WE FOUGHT THE LAW AND THE LAW WON
In the Internet business, At the peak of the boom, NSI was city hall, as in, “you can’t fight City Hall.” Reading the history of NSI’s judicial victories is about as uplifting as re-reading the Dredd-Scott decision. (You’ll recall that was the case where the United States Supreme Court held that a man was property, if he was a slave in his state of origin, even when he crossed over into a state where slavery was illegal.) The NSI cases evoke the same sense of the courts bending to the winds of powerful interests. By accepting NSI’s version of the facts without exposing them to scrutiny, the courts helped NSI build fortresses of judicially-reinforced ignorance. From reading the opinions, it is apparent that the judiciary collaborated with NSI to allow the wool to be pulled over its eyes. How could this happen?
The history of NSI may give some clues. How did it get going? How did it gain so much power? What interests does it really represent? So now for the quick history of the Internet that will soon be learned in the cradle from the mouths of Internet Barbie and Virtual-War G.I. Joe.
In the beginning there was DARPA and ARPA, and their mission was to provide a redundant network for communication among defense and educational computer installations. By redundant we mean that even if one link in the system breaks down, the entire system continues to function. What is amazing is that these scientists managed to collaborate to create the system. Under the benevolent aegis of defense spending, unimpeded by the concerns of crass commercialism, the engineers designed communication protocols based on the UNIX computer language that enabled many different kinds of computers running different kinds of software to communicate over the phone lines.
In essence, the early Internet pioneers put universal communication functionality at the top of their priority list. The most important thing was for all of the computers to be able to talk to each other. Historically, the Internet was foreshadowed by the legend of the Tower of Babel. This early public works project was made possible only because human beings possessed a common language. According to Hebraic historians, utilizing this common language, the nations of the earth attempted to build a stairway to heaven. Jehovah became wroth with this demonstration of arrogance, and rather than burning the tower down as he might have on another day, struck on a more devious solution. Jehovah “confused their speech.” He smote them with the curse of varying languages. The builders suddenly found themselves unable to communicate, and as a result, abandoned their plan to ascend to the heavens by means of a ladder.
Even if we can’t talk to each other, the Internet pioneers decided, our machines can! But the common language developed by the engineers had a disadvantage -- it was numerically-based, cryptic and engineer-like. Sure, an engineer can rattle off an IP address (that’s “Internet Protocol” address) like his newborn’s birthday, but the rest of us have trouble remembering our phone numbers. So if I wanted my computer to call your computer, I would have to input all of these digits.
Along comes John Postel, the patron saint of the Internet, who certainly looked like a Deadhead, and worked for the University of Southern California Information Sciences Institute. While doing contract work for the U.S. government, Postel came up with a clever idea. Since anything arbitrary can equal any other arbitrary thing in the world of mathematics, why not equate numerical IP addresses with arbitrary strings of letters? Since any string of letters can equate to any string of numbers, all you need is a database that interrelates the strings of letters with the strings of numbers and automatically routes messages to the appropriate computer. And thus was born the age of the Dot-Com.
All IP addresses, decreed St. Postel, would end with a three-letter suffix, like .Gov, .Net, .Org, and most importantly to our story, .Com. To the left of the Dot, the user could put anything they wanted, up to a limited number of spaces. (This space has grown so much now, that domain names can practically be a paragraph long.)
Domain names, you might expect, would be a natural-born hit, a marriage of necessity and style. It was sort of bohemian, if you think about it. Bohemians started naming their cars to personalize the smog-belching carriages that rule and enable our lives. Now we could personalize our computers, and with more than mere sentimental effect. Naming your car is something you stop doing when you get your first real job (although the trend resurges among soccer moms, whose Suburbans proudly bear names like “Great White Beast,” inevitably chosen by the micro-Viking progeny of a true trophy blonde).
