Chapter 8. CEO Huang Yimeng, No Longer an Angry Young Activist

A chapter from THE HISTORY BEHIND CHINESE GAMES, the author's unfinished second book, by BBKinG (Liu Yang). Translated from the Chinese original: 八 然而已经不是愤青的CEO 黄一孟

First published in the author's Zhihu column on August 8, 2016: https://zhuanlan.zhihu.com/p/21668259

This translation was reviewed against the Chinese original on August 3, 2026. If you spot a mistranslation, a wrong name, or a factual error, please open an issue or send a pull request.

When the interview was over, I asked Huang Yimeng: "Dash, can I really write about your days as a fenqing, an angry young activist?"

Huang Yimeng (黄一孟) was born on March 18, 1982, in Changning District, Shanghai.

At that point he was still twenty years away from founding VeryCD and bringing eMule, the client Chinese users called the electric donkey (电驴), into China.

And twenty-one years away from being expelled from university.

Huang Yimeng, the angry young activist

Huang's mother was a doctor in a hospital laboratory department. His father taught physical education at a university and was later promoted to head of the department. The family shaped him in two lasting ways.

One was that they were reasonably well off and his parents were fairly open-minded, so he came to games consoles, computers and the internet early. The other was that from a young age his father pushed him into the sports system and put him through every kind of physical training.

His father was an exceptionally cheerful man, Huang says, who played Salamander and Contra with him when he was small. In sport, though, he was extremely strict and expected a great deal.

Teaching him to swim meant repeating a single movement all day long. When he was in his second year of primary school and still very short, his father pushed him into the basketball team, so that while the other children went off to play after school he had to stay behind and practise shooting.

In an era before the NBA and Slam Dunk had caught on in China, all this training was dull and joyless to him, and inwardly he began to resist that way of being brought up.

The awakening of a sense of self

Huang says that until his last year of junior secondary school he drifted along in a daze. His marks were below average, he thought of himself as thoroughly ordinary, he did not know what he liked, and he had never given the future any thought.

Then one day a research team from East China Normal University came to the school and ran an intelligence test on his class. Working through it, Huang simply felt the questions were not particularly hard. He finished and handed the paper in.

Afterwards a teacher took him aside and said: you are this clever, if you put all of it into your schoolwork your marks would be much better.

That was how he found out that he and one other pupil, one of the strongest students in the class, had scored the highest in the test.

From that moment Huang Yimeng began to realise that he might be a little different after all.

He thought back to programming on a learning machine (学习机) as a small boy, to the games he had played in the early 1990s on his father's laptop, and to Pascal, the structured programming language he had come across at the Changning District youth science centre (长宁区少科站) while at junior secondary school. Somewhere in all of that he sensed a direction.

Something clicked, Huang says. It coincided exactly with the rise of the internet in China. He learned about Bill Gates and other well-known figures in technology, and learned that they had dropped out of education and gone on to found successful companies. For someone who hated school that settled it: he would be a programmer. He began teaching himself programming from the internet.

His father, seeing that he was not lost in games but was genuinely learning something on the computer, did not stand in his way.

The senior secondary syllabus, though, left him deeply at a loss. Why learn all this material with no practical use? His dislike of school grew deeper.

Naturally, he did not do well in the 2000 national college entrance examination.

What he found both funny and painful was that the sports training that had made him so miserable turned out to matter enormously. Ranked as a national second-class athlete, he entered Shanghai University to read business administration as a sports-talent entrant.

Only then did he understand that his father had been playing a long game for eighteen years, Huang says drily.

At university he found that nobody was watching him at all and he could do whatever he liked, and his internet skills came on in leaps and bounds.

In his first year Huang Yimeng was already making decent money building websites, and he wrote his own forum software in PHP. Young and full of himself, he came to believe there was nothing on the whole internet of that time he could not build.

