It wasn’t hard to put on a brave face for Karl. They met almost exclusively by computer, after all. In fact, Paul was in a bad way. He had talked Billo and Schwenk and Karl and others into following him to NetCentric, and he had let them down. He had reason to feel depressed, in the ordinary sense of the word. But this was something different. This was the other side of mania, despair out of all proportion to whatever brought it on, to any apparent cause. It didn’t allow him to feel that he’d simply had a setback. It insisted on his worst fears. Maybe I’m not supposed to be a VP of engineering. Maybe I’m just incompetent.
The undercurrent of such thoughts was composed of gloom and lassitude and fearfulness, at times panic, and it felt like something deep and wide moving inside him, like a chemical flood. He tried to hide his distress from his family, but of course there was no hiding it from his wife. She did all she could to help him. She was always sympathetic, struggling to comfort him. But he felt sick beyond all help, a state of mind that all but guaranteed the fact.
He wasn’t taking lithium, nor did he go back on it again. Many nights he couldn’t sleep for more than a few hours. During some he couldn’t lie in bed. A bed seemed too insubstantial. While his family slept, he would climb out and crawl across the floor toward a window, crawling because he felt the need for something solid under him. He’d lie there for hours, watching for the sun to appear above the sill. Afterward, he would remember telling himself, This cloud’s going to lift sooner or later. It may take a year, but it will lift. The sun was the evidence he counted on. Even though on some nights the wait seemed perpetual, one knew after all that the sun would rise and then there would be forward movement in the world, promising relief.
Paul had long practiced his own form of adult education, partly trying to make up for his truancy at Boston Latin but mostly out of a curiosity that ranged from rap music to international medicine to fine arts. He had read about bipolar disorder ever since receiving his diagnosis, and he had determined that he wasn’t one of those especially prone to suicide. He knew he’d never kill himself. In the past when he had lain awake like this, staring out windows, he had sometimes summoned up an image drawn from high school English class. He was Gulliver imprisoned by the tiny Lilliputians, laid out on his back, tied to the ground with a web of tiny ropes. In his mind, he would roll his shoulders from side to side, and eventually, after long exertions, one of the ropes would snap, then another and another.
One day up in the attic, memories of playing xiangqi, Chinese chess, floated into Paul’s mind. The World Wide Web hadn’t existed when he had first discovered the game at UMass. He looked to see if you could now play online and he found a xiangqi club, not a website, just a link to a program that would let two people play each other. It had been created by a couple of Chinese graduate students at Boston University.
He wrote to Karl, asking for help in setting up something much better. His emails were buoyant. “Xiang Qi (chinese chess) is the most popular game in the world, with an estimated 500,000,000 people who know how to play it,” he wrote. “I think there’s a very good chance that with little work, we could setup the #1 web service for the #1 game in the world.” It would also provide a public service: “I’m eager to have westerners learn about this game—anything that can increase cultural exchange is a good thing.”
Karl’s reply was guarded. What had happened to Paul’s other ideas?
Paul replied, “I decided that I work too hard in general, and deserve a sabbatical before launching into something full force.” Setting up xiangqi for the Web would be “a lower impact effort,” and almost certainly turn out to be “a neat little profitable venture.”
All this sounded like the optimistic Paul, the only Paul Karl knew. While writing this way, Paul could believe he was on the verge of getting back in the game of software commerce. But inevitably such a cheery thought summoned up its opposite, and he would think of how far he had fallen—from executive suites and million-dollar stock options and hundreds of people reporting to him, to a lonely seat in an attic, chasing fantasies.
And yet this new adventure had intrinsic appeal. The xiangqi board and pieces had been for Paul an introduction to an ancient and artistic culture, and they still seemed wonderfully fanciful.
A river divides the two halves of the xiangqi board; crossing it, pawns acquire added powers. There are two kings, but the rules confine them and their counselors to their respective palaces, a section of squares marked off on opposite sides of the board. The horse, roughly equivalent to a knight, is said to be “hobbled” when certain configurations of adjacent pieces prevent it from moving. There are chariots and cannons—the Chinese name for the cannon sounds like “Pow,” which delighted Paul.
Xiangqi was like a magic carpet. While he rode it, the dreary attic disappeared. Long days and nights were shortened. Over the next two months, as New England’s springtime arrived—invisible through the slits of windows under the eaves—details of the website assembled in his mind. It was like a place in a dream, like a vision he received, a little pleasure dome for modern lovers of xiangqi. In this vision, members of a Worldwide Xiangqi League would come from near and far to gather at Club Xiangqi. They would sign up, gratis, and choose their playing names, and then each would receive a password and a special league email address for staying in touch with members and administrators. Each would also get a special webpage. You could open yours and look at the list of all your past games and order up replays. Looking for a game? You’d enter the lobby and take a look at the names and ratings—from novice to expert—of other unpaired players, including “robots,” xiangqi-playing programs. When you found an opponent, you would order a game room, each equipped with the finest of boards, colorfully painted. On the screen, the pieces, disks three-dimensionally rendered with their Chinese characters painted in red, were going to be imbued with the vigor that players like his Vietnamese friends at UMass had given them. When you moved a piece, it would bounce to its new place, emitting the champagne-cork pop. When you took one of your opponent’s pieces and removed it from the board, that slap-slap of wood on wood would resound as if to say, “Take that!” You could play unrecorded practice rounds, or, if you only wanted to watch, you could go into the lobby, consult the list of games in progress, then enter your chosen room. Trusted players, “coaches,” would have commands at their disposal for curbing bad behavior. A “Ban” command would allow them to expunge the chronically, egregiously obnoxious.
