In the small hours of the morning on May 7, 1999, Wang Zhidong lay awake, worrying about the usual. Missed deadlines. Server crashes. Post-ship bugs. But he never contemplated the real-life nightmare that hit at dawn.

For the previous six months, he had been busy integrating his former company, SRS - the largest Internet and software company in China - with Palo Alto-based Sinanet, the primary American and Taiwanese portal for the world's Chinese population outside of China. Once negotiations began last November, it took 32 days and many sleepless nights to iron out the details of this East-West alliance. But the promise of reaching upwards of 100 million people by 2003 was too great to pass up. And while every new Net-era company gets hyped as revolutionary, Sina.com, as the new multinational is called, really is. It connects Chinese speakers worldwide and, as such, can unite them socially, culturally, politically - even commercially - as never before.

In the months following the merger, Wang was confident, but the late-night worries were inevitable. Mergers are messy, and Sina's involved the high-wire act of integrating the Chinese and US executive teams. The alliance also brought challenges to Wang's leadership, and debates raged about how best to position the company for its upcoming IPO.

And then, at 6 a.m., American jets bombed the Chinese embassy in Belgrade, Yugoslavia, killing four Chinese staffers and maiming two dozen others. Following the mind-boggling cock-up, justifiably outraged Chinese citizens took to the streets of Shanghai, Beijing, and other cities for government-sanctioned protests against the Yankees. Marching within a mile of Sina, Beijing University students railed against the US, while online they posted sentiments like "Pay blood debts in blood." This was not part of the business plan.

There wasn't much time to reflect on the fact that Sina had just joined forces with an American company. Moreover, "a different truth hit us," says Yan Yanchou, Sina's R&D chief and one of China's most respected computer scientists. "People in China were reading real-time news from Belgrade - talking to one another and to others, whether they were in Beijing or Paris or New York. Westerners cannot understand the importance of this. For Chinese people, this was historic." Indeed, the Belgrade fiasco proved that Sina.comhas a big role to play and, ironically, that Wang had been right to worry about server crashes. Within three hours of the tragedy, there were 7 million pageviews at Sina.com, four times a typical full day's traffic. "We would reset the servers, and three minutes later the limits would be reached again," Yan recalls. There was no letup for a week.

In his modest office at Sina's drab headquarters in the district of Zhongguanchun - Beijing's Silicon Alley - Wang, clicking away at his Legend PC, looks up to see Yan enter the room. The two men briefly confer; they've been quite comfortable with each other ever since meeting at a Chinese Windows convention in 1985.

After jotting down the answers to a couple of questions, Yan shuffles down a dingy hallway and enters a low-ceilinged room, where tables are pushed together to form a U. He sits down for a box lunch amid a dozen PCs, cables snaking underfoot. Yan pushes back his silver-framed glasses and pulls up the pant legs of his khakis before reflecting on Sina.com'sggering potential.

While the Net still isn't available to China's masses, the number of users is doubling every six months. In 1997 there were 200,000; at the start of 1999, there were 4 million; and there will be at least 7 million by 2000, says Rajeev Kupta, Goldman Sachs' Internet analyst in Hong Kong. By 2003, some 35 million Chinese will be online, according to BDA, an independent Chinese research firm. BDA's analysis also shows that the actual number of Chinese users is probably four or five times that figure because several people use each registered account. This means there could easily be 100 million-plus users by 2003. Either way, China will have more Internet users than any other Asian country - including Japan - by 2004. By the following year, it will have more desktop surfers than any country but the US.

Sina.com already has more than 1 million Chinese-speaking users, and the site is registering 4,000 more every day from China, Taiwan, Hong Kong, and throughout the rest of the world. Wang says that the company expects 8 million users within a year, and 15 to 20 million by the end of 2001. The number of Sina customers in Taiwan recently shot up when Sina bought a large stake in PC Home, a profitable electronic newsletter that earns a quarter of the entire take of Taiwan's online advertising revenue. The numbers in the US and Hong Kong grow steadily, and the numbers in mainland China are expected to jump astronomically as the cost of Internet access diminishes.