The birth of VeryCD

In 2001 the Huang family were among the first households in Shanghai to have broadband installed, which meant they could download all kinds of material at speed. Most people had no such convenience, yet the appetite for downloads was growing all the time.

So he came up with a business model: a catalogue website that listed everything. People would simply tick the items they wanted and generate an order, and he would burn the content onto discs and post them out.

The original idea was simple, Huang says. He would sell CDs that way, and that is where the name VeryCD came from.

While he was busy gathering material online to sell on disc, a new technology upset the plan. It was peer-to-peer downloading, and he found it far faster than burning discs.

So Huang changed tack. He set up a VeryCD forum, put his catalogue of collected material there, and turned the computer at home into an FTP server so that people could upload and download material through the forum themselves.

Word of mouth soon made VeryCD the favourite forum of Shanghai University students. It grew busier still during the SARS outbreak of 2003, when many campuses were sealed off and university students all over China had nothing to do.

Just as Huang Yimeng was enjoying his site's popularity, an incident that shocked the whole country took place at his university.

The Shanghai University secret filming incident

Anyone who was online at a Chinese university back then probably still remembers it. A few students pointed a camcorder at the teaching block opposite and happened to film a couple having sex in a classroom. The camera was a long way off and their faces were not even visible. The video was put online.

Today nobody would think twice about it, but in 2003, in the early days of the Chinese internet, the video set off an astonishing chain reaction and was uploaded to and downloaded from countless FTP servers across the country.

Among them were several FTP servers inside Shanghai University itself, and Huang Yimeng's VeryCD forum.

The university moved quickly. It suppressed the news, handled nothing openly and wanted the whole thing wiped away fast. The FTP servers on campus deleted the file under pressure from every direction. Huang Yimeng flatly refused to.

Huang says he was simply a fenqing, an angry young activist, at the time. The footage had nothing to do with him and he had not uploaded it, but he objected strongly to the way the university concealed what had happened and used administrative force to bury it. That kind of thing, he thought, should not be happening on the new internet.

Too young, too naive. Huang Yimeng paid for his youth.

One day the campus police station called him in. An officer told him he had distributed obscene and pornographic material, issued him with a bail order pending trial on a charge of criminal insult, and confiscated the computer at his home.

That was not the end of it. He was expelled by the university along with the students who had shot the video.

It devastated Huang's parents. They had spent their whole lives inside the state system and could not accept an "accident" like this. Huang himself seemed hardly to mind. He was, in fact, rather pleased.

Being expelled, Huang says, was an unexpected windfall. He had wanted to leave for a long time, but his parents would never have agreed to him dropping out of his own accord. Now there was nothing they could do.

On the night he was expelled Huang Yimeng was extraordinarily excited. The thought that from the next day he could do nothing but what he wanted kept him awake.

When he got up the following morning he set about turning the VeryCD forum into a real website.

After that it moved fast. Traffic kept climbing, and it climbed so fast that one shared hosting provider after another threw him out. He had to find a server of his own. But he had no income, only the 2,000 yuan the forum's users had donated, which was not enough to buy a server, so he rented one instead.

So Huang signed up to an advertising network, which brought in a little money, and at the same time went out and found himself a job.

In 2003, through the webmaster of PHPcn.net, a site he visited often, he got work as a web programmer for 4,000 yuan a month, more than either of his parents earned, which gave them some small comfort.

Traffic grew, advertising revenue grew, and every time Huang had saved enough he rented another server. VeryCD went from one rented server to ten in short order.

It barely broke even, but the sheer speed of that growth showed Huang Yimeng an opportunity.

Four months later he resigned. He and three others, Dai Yunjie (戴云杰) among them, rented a place on Tianshan Road just big enough for four desks, and formally went into business.

Dai Yunjie is an interesting figure in his own right. He and Huang Yimeng had been at Fudan High School together and lived not far apart, but they never crossed paths, because Dai was famous for being an outstanding student while Huang was famous for poor marks and a gift for computers.

