by Mark Joseph
Judd Fernandez’s Mexican mother had raised him alone in the barrios of San Jose. Surrounded by poverty and gangs, he had enough sense to look beyond the low-rider life around him to the neatly trimmed industrial parks that sprouted like mushrooms right next door in Silicon Valley. He wanted a job as a programmer, and there were jobs galore, but not for him. He was a punk from the barrio, self-taught, with no recognized merit badges and unskilled at writing sophisticated resumés. His talents went unrecognized. Rebuffed when he applied for work as a programmer at Intel, he decided to get even. He hacked into Intel’s most secure systems and stole schematics for a dozen new chips. Then, full of himself, he took his stolen goods to Los Angeles and was snared in a sting by the LAPD. Tried in superior court, he was sent to state prison for three years. Getting caught selling the goods in no way diminished his achievement of hacking into Intel and downloading their most precious files. In hackerland, that was extreme, perhaps beyond belief, and no one would have believed it if it hadn’t been proven in court.
And finally there was Adrian Hoffman whose life might have been a sequel to Natural Born Killers. He grew up in seven different trailer parks in Central Florida. His mom was a ticket seller at Disneyworld, and his dad drank himself to death when Adrian was twelve. Adrian was a totally dysfunctional human being from K-Mart America with one saving grace. He understood the relationship between trains and computers better than anyone on the planet.
The day after Judd dressed him down, Adrian disappeared for three days. Doc was ready to replace him when he realized the kid was riding subway trains and exploring the city’s private railroad.
“For God’s sake, Adrian,” Doc pleaded when the prodigal returned, “just tell somebody where you’re going and what you’re doing.”
To keep him around Nassau Street, Doc bought HO scale models of the subway trains and let Adrian build a layout in the computer lab. With gentle coaxing, Doc got him to model the subway’s signal system and use the toy trains to test his software patches.
* * *
Doc bought a huge condo a few blocks from Nassau Street in Battery Park City for them, and they had no choice but to learn to live and work together. Their constant proximity produced countless squabbles, two relationships and one breakup, but ultimately, using gentle persuasion, sensitive guidance and the discipline of an implacable deadline, Doc forged them into a unified team with a mission. By the end of 1997 the mission had taken possession of their psyches, and they gradually began to withdraw from the world. Outside relationships became difficult as the constraints of secrecy and silence made anything beyond small-talk impossible.
Judd helped by recounting how he’d learned to live in prison. “Just keep yourself to yourself and live in your heads,” he told them, and their heads became a collective head as their ethnic and personal identifications faded away. It wasn’t long before they were swept up in the urgency of their task. Two and a half years seemed like a long time, but it grew shorter every day. Nothing could be put aside or left undone.
Ronnie quickly became an expert in the city’s water supply that flowed from the Catskills through hundreds of miles of tunnels, aqueducts and reservoirs. Adrian played with his model trains and took sight-seeing trips through the tunnels and into the yards. Carolyn walked the streets of Manhattan, following phone lines from junction box to building to Bell Atlantic router. Doc bought her a truck, a hard hat and genuine Bell Atlantic ID, and she went into the tunnels beneath the streets, inspected the lines foot by foot and tested every chip with a logic probe. In a year she worked her way a hundred blocks from Nassau Street to the ConEd Systems Command Center at 65th and West End Avenue, and from there to the four power plants on the island and the huge Ravenswood plant across the East River in Queens. When she finished hardwiring the lines to the power plants, she started on the lines from Nassau Street to the Department of Environmental Protection offices in the Bronx and the subway’s central dispatch office in Brooklyn.
Bo visited the bookstores around Columbia University and bought himself an education in electrical engineering for a few hundred dollars. An ordinary computer nerd would have hacked into ConEd and stolen applications without knowing or caring about the physical systems they controlled. That wasn’t good enough. Bo had to know how and why and where, and in the end, who.