But big numbers, Yan says, don't capture the story the way the response to the bombing does. Even as the US offered its lame excuses (old maps!), millions of Chinese, inside and outside China, gathered to read the news, console one another, and witness the extent to which the Net can open up China.

"To go from no exchange to such a vibrant exchange is profound," says Yan.

"China has been held back because of a century of stagnation while the world moved ahead," Wang adds. "The Net provides a way to catch up to, and maybe even surpass, the technologies from the West."

Not coincidentally, Sina has created a buzz in the investment world, too. Since the merger was announced, Sina.com's has been intensely anticipated - especially after last summer's successful offering of China.com, a virtually inconsequential competitor (at least until AOL became a partner this fall). China.com more than tripled on its first day out, then settled back at a level well over double the offering price, even though the site's ranked 39th for traffic in China. In the Hong Kong-based IT Daily, Neil Taylor wrote that everybody save its founder "seems to admit that China.com'stent sucks." By comparison, Sina is ranked, based on traffic, as the top destination for Chinese speakers, both inside and out of the People's Republic.

Still, Sina doesn't have the market cornered: It has at least two rivals to take seriously - Chinese Yahoo!, which has roughly half Sina.com'sffic, and Sohu, the number-two China-based site. Sohu was founded in 1997 by Charles Zhang, who wrote the Sohu business plan while he was a physics student at MIT. He showed the plan to Media Lab director Nicholas Negroponte, who loaned Zhang some seed money and introduced him to other potential investors. They agreed to back Sohu, and Zhang returned to Beijing to build the site, which squandered an early lead in market share.

In addition to ad revenue, Sina is looking to generate income from ecommerce. As Michelle Yeh, Sina's senior marketing director, says, "We find that people come to us to send flowers to their mother in Taiwan before they go to 1-800-Flowers, because they trust that we will do the job better in a country we are part of." Many who visit Sina from outside China use the site to buy hard-to-come-by products, such as mung cakes. Yeh and company expect huge jumps in sales on the mainland when better payment systems are developed. (Currently, few Chinese have credit cards, but IBM and the Bank of China are developing a system that will allow Chinese users to pay with debit cards online.) One Sina offering that has been incredibly popular is its dating service, Club Yuan (yuan means "destiny"). The company doesn't have to look far for testimonials: Three of its managers met their wives through the service.

Wang Zhidong, 31, built his company much as he built his career - from scratch. Raised in a poor family in the rural Guangdong province, he studied hard and was accepted by Beijing University. While in school and after graduating, he worked as a freelance software engineer in local tech companies before being hired in 1989 by the Beijing Founders Group, a branch of the university. Then he started a company, Chinese Star, which produced a program that let Chinese PC users run Western applications.

__ Sina.com already has more than 1 million users, and the site registers 4,000 more every day. __

When Wang left C-Star, IBM, Microsoft, and a number of state-owned enterprises offered him a job. "He couldn't just take a job," says his wife, Liu Bing, who is now Sina's office manager in Beijing. "He wanted to do something important for China. He is very patriotic, believing that we must work to build China and help the Chinese people." Liu says she tried to discourage him ("I wanted him to live a calm life with regular vacations"), but Wang didn't listen.

Yan Yanchou, meanwhile - the son of a government official accused of counterrevolutionary activities and jailed in Mao's China - attended a junior high school for workers, peasants, and soldiers in Hunan. He so excelled that by 1978 he was eligible to study at the National Research Institute in Beijing. He taught himself hardware design and programming, then created the first and second computers used in Chinese industry - a payroll system for a private company and a computer that controlled a flour mill. In 1983 Yan was given one of the two IBM PCs available in China at that time. He created a program, CC-DOS, or Chinese Character DOS, that bridged the considerable gap between Microsoft's code and the Chinese character system.