It was in the summer after they left school, when Dai Yunjie came to Huang's home to learn how to build websites, that the two discovered how keenly interested they both were in anything new. Dai later got into Shanghai University as well and worked on VeryCD alongside Huang. They have been friends and colleagues ever since.

In the summer of 2004 VeryCD's servers were getting busy.

The downloads used peer-to-peer technology, so the data itself did not pass through their machines, but traffic to the site at the evening peak kept rising and they could not afford to rent any more servers. They had to solve load balancing some other way.

Cloud computing did not exist then. Software could not balance load between servers automatically, so it had to be done by hand.

Huang says he still remembers how, every evening when the load was heavy, he and Dai Yunjie had to micro the servers by hand, distributing traffic across them so that everyone could get through. Once they had finished microing the servers, they would go and play a few rounds of Warcraft III and micro there instead.

That experience later left XD Network (心动网络) with a great deal of knowledge about handling server load when it moved into games.

In 2005 Cai Wensheng (蔡文胜) invested one million yuan in VeryCD, which finally gave Huang some working capital. He bought servers and gradually replaced the rented ones, and hired a software team to develop the eMule client.

The name electric donkey is worth explaining here.

They did not choose it, Huang says. Users called it that. When people later accused them of passing a mule off as a donkey, he felt rather hard done by.

eMule literally means electric mule. eMule was a foreign open-source download program, and the electric donkey Huang Yimeng built was based on it. eMule itself, though, had been built on a commercial program called eDonkey2000, and eDonkey literally means electric donkey. To signal that they were better than eDonkey2000, the eMule developers had picked the mule, an animal stronger than either a donkey or a horse.

So the software Huang Yimeng made was dragged all the way back round to the donkey by its users. It is a thoroughly muddled account, however you look at it.

From 2005 to 2008 VeryCD kept growing, but its costs were heavy too. The business still ran mainly on advertising, payment for which was often in arrears, and the company could do no more than break even.

What worried Huang Yimeng more was how fast the market around him was changing.

First, public awareness of copyright had risen and policy had tightened. In the early days nobody paid any attention; then came copyright disputes; then the state began to take film copyright seriously. The pool of material that could be shared was shrinking, and that was the direction of travel.

Second, Xunlei and the online video-on-demand sites were rising. They had taken larger investments and had better technology, and growth in eMule's user numbers slowed.

Most important of all, Huang Yimeng believes his own vision and judgement were not good enough while he was building VeryCD. He had no clear idea of where the site should go in future, he was too idealistic to think about return on investment, and he had no experience of running a company or structuring its equity.

The consequence was that by the time he wanted to change direction, the company could no longer be turned.

Between 2008 and 2009 Huang tried a great many directions. He thought about television set-top boxes and about a video site, but neither had a good business model of its own.

Then browser games took off. VeryCD began acting as a distribution channel and joint-operated Wulin Yingxiong (武林英雄), Sanguo Fengyun (三国风云) and Shangye Dahen (商业大亨), an arrangement in which a channel promotes another company's game and takes a cut of the revenue. To everyone's surprise the company found itself taking a million yuan a month.

Huang found joint operation both more profitable and more stable than advertising. He had also played games since childhood and loved them. He had played Command & Conquer and Quake II obsessively on the computer at home, and when esports took off he was highly skilled at Counter-Strike, Warcraft III and StarCraft and even competed in tournaments. He owned consoles and handhelds from every era: a Game Boy, a Nintendo DS, a PlayStation, an Xbox.

So the way he saw it, the games out there were crudely made and still made money. Why not make a game himself and operate it himself?

In 2009, to break out of the old company's equity structure and make the change, Huang gritted his teeth and sold 7999.com, a web directory site that was still doing well. It went for 8 million yuan, of which only 4 million reached him. He then spent 1 million on the domain XD.com and 500,000 fitting out a new office, and used the remaining 2.5 million to hire a large intake of new staff to develop games in-house.