But a name for your computer is not just a sentimental tag attached to a mechanistic object. It is a grubstake in the new economy. It is a homestead on the information frontier. It is a doorway to a community of knowledge, information and entertainment that is so compelling in its current, high-speed incarnation, that broadband users, who get more information faster, spend 28% more time at their computers than dial-up users. Dial-up now has market penetration roughly equivalent to that of television in the late 1950’s. Meanwhile, the demographic profile of the average Internet broadband user is identical to the demographic of the dial-up user five years ago. I couldn’t have rattled off this stuff three years ago, and by the time you read it, the numbers will have changed. And yet it was all there in potential form in the early 1990’s, a glimmer in John Postel’s eye.
When Postel invented them in 1984, domain names were adopted only by a tiny group of engineers spread around the country. They started emailing each other about their typical geeky topics, one of which was the future structure of the Internet. For about ten years it was basically managed by volunteers. In 1994, Postel wrote an email to some friends who were also helping to run the system, laying out the plan for an agency to administer the registration of domain names. Postel and his friends created an entity known as IANA, the Internet Assigned Names and Numbers Authority.
IANA operates out of a building in Marina Del Rey, California, and hosts a website at IANA.org; however, it does not have any corporate or governmental existence that I’ve been able to discern from my research. In response to a lawsuit seeking to compel IANA to create a “.Web” top level domain name, Postel filed a declaration stating that IANA was merely a “function,” performed under contract for a security agency of the United States Department of Defense. Apparently, at one time, IANA had control over all of the domain names. Later, IANA transferred responsibility for managing the four main top-level domain names (.Com, .Net, .Gov and .Org) to the International Corporation for Assigned Names and Numbers (“ICANN.”)
Meanwhile, the cash for funding the registration program came to be administered through the National Science Foundation, which put the domain name registration function out for bid to data processing companies. The contract was just for a few million bucks, so in May 1993 it went to NSI, a small company in Herndon, Virginia.
The National Science Foundation paid NSI to register domain names for free. In December, 1994, seven months after Gary registered Sex.Com, the NSF recommended that users start paying to operate the system. In May, 1995, NSI was bought lock, stock and barrel for $4.8 Million, by SAIC, Science Applications International Corp., a privately held San Diego company with a Board of Directors peopled by defense establishment magnates, including former Secretaries of Defense Melvin Laird and William Perry, former NSA chief Bobby Inman, and former CIA Directors Robert Gates and John Deutch. On September 14, 1995, NSI announced it would start charging $5 per domain name registration.
The cash cow started producing immediately. Between September 1995 and March 1996, NSI took in $20 Million in registration fees. The arrow just headed straight up from there, like an X-15 pilot in the movie, The Right Stuff. Boy, did NSI have it. Fueled with the power of absolute monopoly over a desirable resource, NSI went from being a cheap date to a supermodel in two short years, turning into a stock market darling at the beginning of the tech-IPO craze. NSI proudly described its glowing success in a July 1997 stock prospectus:
“Net registrations within the TLDs maintained by the Company increased by 233% from approximately 246,000 domain names registered at March 31, 1996 to approximately 818,000 domain names registered at March 31, 1997. The Company believes that commercial enterprises and individual Internet users worldwide are increasingly recognizing the .com TLD as a desirable address for commercial presence on the Internet. *** [T]he Company believes that the potential for continued growth of domain name registrations by commercial entities and services related to those registrations is substantial. Net revenue from Internet domain name registration subscriptions accounted for 76.5% of the Company’s net revenue for the three months ended March 31, 1997.”
The day NSI went public, at the opening bell on September 26, 1997, the market was drooling, and the first trade closed at $25/share, 40% above the $18/share offering price. By the end of the day, NSI had a market cap of $382.5 Million, and SAIC had sold enough stock to harvest a $9 Million profit, while keeping a 28% interest in the company. A little champagne was consumed in San Diego that night. The fruit of the Internet, born in the soil of DARPA, the Defense Advanced Research Project Agency, fell quite close to the tree.