After two years, Yan left the institute for a private company, "to go swim in the sea," as he puts it. He was working for an electronics firm, writing software for power plants, when he met Wang. A year later, the two men decided to go into business together. They were turned down for financing by several companies before the Stone Group - a Hong Kong-based importer of typewriters and Compaq PCs - funded their venture, which they named Stone RichSight, or SRS. They set up shop in an abandoned school building.

Three years before Microsoft released a Chinese version of Windows, Wang and Yan developed a program called Rich Win, which translated standard English Windows to Chinese, so it could read and display Chinese characters. Rich Win became essential to Chinese PCs; eventually, 80 percent of the PCs in China used it. Though as many as 2 million copies were bootlegged, SRS sold more than 3 million, making it the most successful software company in China.

In 1995, Wang watched intently as the Net exploded in the US. Though China had relatively few Net users, he saw how it could be "China's salvation." He began what he called Project Surf, the transformation of SRS from a software company to an Internet company - in spite of opposition from most of his managers and investors. (When one of them heard about Wang's idea, he renamed it Project Stinky Fish.) In March 1996, SRS released its first Net-related product, a seamless upgrade called Rich Win for the Internet. As Rich Win did for PCs, this second-generation software made the Internet Chinese-friendly. It did so by letting Net browsers recognize and display Chinese characters. It includes a cursory dictionary so a user can click on English words on the Web and get their Chinese equivalents.

Next, SRS launched a Web site called SRSnet.com to support its software. There was also a BBS on which, Yan says, "we thought our users would discuss bugs and relate technical tips. Instead, they talked about football, gaming, and the news." The topics ranged from current events to a discussion started by venture capitalist Bo Feng, who arranged a round of financing for SRS in 1996. (See "He's Got Guanxi!" Wired 7.02, page 122.) The topic's irresistible title? "Modern Youth and Existentialism."

At first, the number of SRSnet users was inevitably small - few people were online in China - but SRS got a large percentage of those who were connected. To make sure the site didn't run afoul of the government and kept growing apace, it walked a fine line of self-censorship.

"We told users of the BBS, 'Do not post expression against Chinese law or the government,'" says Yan. "If anyone posted these contents, we deleted them in order for our Web site to survive."

To run the site, Wang hired Wang Yan (no relation), who was educated in China before earning a law degree from the Sorbonne. The firm's flashiest dresser, he was put in charge of expanding the site in many directions, using the model of AOL and Yahoo! Along with AltaVista's search engine, he put up a slick Chinese-character search engine designed by Wang Zhigang, Wang Zhidong's kid brother. He also added chat rooms, forums, news, and multimedia information about Chinese culture - everything from movies and music to traditional theater.

SRSnet quickly became the most popular portal in China. In 1997, when the Chinese soccer team played in the World Cup, SRS Webcast in real time and got 3.1 million hits per day in May and June - a record for any Chinese site. But in spite of this growth, by last year Wang Zhidong felt something was missing. As Yan Yanchou says, "We had the best and biggest Chinese site in the world. But our users were all in China. There was a whole world of Chinese users who weren't being served." Wang wondered: What about the Taiwanese, North American, and other Chinese communities? The link to the Chinese diaspora would turn up in a purple-walled, single-story office building in Palo Alto, California.

__ The dating service, Club Yuan, is extremely popular. Three Sina managers met their wives using it. __

Sinanet's origins fit the Silicon Valley archetype: three Stanford students, a cramped living room, and a shared, restless idealism.

Jack Hong, one of the three, was born in Taiwan and grew up in Ohio, where his father was a professor. His family moved back to Taiwan when he was 10, then returned to Ohio when he was 15. "There was a Korean in my 12th-grade class," Hong says. "Everyone thought he was my brother." Though he had never been to mainland China, he felt drawn to his Chinese heritage, partly because he felt like an outsider in both Taiwan and the US. In college at the University of Texas in Houston, he founded a band that played Chinese music. He started another Chinese band while attending graduate school at Stanford.