In-house development carries a fairly high risk, and the projects both succeeded and failed. The first thing they launched, a small social game, failed, and the whole team behind it was eventually disbanded.

By 2010, when the first game they had developed themselves, Tiandi Yingxiong (天地英雄), launched successfully, money at the company was already tight. The office took up two floors on Tianyaoqiao Road in Shanghai and held more than a hundred people. It was, by any measure, a very large bet.

The bet came off. Tiandi Yingxiong was soon turning over 30 million yuan, and Huang Yimeng's belief in the games business hardened.

Shenxian Dao overturned the entire game channel business

In 2005 Cai Wensheng convened a webmasters' conference in Xiamen. It had a lasting effect on the shape of the Chinese internet that followed, and a great many people first met one another there.

Huang Yimeng met Cai Wensheng at that conference, and there too he met Afei, the boss of ChinaZ.com, who years later would run Feiyu Technology, the company that developed Shenxian Dao (神仙道).

In 2010, after XD had made a success of Tiandi Yingxiong, Afei sent Huang Yimeng a playable demo. Huang thought the quality was high and wanted to sign it as publisher, but several game companies were already competing for it and the licensing fee had climbed to 4 million yuan.

Four million does not sound like much today, but at the time it was extraordinary. Remember that Huang Yimeng had developed several games in-house on 2.5 million.

XD was not the highest bidder in the end, but Shenxian Dao went to XD anyway, because Afei rated what XD had achieved with Tiandi Yingxiong and thought well of the suggestions XD had made about Shenxian Dao.

XD Network at that point had never thought of becoming a publisher in its own right. It was still mainly a development company with very few operations staff. So Huang Yimeng handed the running of Shenxian Dao to Liu Yifeng (刘义峰), a retired professional Warcraft III player who had only just joined the company. Huang had competed himself, and reckoned that anyone good enough to go professional was clever.

Huang Yimeng says Shenxian Dao went extremely smoothly after launch. Within six months it was taking more than 100 million yuan a month. He puts its success down to three things.

The first is that the game itself had unusually good graphics for the time. It was side-scrolling, and you could see a lot of people on one screen, which browser games before it had rarely managed.

The second is the large stock of servers left over from VeryCD, and years of accumulated experience running them, which meant XD's operating costs were far lower than anyone else's. XD was also the first in the industry able to open a new server instantly, at a time when other operators needed at least three to seven days.

The third is that low costs allowed XD to give its joint-operation channel partners, some of them well-known individuals, more than 70 per cent of the revenue, a share no other game operator could match.

Between them, these internal and external factors made Shenxian Dao a success and overturned the channel strategy of the period.

It was also over 2011 and 2012 that Huang Yimeng's son and then his daughter were born. In 2012, for his thirtieth birthday, he bought himself a Lamborghini.

Huang says that on the day his son was born he was driving home along an elevated expressway, and the setting sun ahead of him came into the car and fell across his face, and he was happy.

A successful business, a happy home, a son and a daughter. It was the happiest period of Huang Yimeng's life.

The story does not stop there, though. What comes next is the point.

Before long a browser game copying Shenxian Dao appeared. Its payment systems were designed with brutal aggression, it took even more money than Shenxian Dao did, and legally there was nothing XD could do about it.

The joint-operation channels cared about none of that. They promoted whichever game made money faster and made more of it.

That forced Shenxian Dao to change its own systems too and find ways of raising revenue, and XD had to fall into line. It soon won the market back, but doing so destroyed the game's own future, and the balance of its later stages suffered badly.

Unfortunately, Huang says, none of this made him think any harder at the time. After the enormous success of Shenxian Dao his confidence swelled, and he assumed that all he had to do in future was apply the old formula to new games and success would be guaranteed.

XD Network duly picked up a great many games afterwards. Some of them performed decently, but not one did better than Shenxian Dao.

Huang Yimeng began to ask himself why games in China were becoming less and less fun.