Eventually, administration of the government contract with NSI passed from the National Science Foundation to the United States Department of Commerce and ICANN. A period of conflict between Commerce and NSI took place during 1999, with NSI taking the position that it “owned” the right to register all .Com domain names in perpetuity. NSI also claimed it owned the entire database of five million .Com registrations that it had registered during the period of its government-created monopoly. In late 1999, NSI gave up its death grip on the .Com franchise and database when it agreed with ICANN to allow the creation of additional registrars who could compete in the registration of .Com and other domain names. However, above all of the registrars, NSI continued to hold the privileged position of being both a registry and a registrar.
There is only one domain name registry, and it is NSI. There are now many registrars, but they all must register their domain names with NSI, the world’s only registry. For this service, NSI charges $6. Thus, if you register a domain name with any other registrar, $6 of what you pay goes to NSI. Meanwhile, NSI collects the $6 plus a premium that has steadily declined from around $30 for every name that it registers directly through its NSI.Com registration portal.
The agreement with ICANN in 1999 also provided that NSI would have to choose between being the registry and being a registrar by 2001. However, for those who read the papers closely, it was disclosed in 2001 that NSI had managed to renegotiate the agreement, allowing it to continue operating as both the registry and a registrar until 2005. Since being acquired in March 2000 by Verisign Corporation, NSI and its corporate parent are probably destined for a merger with Microsoft. (Anyone hear the opening strains of Terminator II?) For all of these reasons and more, I took to calling NSI the “stainless steel saint.” Every lawyer knows that some companies are more fun to sue than others. NSI was no fun at all.
NSI’s counsel of record, Dave Dolkas of Gray, Cary, Ware & Friedenrich, had the kind of self-assurance that told you this man had no question where his next meal was coming from. Dolkas was tall, dark, handsome, and well-accustomed to winning. He wore his good quality suits very nicely, maintained unflappable dignity in the face of punk tactics, and seemed to know exactly how little work he needed to do in order to get out of a case for NSI. He relied on precedent, legion in his favor, which he recited in a matter-of-fact, almost bored tone. Without getting excited, he would ask the simple question, “Why are we being sued here?” This tactic has worked so well both with the public and the courts, that Ellen Rony entitled her chapter on NSI in the Domain Name Handbook, “NSI: Caught in the Crossfire.”
NSI was caught in the crossfire in a number of cases. That is the primary problem with the body of law that was created when NSI responded with a coordinated strategy to a series of helter-skelter lawsuits attempting to compel it to do all kinds of stuff. These cases were often litigated by plaintiffs who exercised little forethought and not much staying power, so NSI racked up a series of victories. In NSI’s repertoire of favorable precedents that they plied before the courts like a string of get-out-of-jail-free cards, are many instances of bad facts making bad law. For example, back when they had the registrar monopoly, they were sued for refusing to dot-com the word “fuck.” Guess when they won that case? When it became “moot,” because other registrars had been created that were ready, willing, and able to register Fuck.Com or anything else you could think of. The meaning of that case, to Dolkas, would be that NSI always wins. To me, it says that if you stall the process long enough, you can thwart it. But which meaning would the court infer?
NSI’s opening brief told Judge Ware that the entire process of registering a domain name, or transferring an existing domain name, was so automated that it was almost automatic. In essence they argued that Cohen’s letter would have passed in front of a minimum wage employee recruited from a technical college in Alexandria, Virginia, who would have accepted such a transfer as a matter of course and performed the change of registration as directed therein by the original registrant, Online Classifieds, Inc.
NSI retreated from this position in later filings, when I pointed out that Cohen had faxed the letter directly to David Graves, an individual who was far from minimum wage in the NSI hierarchy. In fact, Graves was the person regularly designated to testify on NSI’s behalf in litigation. Additionally, at Cohen’s deposition that took place after NSI filed its initial moving papers for summary judgment, Cohen testified he had spoken directly with David Graves, who had essentially told him what to put into the forged letter:
CARREON: Did you have any communication over the phone with anyone at NSI?
