mocked the arrangement under which IBM got out from under the 1956 Consent Decree that had constrained Big Blue's practices for more than 40 years.
Platform brought in some high powered lawyers to plead its case, Susman Godfrey, a big and well-regarded outfit based in Houston that has offices in several cities, including New York. Stephen Susman is at the top of the counsel list, which means the firm intends to go at this matter with the same vigor it used as it helped take apart Enron. The battle between Susman Godfrey and IBM's counsel in this matter, Quinn Emanuel, another powerful firm, is unquestionably in the legal heavyweight category. But two computer companies, one the giant IBM, the other the small but feisty Platform Solutions, are only the visible parties in this battle, which will decide whether Platform can compete in the mainframe market with its product, a firmware implementation of the System z architecture on Hewlett-Packard's Itanium-based Integrity servers.
Clearly, HP and Intel both have a lot at stake here, too. IBM-compatible mainframe would give HP a big boost in glass houses, and it would be precisely the kind of showcase Intel needs to bolster its star-crossed effort to popularize the Itanium family of processors.
In fact, HP may be more intimately involved than it admits. Platform's court filing says, among other things, that IBM's behavior threw a monkey wrench into a pending deal that would have resulted in the acquisition of Platform by an unnamed company. HP is probably that company, but outsiders like us can't say so for sure. The talks and other intercourse between Platform and its potential buyer are wrapped in a cocoon of nondisclosure agreements.
There's another party that is probably keeping a keen eye on the situation, too. That interested outsider is Electronic Data Systems, which has a business that could really take advantage of the Platform system. It is also arguably the strongest computer services company with a knack for using IBM-compatible systems in its operations. EDS was, in the past, a very good customer of clone mainframe maker Amdahl, and today it might still be operating more Amdahl processors than any other company in the world. EDS is, of course, not the only company that liked a mix of IBM and compatible hardware, but it is a firm known for its moxie. If the Platform system works well, and if it offers EDS technical or economic advantages, EDS would not be inhibited about buying the machines. And if EDS said the boxes were good, Platform would have an easier time getting its products into other homes.
But nobody is going to buy a Platform machine if IBM won't license its software for the box, and that's precisely the kernel of all this litigation.
IBM said it would not license software for Platform's machines because it believes Platform's technology infringes IBM patents. IBM has also said it has no intention of settling any patent conflicts by working out a licensing arrangement with Platform. This is exceptional. IBM has generally been willing and able to hammer out patent licensing deals with other computer companies, including Amdahl and Hitachi, who were the two big vendors of IBM-compatible hardware during the golden age of bipolar mainframes. Now it is IBM that is exhibiting bipolar tendencies, swinging from its former practical openness to a stark refusal to work cooperatively if somewhat reluctantly with a rival.
IBM's complaint also alleged breach of contract by Platform. Specifically, IBM asserts that Platform's technology violates a key clause in the IBM Customer Agreement, also called the ICA, under which IBM licensed software to Platform. Any company that buys a Platform machine would have to license software under the very same ICA, which IBM now says would be completely unacceptable.
Platform, in its answer to IBM's complaint for patent infringement and breach of contract presents three basic defenses.
First, Platform denies it infringes any IBM patents and indeed challenges the viability and applicability of these patents to this case, or as lawyers put it in their peculiar language, the instant matter. Patent law is about as simple as cold fusion, so we won't share with you a full explanation of the vast, convoluted, and maybe downright ugly body of law that has grown up over the years. What any layman can and should understand is that most patent disputes, except those that involve outright piracy, end up with negotiated settlements, because nobody, not even the world's most prolific patent holder, IBM, can completely rely on the law and its lawyers for the business results it wants.
The patent dispute in this case is very unlikely to have a sumo moment, where one or another of the gigantic law firms knocks its adversary out of the courtroom.
Still, even if it cannot achieve a decisive victory, IBM could tie up Platform for so long with the patent fight that the whole dispute becomes irrelevant. That is the nature of patent disputes when technology evolves far faster than lawyers can fully argue their positions, and it may well be the case here. The responsibility for moving the case along lies with Judge Stephen Robinson, and it could take all his skill and energy to keep the lawyers, particularly the ones on IBM's side, moving along. They teach a lot of things in law school, but not mule skinning, and lawyers only realize they've missed some vital education in this art if they rise to judgeships.
Platform argues that nothing about the way it runs IBM software on its products violates the ICA, and indeed it does appear that IBM's construction of IBM's ICA, paragraph 4.1.1, which says the customer cannot "reverse assemble, reverse compile, or otherwise translate the ICA Program," is a stretch. Every IBM customer using a mainframe has a machine that, in a fashion more like Platform's than different from it, translates instructions as it performs information processing.
Another point made by Platform is based on a concept with roots in English law that is also a part of U.S. law, promissory estoppel. Basically, if party A promises something to party B and party B acts relying on A's promise, A pretty much is supposed to keep its promise and a court can order A to do so, or to not do something that violates the promise. The classic example involves rental of distressed property that suddenly becomes much more valuable. A rents the property to B and says B can have it at this rent for as long as B wishes. B builds a business around this. Conditions change and A wants to really jack up B's rent. B goes to court and explains the deal. The court tells A he is stuck with B paying the lower rent.