Six months after he started, Bo knew more about Con Edison than the CEO, Mr. Peter Wilcox, whose e-mail he could read although he didn’t have time. He knew where the different subsystems worked together efficiently and where they didn’t, and which sub-stations between Manhattan and the power plants were likely to blow like weak links in a chain. He’d figured out how to correct the Y2K problems and build a system that would keep the power on in Manhattan, but he had no way to trick the power company into switching over to the computer on Nassau Street. He needed secret override passwords to put his knowledge to work.
“Well,” Doc asked him one day in 1998, “can we save New York?”
“Not without help,” Bo answered. “I need someone inside who can give me passwords. Without the override codes, at the moment of the century rollover, I’m dead in the water. Find me a spy.”
“No problem,” Doc replied. “They’re all over the Internet.”
Finding an insider who occupied a key position in the power company’s control rooms was less difficult than finding competent hackers. No one had a better understanding of Y2K problems than the hardcore professionals who sat at the controls. These men and women communicated extensively via the Internet, often using pseudonyms to hide their identities. Power plant engineers, telephone system programmers, transit system authorities, and dozens more exchanged ideas and technical information in cyberspace.
Doc logged on to the Usenet and within a month had found an engineer who posted messages under the name “Deep Volt.” A senior operator at ConEd’s system operations command center, Deep Volt was convinced the power grid was going to fail. Doc started sending the anonymous engineer the locations of Y2K bugs Bo had found that were missed by the company’s own Y2K programmers. In exchange, Deep Volt provided secret documents, applications and system schematics. They had their insider, and from that moment on, the work progressed much more quickly.
As they got into it, they started to call themselves the Midnight Club. Ronnie had T-shirts printed with the skyline of Manhattan at night. They worked like dogs, and by the end of 1998 they’d ascertained the city’s true state of readiness for the coming assault. The prospects were dismal.
After spending several hundred million on Y2K, Con Edison had the most advanced remediation program of any power company in the world, but the giant utility was at the mercy of malfunctions anywhere on the Northeast electrical grid. The interconnected systems were so complex that the chances of every company on the grid making each fix exactly right were nil. Probability studies predicted failure. If the power went down, nothing else would matter.
The telephone system had the best chance of survival because Bell Atlantic had done everything they could think of at a companywide level to correct their Y2K problems, replacing embedded chips in routers and switches and checking every application for flaws. At the local level the work was uneven and inconsistent. Like every other system, Bell Atlantic was part of a wider network of interconnected systems, and the interconnects were extremely vulnerable and subject to failure.
The Department of Environmental Protection had one of the oldest, most decrepit computer systems the Midnight Club had ever seen. Like many city departments, the DEP was the private bailiwick of senior managers who dwelled in their own hermetically sealed universe. Their system had worked efficiently for decades and they saw no reason to change.
“We’ll just have to do it for them,” Doc said.
The Metropolitan Transit Authority had recognized the Y2K problem early, and the overall compliance was remarkably advanced, but the subway’s train control system was ancient, having been hardly touched since the 1950s. The advantage of runnin
g an antique was that the electrical relays were mechanical devices without embedded chips. The trains would be safe, but the problems lay in the automated train boards at the MTA’s traffic control center on Jay Street in Brooklyn. Adrian found undetected millennium bug flaws that would shut down the screens, and the system operators wouldn’t be able to see where the trains were. They’d have to rely on the MTA’s radio communication system, which was dependent on flawed chips in the radio transmitters. When the screens and communications failed, the system would come to an immediate halt.
Left alone, New York’s most vital systems would break down. In midwinter, the technological overlay that made the city work would be stripped away like bad varnish, reducing New York to medieval conditions. Projecting what would happen after that depended on one’s view of human nature. If people were evil and wicked, then looting and pillaging would be the order of the day. On the other hand, if the citizens of New York were concerned with their mutual benefit and survival, people would adapt and help one another. Doc believed he’d see plenty of both.