Huang says that after Shenxian Dao he lost interest in new games. They all seemed much the same and he could not get into them, not even the ones on the download stores' top-grossing charts. At the time he did not think this was a problem. He put it down to having played too many games and grown jaded.

One platform did start him thinking along new lines, and that was Steam.

Have you noticed, Huang says, how many games on Steam are strange, offbeat things? The ideas run wild, they light you up when you play them, and beyond the one-off purchase there is almost nothing else to pay for. Why are games in China not like that?

In 2013 the golden age of browser games ended and the big game companies moved to mobile. XD was no exception. It plunged in, into another round of fierce competition.

Before long Huang Yimeng noticed the old problem again. The games on the major channels' top-grossing charts were all much of a muchness, and he could not get into them.

He thought back to fighting a copycat game for channel support.

Could the problem lie in the channel joint-operation model itself?

Following that thread, Huang found where the problem was. He calls it the original sin of joint operation.

Back then, the generous revenue share on Shenxian Dao had roused the joint-operation channels to the highest possible pitch of enthusiasm. It brought enormous benefits to game companies and channels alike, and the whole industry rushed to copy it.

After a few more years of that, the joint-operation channels had grown into a colossus. They had turned the games industry into a "traffic monetisation industry", and a great deal of foul practice was sheltered inside it.

The reason is easy to understand, Huang says. A channel's traffic is finite. The company has to make money and the editors have targets to hit, so they will not promote the wildly imaginative games on Steam, or indie games, or games you pay for once. They will pick games with more mature progression systems and higher revenue per user.

Competition between the channels is fierce as well. To pull in users they will resort to guild discounts, cut-price top-ups and other destructive tactics, which does serious damage to the balance of a game and to its fairness.

That disorder drove the game operators to extremes of their own. To get onto the revenue-ranked top-grossing charts and win a channel recommendation, they would put their own money into their own games to move up the rankings, dig out every possible payment point, and do whatever it took to convince the channels that their game was highly profitable.

So the games you saw on China's major charts became more and more alike, and more and more expensive to play.

Huang says the more money a game makes, the bigger the channel's cut and the more advertising the channel will put behind it. That is the strength of the joint-operation model and it is also its evil. People are greedy. As long as joint operation continues, the channels will only ever promote games that cost a lot to play rather than games that are fun, and the vicious circle will keep getting worse.

But could you get a channel to abandon joint operation, give up its cut and stop caring about revenue, to hand back what it already had? Was that ever going to be possible?

Huang Yimeng's answer was that if no such channel existed, he would build one.

And so TapTap.com came into being.

While designing the platform, Huang Yimeng thought back to VeryCD. That had been a sharing platform too, helping people find the material they were looking for, but it did nothing to help the people who made the content earn money. If anything it harmed them.

So the new platform had to help players find games they liked and also help game makers earn money.

Guarding against human greed meant decentralising and reducing human intervention.

Out went the revenue-ranked charts, out went the editors, and joint operation was ruled out. Recommendations would rest only on the people a user knew and on reviews from players in the community.

Finally, the platform had to be able to make money, because Huang Yimeng is no longer the too young, too naive boy of years before. He has to keep the model sustainable.

His plan is to sell game advertising once the platform is big enough.

There are a great many game companies now, Huang says, and they are keen to advertise. TapTap's users are strongly defined by games, and the market is in far better shape than it was ten years ago when he was building VeryCD. Label the advertising as advertising, and if a game really is fun people will download it.

XD, which overturned the games industry with joint operation, now means to end joint operation and overturn the games industry again.

Huang says the hundred days of building TapTap have brought a flood of praise. Players and game companies both like the model, and it has given him back the sense of achievement he had while building VeryCD. He is only thinking more clearly now and looking further ahead. TapTap is to become Steam on the phone, with far more weight put on building the community.

When the interview was over, I asked Huang Yimeng: "Dash, can I really write about your days as a fenqing, an angry young activist?"

Huang Yimeng said: "But I am not a fenqing any more."