When Paul left the xiangqi palace and looked around his attic, he would be reminded once again that only a month ago he’d been the VP of a hot new tech start-up. But at times up there under the rafters, transported into the realm of xiangqi, he became a leader with an idea for an enterprise and a little team to enact it. At first this team included the Boston University grad students, but eventually they had to leave for paying jobs, and then the team was only he and Karl. Periodically, Paul would write to Karl that he was courting financiers, assuring Karl that someday they’d get paid for creating this site. One time, Paul offered Karl a token payment of a thousand dollars, Karl wrote that it wasn’t necessary, Paul sent the check anyway, and Karl wrote, “Gee, thanks.”
Karl had left Interleaf and had another paying job. He had some time to spare, and he’d always liked to work with Paul. Paul was so energetic and generous, his management style so different from the bureaucratic, rule-bound norm. By summer he and Paul had put together a functioning, though rudimentary, website. Karl wrote the administrative programs. Paul, meanwhile, was planning to write new foundations for their site—a big coding job.
In better days, Paul had bought a summer house in Hull, larger than his parents’ cottage and situated on the shore of Massachusetts Bay. In mid-June he and his wife and two small children moved there for the summer. For Paul and indeed all the English family, Hull had always felt more like home than Perham Street. Hull was the place where he and Danny had done their most inventive mischief, the place where he had felt a part
of a truly happy family. He knew the history of the town and of the harbor islands in its panorama. Every upstairs room of the house commanded lovely views, boats on blue waters, the grand old Boston Light, and in the distance to the west across Boston Harbor, the towers of the financial district. There were times of relief this summer, when for instance he took the ferry to Boston and returned to find his infant son and daughter and wife waiting for him at the terminal in Hull, and his heart lifted. Or the times when he made chocolate cake, his daughter standing on the counter covered in flour. But the main escape from the gloom that lingered in Paul’s mind lay on his computer screen.
In moving to Hull, Paul had traded his attic office for a basement room. It was barely large enough for a desk and a couple of tables to hold computing gear. Down there, all that Paul could sense of the oceanfront world came through a small window near the top of one wall, high above eye level, which he had installed to replace an even smaller window with a rotten frame. The room had a concrete floor, usually damp, and the place smelled musty, like the converted garage on Perham Street. There at his desk in the basement, with just a patch of sunlight or moonlight in the window above his head, the dank little room seemed to him like a fair representation of the feelings that kept returning, his interior darkness made tangible.
Emails to Karl still read like descriptions of how Paul wished he felt. He and Karl were going to learn everything “soup to nuts” about what it took to make a good website. Theirs was sure to be a hit.
The sound in Paul’s mind was much more tentative. The challenge he had set himself, and had postponed for some months, was to rewrite from scratch two large programs, the new foundation of code for the xiangqi gaming palace. It had been a long time since he’d done a coding job this large and complicated. Could he do it? Can I create again? I’m not sure I can create again. But when he sat down at his computer in his basement office, on a day in July, he knew—it was reflexive—just how and where to start.
Some coders draw diagrams of their programs before they begin to write them. Paul had never felt the need. He’d always been able to hold a large, complex construction in his mind. His picture of the two programs was like a map of all their parts and the connections among them. A kinetic map. He loved populous Manhattan avenues. Imagined from above, they were like an efficient program, he thought, the crowds moving in opposing currents and crosscurrents that never quite touched, that seldom interfered with each other.
Donald Knuth once said, “Computer scientists seem to have an uncanny ability to jump between levels of abstraction—to see things ‘in the small’ while at the same time seeing them ‘in the large.’ ” Paul could still do this. He could still write something like a cascade of coded routines, each hanging from the one above, like a Calder mobile.
As the programs took shape on his screen, he felt more and more nearly obsessed with this odd adventure at every level of abstraction, as in the old days at Interleaf and before Interleaf. He still lay awake in the middle of many nights, but he wasn’t huddled on the floor waiting for the sun. He was lying in bed thinking about the code he had written and the code still to write, or he was down in the basement with the lights on, writing it. This wouldn’t be the road to mental health for most people, but for him it felt more and more like a way back, a new beginning.
By September, when he and his family returned to Arlington, the bulk of the two programs was written: about 10,000 lines of code, which, if printed out, would fill about 170 dense and very neat-looking pages. There had been much more prodigious feats of programming, but this was a large job for one person to have accomplished in about two and a half months.
I still have it, he told himself. I still have my chops.