* * *
After two years, the Midnight Club had hardened into dedicated, militant soldiers of cyberspace. They were ready to take over and operate the deep infrastructure of the island of Manhattan, but they needed the essential Con Edison override passwords for the final switchover. With them, they had a chance to wrest control from the ConEd computers and save the island from a massive power failure. Without them, New York was kaput.
With six months left, they had almost everything they needed, but not quite. Deep Volt was unable to obtain an essential set of ConEd passwords that unlocked the override controls that switched the primary system to the backup. Without those passwords, Bo’s work would go for naught. Instead of switching to Nassau Street, the ConEd computers would switch to their own backup systems that contained the same flaws as the primary, and power would be lost. The passwords were kept on a locked PC with no connection to the outside world, and there was no way to hack in and get them.
“Your spy isn’t worth a damn,” Bo complained.
“She’s doing the best she can,” Doc said in her defense. “She’s risking her job. If she breaks in and steals the passwords, she’ll be caught and they’ll change the passwords.”
“There must be a way,” Bo insisted. “Find a thief.”
“Thieves don’t advertise on the Net.”
* * *
The Midnight Club was isolated from the rest of the city, but they weren’t operating in a vacuum. By March of 1999 news of the coming millennium bug was beginning to reach the general public. In that month companies whose fiscal year began in March started encountering problems they hadn’t expected. If no man is an island, neither is a company. The world was learning that virtually all the computers on the planet were part of one huge network that was as vulnerable as the weakest link. Firms who believed their systems were Y2K compliant suddenly found their expensive new software was corrupted by files imported from suppliers and customers who were not compliant. The technical difficulties of finding corrupted code in a maze of computer languages, lost documentation and shoddy programming became apparent. Bankruptcies mounted and panic began to creep into the financial community. Companies that had delayed the allocation of funds for Y2K because they believed the return on investment was zero realized they’d made a mistake. To their horror they discovered that no programmers were available to sort out their problems. It was too late. Every month as another round of firms began their fiscal year, the problems intensified. Billing and accounting programs went haywire, exactly the way Doc had demonstrated on Old Blue.
People looked to the federal government for leadership, but none was forthcoming. The Senate and House held hearings and experts repeated what they’d been saying for years, and the result was to call more panels of experts and create bureaucracies that argued over definitions. Every day, as the inevitable deadline crept closer, the responsible authorities argued and debated and discussed, and another day was lost. Unclear on the concept, the government foundered. Agencies such as the Social Security Administration that had been working on Y2K since 1989 learned that inadequate measures in other agencies thwarted their meeting objectives. When 2000 arrived, Social Security would be able to distribute funds, but the Treasury Department would be unprepared to cut checks. Faced with impending paralysis within its own systems, the government was revealed as powerless and irrelevant, reduced to nothing more than an obstacle and a nuisance.
* * *
Millennia, years, days, and nanoseconds are convenient ways of measuring time, and accurately measuring time is important to computers. Units of time are finite objects a computer can recognize and count, and it is the counting that allows a computer to compute. Computers count time as a way of sequencing processes. One of the first things a computer does when it is turned on is tell itself the time and date.
The millennium bug was not the only date-related problem confronted by computers in 1999. Many systems were based entirely on date and time, including the Global Positioning System, GPS, a satellite-based system of radio-navigation used world-wide by American and allied military units, civilian ships at sea, aircraft, motor vehicles and geophysical researchers. Every telephone company in America used GPS to set the clocks that ran their switching stations, billing procedures and maintenance schedules.
GPS was operated by the United States Air Force Space Command who maintained a fleet of 24 satellites, and a radio fix from three birds was sufficient to locate a position anywhere on the globe. Anyone could buy a GPS receiver, tune in the satellites and use the system. Long before the night of August 21–22, 1999, the Air Force widely disseminated the news that on that date the nature of the data beamed down from the satellites was going to change. Older GPS receivers required a new chip in order to correctly interpret the data and determine an accurate position. So many millions of craft depended on the system that failure in only one percent amounted to thousands of receivers that hadn’t been upgraded.