CARREON: Who were those people? Pertaining to this transfer right here.
COHEN: There would only be one of the -- Dave Gray (meaning David Graves).
CARREON: And what did you talk to Mr. Gray about?
COHEN: Mr. Gray asked me to send him a copy of the documentation I had for the transfer.
CARREON: And this was all you were able to send him, this -- Exhibit 2?
COHEN: That’s correct. Unless there was a cover sheet, which I don’t remember, and there may very well have been.
As a result of this additional revelation, I made a motion to Judge Ware requesting to take the deposition of David Graves. Judge Ware denied the motion two months after granting summary judgment. Other motions I made while NSI’s motion for summary judgment was pending, like the one protesting Phil Sbarbaro’s improper contact with Ellen Rony, were also ignored until after NSI was safely out of the case.
NSI’s exit from the case became final on May 8, 2000, nearly six years to the day after Gary registered Sex.Com, on May 9, 1994. Judge Ware’s order granting NSI’s motion for summary judgment recorded the date in the first line of the factual recitation. Judge Ware concluded that Kremen had no contract with NSI, making no mention of my argument that, by agreeing to become one of the registrants of domain names, Kremen had contributed to the growth of a user base that had concrete value to NSI, which it exploited specifically in its IPO stock offering by boasting about the number of registrants in its database. Judge Ware also concluded that Kremen was not a “third party beneficiary,” that is to say, someone entitled to make a claim under the contract between the National Science Foundation and NSI, pursuant to which NSI became the registrant of domain names in the first place. Since NSI was paid millions of dollars to register domain names by the NSF, it looked to me a lot like a situation where the NSF had sort of crowed out, “drinks are on the house,” allowing anybody and everybody who wanted to register a domain name to have one for free. So it would seem that, in order to maintain quality control, anybody could complain about the quality of his drink. For purposes of quality control, it makes more sense to give registrants like Kremen the power to sue NSI, since as a practical matter bureaucracies that grant contracts to companies like NSI rarely police them effectively. And in this case they certainly did not.
Then Judge Ware got around to the big important argument. Was it possible to nail NSI for conversion? Judge Ware said no, concluding that there is “very simply no evidence establishing that a domain name, including Sex.Com, is ‘merged in or identified with’ a document or other tangible object.” He acknowledged that “California law does recognize ‘conversion of intangibles represented by documents, such as bonds, notes, bills of exchange, stock certificates, and warehouse receipts.’” But he adopted NSI’s argument that “a domain name is a form of intangible property which cannot serve as a basis for a conversion claim.”
Wasn’t the registration agreement at least as much of a “document” as a warehouse receipt? The registration agreement gave the administrative contact sole authority to control registration of the domain name just like a warehouse receipt entitles the holder to claim the warehoused goods. When Gary printed out the online registration agreement and mailed it to NSI via certified mail, he emphasized the importance of this document. And the online list of names and numbers known as the Domain Name Server Tables is an electronic document.
Electronic documents are old hat under the Federal Rules of Evidence, which recognize all systems of symbol recordation, from clay tablets to palmtop advertising, as legally entitled to the name of “document.” Smoke signals are probably out. But the registration agreement and the DNS Tables are, respectively, the deed and the land office records of the world of cyber real-estate. If a claim check for a parking lot or a dry cleaner is good enough, surely an online, printable registration record satisfies the document requirement.
Perhaps to avoid dallying with this uncomfortable notion, Judge Ware’s opinion then launched into a litany of justifications for keeping to a conservative course of action, noting that Kremen’s claim “invites abandoning the traditional strictures of conversion to encompass forms of intangible property never contemplated in its formation.” Well, I’ll agree that when the first common law judge held the stable liable for conversion when the wrong fellow rode off on the wrong horse, he wasn’t thinking about Internet domain names. But the basic obligation that the stable operator breached when he gave the horse to the wrong rider is the same one that NSI breached when it let Cohen steal Sex.Com. Judge Ware was really saying that a law that has long been good enough for the stable hand or the warehouseman is too harsh for NSI, an enormous corporation awash in Internet money thanks to a tax-payer funded monopoly. He said, “The Court finds it inherently unjust to place NSI in this untenable position by virtue of performing a purely ministerial function.”