Studying art, product design, and mechanical engineering, Hong tutored on the side. One of his students was Ben Tsiang, who came from a prominent Taiwanese family. (Tsiang's grandfather was Chiang Kai-shek's chief of staff.) Tsiang had his sights on the Harvard Graduate School of Design. Before graduating, Hong also met Hurst Lin, who was born in Taiwan and grew up in New York before earning his Stanford MBA.

After Lin graduated, he worked as a financial consultant while Hong stayed on at Stanford for his PhD. At night, he'd invite Lin and Tsiang over, and the three would sit around Hong's living room discussing their futures. Deciding they would work together in some business, they considered several.

By coincidence, Lin and Hong attended a presentation by Jim Sha, one of Netscape's founders, in late 1994. Sha was recruiting for Mosaic, the Web browser company that would become Netscape, and he gave the audience of Stanford students a preview of things to come. "He was showing these hockey-stick charts of the growth of the Internet, and we thought, 'Yeah, right,'" Lin says.

Soon after they missed the Netscape boat, Lin says he began to "get it." Hong, too, got online and into the Net's implications for distance learning. Then Tsiang came up with the idea of a Chinese site.

"We didn't have a competitive advantage in education services," Lin says. "But we had one in a site for and by Chinese. We were Chinese. We knew the needs of Chinese-Americans."

The more they discussed it, the more they became inspired. "Chinese kids in the States lose their Chinese, lose touch with their roots. I thought the Net might be a hip place for kids to read in Chinese, as opposed to boring newspapers," Lin says. "Also, there was some chauvinism thrown in: Why should the computer world be dominated by English?"

In April 1995, they made their move. For their company's name they rejected Great Wall ("too trashy," says Lin) and China.com (it didn't include Taiwanese, Singaporean, and other Chinese people). They settled on Sinanet and, courtesy of Stanford computers, launched. Hong used his art and engineering skills to create software that rendered Chinese characters on the Web. For content, the new partners called Chinese embassies and consulates, both of which receive a daily newsfeed from China's state agency. "Sure," Lin admits, some of it "was official news and propaganda. But we had to put up something."

Every morning at 6, one of the Sinanet founders uploaded the news from New York via a 14.4 modem. It took about half an hour. If the modem disconnected, they started over. They sought other news sources and hit pay dirt when the second-largest Chinese-language newspaper outside China agreed to let Sinanet have its content free. (The publisher viewed it as good publicity.)

"Before us, news from China was always really late - 16 hours late if you lived in San Francisco, where there is a large Chinese population and some Chinese newspapers, and as much as three days late in other parts of the country," says Lin. "We were able to beat the newspapers. We became a big hit among Chinese students everywhere." Hong says serious news from China was the most sought after, followed by entertainment and cultural news. There was also a hearty audience for tabloid stories - grisly murders and sex scandals.

Sinanet put up chat rooms and bulletin boards, too - "self-generated content," says Hong. Chinese students on college campuses spread the word to their families, and Sinanet's traffic steadily grew. To pay for the site, they began selling Chinese food online. "At least it solved our lunch problems," Hong says. Stan Saih, the founder of Acer Computer, heard about the site and offered operating funds and servers and computers - which became particularly useful once Stanford kicked Sinanet off its system.

As Sinanet's traffic grew, the founders settled into a real office and wooed a CEO, Daniel Chiang, then president of Trend Micro, an Internet virus protection and security software company. With Chiang in charge, the three founders took VP roles and focused on supplying editorial content through deals with 200 providers, beefing up the arsenal of servers, and selling advertising to a variety of companies interested in reaching the Chinese-American community. But their plan lacked something: China. And, Lin adds, "As a bunch of badass Chinese-Americans, we could see pretty clearly that there was no way we could parachute into China."

Just as Sinanet's trio of founders and new CEO researched a move into China, SRS's Wang Zhidong and Yan Yanchou contemplated a way to expand beyond China. The two companies first got together when a pair of Wang's investors brought him to visit Sinanet. Both sides immediately saw the logic of a merger.