This concept does not apply to every similar situation. If it did, nobody could ever be laid off at work. But even in employment situations, the concept has influenced law the world over — particularly in wealthy countries — and it is embodied in various restrictions governing layoffs and the social safety nets available to people who lose jobs though no fault of their own.
So promissory estoppel applies in contractual matters, but not all of them, and it turns out to be applicable to the situation in which Platform Solutions finds itself. IBM has for some time licensed its software for use on Platform servers, not only to Platform but also to others, including LL Bean and Lufthansa. (And maybe to EDS.) IBM and its licensees, and in particular Platform Solutions, acted on the basis that this licensing policy would continue. That all changed on November 29, when IBM filed its lawsuit. Platform asserts that the court should order IBM to continue licensing software to Platform (and therefore to other users of the Platform machine) because it has done so in the past (going back to the beginning of the plug-compatible mainframe era).
One aspect of estoppel that may play a role in IBM's response is a concept in the body of judicial power from which it emerged, called equity. Equity is a body of procedures and decisions that has cultural rules that distinguish it from other case law and from statue law. Equity was a return to common sense and natural law that arose in English courts at a time when the courts were failing to provide justice. Where other law fails to provide good results, the traditions of equity often provide a basis for fair and reasonable decisions. But there's a catch: to invoke equity, a party must come before the court with "clean hands". In this case, that means Platform cannot ask for an equitable judgment if the court believes it is a patent infringer, which would make it a party with unclean hands. If all this seems as clear as mud, that's because equity is a tradition that can be opaque to lawyers and laymen alike.
Platform's lawyers chose not to mention the case of Flex-ES mainframe emulation systems, which run on ordinary IBM X86 servers, and which more closely fit the description of "translation" in the ICA than Platform does. Nor does the suit mention the Hercules emulator, for which IBM will not license software. Yet one of the risks IBM faces in its litigation, even if it's a small risk, is that it could lose the case in a way that opens the door to other firmware and software products that provide mainframe functionality on hardware that's substantially different from the equipment IBM and Platform use.
The third element in the Platform defense is antitrust.
When IBM got the 1956 Consent Decree lifted in 1997, with a four-year phaseout of its various provisions that was completed in 2001, the court warned IBM that the Justice Department could go back after Big Blue if IBM was a recidivist, and particularly if it engaged in the tying of products. Tying means that you must but one IBM product in order to obtain another. For example, if IBM says that you can't buy IBM systems software unless you also but an IBM mainframe, that is tying. Well, that's how Platform sees it, anyway.
Platform has accused IBM of a number of antitrust violations. The accusations boil down to the same concept: That IBM has monopoly power in mainframes and is using this power illegally to restrain lawful competition. Monopoly power per se is not illegal under U.S. law, which concentrates on how such power is used or abused.
The application of antitrust law is further restricted by the way the law, including case law as well as statute law, defines monopoly. In legal terms, a monopoly is defined as control over a particular market, and there has been lots of argument about just what the relevant market happens to be when it comes to mainframes. In the past, IBM was able to argue with considerable success that mainframes alone do not constitute the relevant market when it comes to antitrust law because users of mainframes can and do migrate to alternatives. But just because IBM succeeded with this in the past does not mean it would get its way today.
In the past, the focus of antitrust suits against IBM was hardware. At various times, the alternatives to IBM mainframes were incompatible systems made by other computer manufacturers, most of them long gone, plus in some cases midrange computers, also made by companies that have generally disappeared. These days the mainframe market is defined largely by software. And what Platform would like the court to do is to view the IBM and IBM-compatible market as the relevant market for legal purposes. In that case, it is a tautology that IBM has pretty much complete control over the market in machines that can run IBM mainframe software. (It's not total control as long as there is one Fujitsu/Amdahl , Hitachi, or Flex-ES box running somewhere, but it's pretty close in value terms even though there are hundreds of little machines running Flex used around the world.)
It's hard to believe IBM would be forced to give up hard-won legal ground taken by IBM's preeminent lawyers, Cravath Swain & Moore a decade ago. And any action taken by the Antitrust Division within the Department of Justice would be quite a surprise. But it's possible for IBM to find that it can in fact be put under a bit of pressure by its adversaries and by the Justice Department. IBM did bring its case in the same venue that produced the 1956 Consent Decree and its dissolution. So, while IBM might have a home court advantage in asserting that Platform's antitrust claims are old, settled business, particularly in the Southern District of New York, IBM nevertheless would be unwise to shame the court by acting in a way that is disrespectful of the process and court that ultimately freed it from the restraints imposed in 1956.
Judge Robinson will not only seek justice. He will also seek justice that is consistent with the distinguished history of the Federal Court for the Southern District of New York. And because the lawyers before him are, in a sense, outsiders, litigators from firms with roots in Houston and Los Angeles, the court might teach both parties' lawyers a few things about a New York state of mind.
— Hesh Wiener February 2007