The Club Xiangqi caught the attention of a few companies, including Yahoo. One of its founders came to Boston and talked about buying both the site and Paul. Yahoo made him a handsome offer, and Paul wanted to accept it, but the company was situated in California, and he couldn’t move there and leave most of his extended family. For a time he thought he was slipping back toward despair. The world was going mad for people who knew how to create things for the Internet, and for a while he felt stuck outside looking in. He still had money in the bank, but it was ebbing. He worked briefly as a part-time programmer for a small company, earning $80 an hour, thinking again of how far he had fallen, thinking, as he sat in a cubicle making boring repairs to badly written code, My normal self would rewrite all of this.
But his gloom didn’t last long this time. Soon after he turned down Yahoo, he had a chance encounter, on a Massachusetts golf course, of all places—he could scarcely play the game—where he met a man who was working for The Boston Globe, managing the paper’s online store. Soon afterward Paul had a deal to improve the Globe’s website, to make it a place where visitors could easily buy all sorts of Boston-related stuff, such as T-shirts and snow globes and sports team paraphernalia. The contract was for $50,000, a contract for a product before Paul had a company to build it. He had a name, though—Boston Light Software.
If all went well, he’d assemble a great team and they’d build Boston Light into a website creator that would serve all sorts of giant corporations, once they’d done their first job for the Globe. He imagined that he and Karl would make good use of their Xiangqi experience, not the actual code they’d written but what they had learned about creating a complicated site on the Web.
He felt proud of what he and Karl had done, constructing out of the ephemeral stuff of code what seemed like a real community, indeed a real place where tens of thousands of people all over the world were hanging out electronically and competing and learning the game of xiangqi. He liked to go to the club himself and enter a room to watch a game, and if a player in Malaysia made a bad move, it was as if Paul could hear his fellow spectators—one might be in Japan, another in China—murmuring: “Can you believe he did that?”
Paul felt that Xiangqi had saved him from despair. But he had been spending fifty to sixty hours a week on the Xiangqi League, and none of his dreams of making money from it had materialized. He had two houses and a wife and young children. He had taken a step back to a former self. He had programmed his way out of depression. Now it was time to go back to work.
8
When Paul founded Boston Light—in 1998, three years after he quit Interleaf—impermanence had become the norm, new inventions displacing ones that only yesterday seemed new, computer-related businesses forming and dying, flourishing and dissolving and being bought for their patents, or disappearing, as Interleaf was soon to do, into other companies. The latest craze now was business on the Internet. Part of the promise lay in the old idea of making money through advertising. The medium was new, though, and so were the terms for it: e-commerce, information technology, the tailored Web experience, dot-coms, all amounting to nothing less than a New Economy.
Virtually all the dot-coms listed their shares on the NASDAQ exchange. That index doubled between 1999 and March 2000. One estimate has it that in the midst of the boom the value of the 371 publicly traded Internet companies amounted to $1.3 trillion, about 8 percent of the value of all the stock traded in the United States.
To take another measure of the era, the number of venture capital firms—many “VCs” specialized in financing new dot-coms—also nearly doubled between 1996 and 1999, and the capital they had to spend grew threefold, from about $50 billion to more than $160 billion. In retrospect, these amounted to clear symptoms of a bubble. Some people involved in the financial markets saw trouble coming, but in finance as in technology, being right too soon is as punishing as being right too late. Even some wary investors felt they had to follow the herd or risk a fatal loss of reputation, or at least of opportunity.
In the context of that period’s economy—manic, not just hypomanic—Paul’s Boston Light was a model of sanity. He had a paying customer and something like a business plan—to make handsome and reliable websites for all sorts of companies. He didn’t even t
ry to raise money from outside investors. Making software didn’t require anything like the capital involved in manufacturing. He figured he had enough to open the doors of Boston Light with the $50,000 from the Globe and what remained of his savings.
He rented an office in Arlington, the cheapest space he could find. It had no cleaning service and the heat was unreliable, but it cost only $8 a square foot. He fitted it out with used furniture and bought some new server computers. He recruited a team of fifteen, a couple of smart young businesspeople, support staff, and programmers. Most of these were old colleagues—such as the versatile programmer and manager Jim Giza, and also Karl, who would take care of the smooth operation of the company’s systems. It was a big moment for Paul when Billo agreed to abandon his little start-up. When Schwenk and Rago followed, Paul felt he had the team he needed. He couldn’t afford to pay them much, but for Billo and others there was the promise of adventure and money to come.
When Paul called asking Brenda to join Boston Light, she told him, “Yes. If I can work part-time.” Paul said she could, then put her in charge of creating the UI for the Globe’s website, a more than full-time job. But to be working long hours among old friends, at communal tables in one big open room, felt like a return to what she thought of as her “glory days.” Karl was one of those old friends. She remembered asking back in Interleaf days why he kept wearing the same old wool sweater with holes in its sleeves, and Karl had told her his mother had given it to him and he was damned if he was going to replace it. It made her happy to see that he was still wearing it when she saw him again, for the first time in years, at Boston Light.
Paul made Karl a co-founder of the new company and endowed him with about a quarter of the stock, about half as much as Paul took for himself.