"We had a common willingness to be the best in the world and a common willingness to be the biggest in the world," says Yan. Wang, though, was still unconvinced. Over a weekend in November 1998, he sequestered a group of advisers in a conference room at the Dragon Club in Beijing. For 40 hours, nobody slept. Finally convinced, Wang decided to go forward with the merger. Sina.com was born.

In Beijing, Yan Yanchou remained head of R&D while Wang Yan led an ad hoc team - made up of staff from both companies - in an effort to combine the sites so the look and feel was the same whether one logged on from the US and Europe (sina.com), China (sina.com.ch), or Taiwan (sina.com.tw). Hong Kong (sina.com.hk) was added later. Though the design was now integrated, political and cultural differences dictated that the sites remain unique. Local editors in each country were in charge of selecting content relevant to their locale and supplementing it with articles on local affairs. Some choices were obvious - entertainment listings, for example - but the Chinese and Taiwanese editors had to make others as well, scouting for content that wouldn't offend government censors. The initial political caution has given way to discussions that are, by Chinese standards, freewheeling. Increased usage has brought greater freedom of expression.

__ Sina's September shake-up confirmed that while it is a global company, its top priority is China. __

In the meantime, Wang, who became chair of Sina.com, and Chiang, Sina's vice chair, prepared for the IPO by bringing in Silicon Valley veterans to run the company and sit on the board of directors. They convinced Daniel Mao - an investment banker who helped secure funding for SRS and had become Wang's trusted friend and adviser - to come aboard as COO. For CEO, they tapped Jim Sha, the man who had given Sinanet's founders their first look at the Net back in 1994. Sha, who'd been at Oracle and Intel before he helped Jim Clark and Marc Andreessen found Netscape, had since retired from "the Internet roller coaster" and planned to manage an investment fund. But after visits from Chiang, Mao, and Wang, Sha joined. "The call of China was too much to refuse," he says. "I felt obligated."

Although, under Sha, Sina's numbers kept improving and advertisers embraced the site - the Sina Mall began to rack up impressive sales - Wang disagreed with Sha's emphasis on the non-China market ahead of the nascent, but exploding, Chinese market.

Some advisers told Wang to wait until after the IPO but others supported him when, in September, he radically shook up the company's leadership. "I learned what Sina.com needed to go forward," Wang told an associate in San Francisco, "and felt the need to set the company on a clear, directed course."

Sina was already in its pre-IPO quiet period when Wang consolidated his position (Business Week called it "a management purge"). Wang replaced Sha as CEO, leaving him a small role as a company director; Mao, meanwhile, became CFO and COO in place of a Sha hiree. This change, plus a switch of lead underwriters (from Goldman Sachs to Morgan Stanley Dean Witter), reestablished a single vision and a single direction for Sina - Wang's. In practical terms, it meant that the Chinese market would take priority over other Sina markets elsewhere in Asia and the US.

Even as Sina endured its org-chart tumult and recommitted to users in China, it continued to unite Chinese speakers everywhere. On September 30, two weeks after the board meeting where Wang shuffled the deck, and right about the time these changes were made public, an earthquake measuring 7.6 on the Richter scale hit Taiwan, killing 2,295 people and causing billions of dollars in damage. Minutes after the tremor hit, people in every country of the world booted up Sina.com.

"It was as close as I could get to going home," says Mark Yang, a Chinese-American student whose parents live in Taipei. Once again, the company's servers were stress-tested as thousands checked in to get news - for some, news of family in Taiwan.

Sina's founders on both sides of the Pacific will likely strike it rich when the company goes public. And, as Daniel Mao says, some of the excitement they feel undeniably comes from being able to stop playing "the poor relation" and "level the playing field." But Sina's founders say that being there for people like Mark Yang will always be their greatest reward.

"We knew that the ideological and cultural problems were bigger than any technical problems," Hong says. "First Amendment? China doesn't have a First Amendment. Yet what brings change but openness - access to information and the ability to communicate. It's happening at Sina and it's all very gutsy and exciting. It's like a spiritual awakening."

"Suddenly, we are no longer separated by geography and nationality," says Wang. "Finally we are genuinely tied together by the Internet."