Despite the warnings, on August 22 the world was surprised by a wave of disastrous accidents caused by simultaneous computer malfunctions all over the globe. In foggy San Francisco Bay two oil tankers collided, polluting the harbor with millions of gallons of crude. In Rotterdam a liquid petroleum gas ship rammed a pier and the ensuing explosion destroyed the pier, a railroad yard and killed 327 people. Small boats were especially vulnerable, and dozens of inexperienced sailors who depended on the system and had never learned the basics of navigation without instruments were lost at sea.
The GPS rollover was not a millennium bug problem, per se, but it was similar and came at a time when Y2K was in the news every day. People began to look at computers as Trojan horses, and all the old fears from the early days of computing were brought back to life. In the ’50s and ’60s the huge, cumbersome machines of that era had inspired dread. Workers had believed they would be replaced by automated machines, and they were. Students had rebelled against being treated like computer punch cards, and movies like 2001 had envisioned menacing machines making war on humanity. Cartoon characters had stalked the pages of The New Yorker with signs that read, “The end is near, computers are here.” Computers were strange, threatening, and misunderstood, but over time they’d been accepted, embraced, and ultimately ignored, if never properly understood. As Y2K loomed nearer, the cartoons reappeared depicting a life-and-death struggle between Man and Machine. In late 1999 pundits began asking the question: What have we wrought? No one had answers that made sense. Newsweek and Time featured articles on cybernetically isolated colonies of computer scientists that had sprung up in Arizona and New Mexico. These knowledgeable men and women believed a disaster was on its way, and they’d sold their securities, withdrawn their money from banks and built fortresses to protect themselves and their families. As experts continued to contradict one another and wild predictions of an apocalypse crowded the headlines, people became confused and frightened. Every error on a bank statement wa
s blamed on the millennium bug.
On October 1 a severe round of fiscal year computer meltdowns delivered yet another blow to the economy. On that day the federal fiscal year began, and inadequate remediation inside dozens of federal agencies became apparent. Projections failed. Planning became impossible. Government procurement agencies suddenly couldn’t buy anything because they’d lost track of purchase orders and invoices. Military logistics systems lost control of entire supply chains, and the ripple effect sent a shock wave through the economy. Companies who depended on government orders saw the bottom fall out of their stock value. Technology stocks tumbled, and lawsuits, long predicted as the most expensive part of the Y2K problem, began to clog the courts as the real debate began on the issue of who was responsible for the millennium bug, and who was going to pay for the losses. The answer: everybody.
Despite all the publicity, opinion surveys in late December 1999 revealed that a third of the American people had never heard of Y2K or the millennium bug. People knew that many government computers were not working properly without understanding why. Millions of New Yorkers carried on with their lives, not completely unaware of what was coming, but uncertain as to its exact nature and what it meant. People who understood Y2K tried to explain it to those who didn’t, and you either got it or you didn’t. It didn’t matter. It was coming, ready or not.
Those who did get it took it seriously. A New Yorker didn’t need a degree in computer science to understand that a technological breakdown that started with a power blackout had severe social consequences. Previous blackouts had come as a surprise and prompted spontaneous looting. This time people were forewarned. Steel doors, window grates and security guards commanded premium prices. When guards became too expensive, small business owners bought shotguns and planned to sit in their stores themselves. In every borough, neighborhoods created community patrols and made plans for mutual protection on New Year’s Eve. The police didn’t object. On New Year’s Eve, the department would be stressed to maximum capability even without Y2K. In the most squalid ghettos of the South Bronx, the Colombian district of Queens, Bedford-Stuyvesant in Brooklyn and Spanish Harlem in Manhattan, community leaders spread the word: Don’t burn down your own house.