Judge Ware then gave an economic argument in favor of his decision: “Furthermore, the threat of litigation threatens to stifle the registration system by requiring further regulations by NSI and potential increases in fees.” Where he got that information, I’ll never know. There were no facts presented to support that conclusion. The Judge clearly had not paid any attention to NSI’s fees, which had skyrocketed far beyond the basic charge originally authorized. NSI is currently the most costly provider of domain name registration services, adding on oodles and oodles of extras whenever possible above the basic $6 registration fee, which of course, it also collects. And in any event, the “security costs too much” concept just doesn’t make any sense to me, when the essential service NSI provides is keeping accurate records of domain name ownership. That’s the bare, central task. It’s like saying we don’t have enough money to train nurses how to draw blood with clean needles. That’s what they are doing. They should do it perfectly every time.
Trying to avoid “creating the proverbial slippery slope,” Judge Ware buttressed his decision with the language of judicial restraint: “The Court is compelled to uphold this distinction rather than contort the cause of action to encompass property never contemplated.” But this was not true judicial restraint -- in fact its very opposite. The trend of California law has always been to use the tort of conversion to deter the theft of property, whether tangible or intangible, so long as the property could be said to be “merged with a document.” Only by ignoring the evidence and refusing to acknowledge that the registration agreement is a document of title, was Judge Ware able to avoid the effect of this rule.
When judges declare themselves powerless to act, you can bet they’ll be passing the buck to our elected representatives, in lines like this: “The Court leaves it to the legislature to fashion an appropriate statutory scheme to protect dormant domain names unprotected by trademark law.” But the reference to trademark law and “dormant” names is irrelevant to the case of Sex.Com. Even if Gary had activated a Sex.Com website on the Internet, that wouldn’t have given him trademark rights. Indeed, Judge Ware so ruled six months later, when he denied Cohen’s claim of trademark protection and granted Gary’s motion for summary adjudication. And even if the website had been heavily-used, that wouldn’t have given rise to a conversion claim, either, since the theft of “eyeballs,” as dot-commers refer to web-surfer traffic, would only be the loss of business goodwill, which is clearly not subject to conversion. So what’s all that about?
It’s all about avoiding any analysis of my claim for conspiracy to convert. Remember that one? Where I alleged that Cohen and NSI had collaborated to help Cohen steal the name? Well, I had found some evidence of that, when Cohen testified that he personally talked to David Graves, who told him what needed to be put in the forged letter in order to arrange a transfer without contacting Kremen. I had that testimony, but of course I needed to question Graves. I noticed his deposition and filed a motion with Judge Ware to compel NSI to produce him. I told the Court in my motion that David Graves had received the fax directly from Cohen, and that Cohen had previously talked to Graves to find out what information he should put in the forged letter. I never got to take Dave’s deposition, because Judge Ware’s opinion neatly undercut the necessity of looking at whether Cohen and Graves conspired to convert the name. Indeed, even if they did, under Judge Ware’s ruling, it wasn’t important.
The judge’s arguments in support of the ruling -- avoiding imposing liability for a routine function, not increasing the costs of registration, and leaving the issue to the legislature -- don’t justify the denial of Gary’s claim for conspiracy to convert. Liability for conspiring with Cohen would not result from the performance of a routine function. Theft is not routine. Holding NSI liable for agreeing to accept a forgery would not create any “slippery slope,” at least in the way we usually consider these things. It would fairly impose liability for an out-and-out theft. Judge Ware’s opinion for NSI is most notable for its failure to discuss electronic documents and the unexamined contents of the Cohen-Graves communications. However, that hardly makes it unique in the annals of legal history.