In June 2003, Doug Rush received a copy of Star Wars Galaxies, a new massively multiplayer online role-playing game, as a self-interested gift from his brother, who needed a medic to accompany him on his adventures.
Rush, then 43, had enjoyed the Star Wars films but didn’t consider himself a fan. And unlike his brother, he really didn’t like video games. They were too straightforward, he thought; they made you play by someone else’s rules. There was no challenge in trying to understand them, and you couldn’t open them up and dig around to see how they worked.
He’s always been a tinkerer. “I was probably 9 years old when I asked my mom for a multimeter from RadioShack,” he says via Skype. “I take things apart and I put it back together and I’m really, really good at it. I learn the details of how something can function, and I use it to the best of its ability. It’s always been that way for me.”
Despite his lack of interest, after two weeks of being hassled by his brother, Rush finally relented and tried the game. In a few hours, he was hooked. Two weeks later, he’d gone from disliking video games to playing Galaxies like it was his full-time job.
Now 55 years old, he lives outside Dayton, Ohio, and works as a service manager for a commercial refrigeration company. He can tell you the boiling point of any substance known to man—including neons and metals—and runs a team of 12 technicians that provides on-site training and consultation for companies in Ohio, Kentucky, and Indiana. “We all have 15-plus years experience with specializations,” he says. “We consult for companies without the skills our techs have. On-site training. My guys are the best of the best.”
He also has a second job, and again it’s related to Star Wars Galaxies—but this time on the other side of the game code. He’s the quality assurance lead on the Star Wars Galaxies emulator project, or SWGEmu, a fan effort to revive one of the most ambitious massively multiplayer online games ever made.
It’s both a labor of love and an abstract, technical challenge. Among veterans of Star Wars Galaxies you hear a mixture of wonder and tragic sadness when they reflect on how great the game was—and how short was its time. Launched in mid-2003, it was developed during the early days of the MMO genre, when developers were freer to experiment with the potential of persistent, online game worlds filled with thousands of players. Galaxies stressed freedom of choice and was set among the sights and sounds of the Star Wars films.
Galaxies players clinked glasses in the cantina at Mos Eisley, rode speeder bikes through the dense jungles of Naboo, hunted creatures in the grasslands of Corellia, and faced deadly Rancors on Dathomir. For two years they lived in the Star Wars universe, not as Luke Skywalker, Han Solo, and Darth Vader, or any of the other recognizable characters, but as—in the universe-spanning, space-opera scale of things—bit players. They weren’t heroes, but they were able to live their own stories. That was the wonder of Star Wars Galaxies.
“I got involved in this because I wanted to play, and it was broken and I didn’t like it. So I did what I could do to help make it right.”
Then change came to their world. The game’s producers decided that not being Han Solo or Luke Skywalker or Princess Leia had limited appeal. They tried to win over new players by making the game easier. But that meant dismantling the complexity that kept people like Rush coming back; fans felt the game was being slowly gutted. What had previously been an enthralling illusion of living in the Star Wars universe—really living in it, not just playing the role of a familiar, heroic character—became just another shallow franchise tie-in. Galaxies came to feel like a cookie-cutter MMO with Star Wars emblazoned on the game box. “I was outraged,” he says. “They took what I know about this game, and they altered it drastically.”
Seven years later, Rush discovered a group of volunteer developers trying to revive the dead world of Galaxies. “I got involved in this because I wanted to play, and it was broken and I didn’t like it,” says Rush. He saw a glimpse of the world he remembered, and he wanted to make it better. “So I did what I could do to help make it right. I’m still doing that.”
“People look at it as sort of an exemplar on the sandbox side of things,” says Raph Koster, the creative director of Star Wars Galaxies, during a phone interview. “Which is to say, a world where players are not being directed moment to moment or guided linearly through the game.”
Most massively multiplayer online games today are akin to theme parks—virtual worlds loaded with adventures and cities and merchants to buy and sell equipment—but the worlds are essentially static. Every few years a game expansion might be released that makes radical changes to the game world, but then that simply becomes the new normal: another template that cannot be changed by the players.
Star Wars Galaxies, on the other hand, was designed as a world to be constructed by players, where cooperation wasn’t merely desirable but necessary for the game to function.
Koster tapped into the most human of behaviors, the need to build things and the desire to consume them, as the kick-start for the social machine he’d created.
“The rule was, the best things in the game had to be made by players, and the other rule was everything had to break,” says Koster. “Both of those are deeply contrary to what you see in something like World of Warcraft.” Without an active player base Star Wars Galaxies would have collapsed not solely owing to the economic considerations of low subscriber numbers, but because without a large number of players to build all the gear everyone would require, the world would just stop. Galaxies was a game that would live or die almost solely on its level of virtuality, to be a world compelling enough that people would want to work as well as live there.
“All of those things were designed in order to have players make friends,” says Koster. “And more specifically, it wasn’t even just making friends, it was making—the technical term is ‘weak ties.’” A weak tie is a relationship based on a mutual relationship with someone else. In the case of Galaxies, a scout who located resources had a weak tie to the manufacturer who accepted the resources from the supplier whose mining efforts were guided by the scout.
“The rule was, the best things in the game had to be made by players, and the other rule was everything had to break.”
“That sense of being part of a social fabric is missing in even most MMOs,” Koster says. “It’s just not there. We don’t have the social infrastructure for it to be able to happen. How can you be a loyal customer of a given weaponsmith in World of Warcraft? None of the infrastructure exists.”
Coronet City, capital of the planet Corellia, was filled with hundreds of players who emerged daily from the city’s starport and went about the business of an average day in Star Wars Galaxies. Players who mostly focused on combat might have wounds that needed healing, so they’d head to hospital where doctors stood near beds and tended to long queues of patients. Or players might be suffering from loss of morale—one way health was measured in the game—and that meant heading to a cantina was dense with other players, to watch dancers or listen to musicians perform.
Later, Galaxies players would strike out from prebuilt cities like Coronet, find empty land, build their homes, and form their own cities in which to build the shuttleports, hospitals, and cantinas that players needed. Interior decoration became another full-time profession. The world was constantly becoming busier, larger, and more rich, all at the hands of the players.
“World of Warcraft hadn’t yet cemented the template that we take for granted today,” Koster says, “and the future of MMOs looked to me like it was going to develop much more in the direction of a richer environment and more ways to play.” Instead, massively multiplayer online games since Galaxies have largely provided static worlds within which players enjoy prebuilt encounters, like dungeons. The worlds have a sameness to them, and they rarely change.
“When I would sit down to play, when I knew it was my time to play, I wouldn’t be bothered, I had to be prepared,” Rush says. “I had to have something to drink.” Mountain Dew and coffee usually did the trick, keeping him sharp while he calculated skill trees and did the math on all the possibilities he could explore. He was so taken with Galaxies’ underlying mechanics that he wanted to sample all the game’s myriad professions.
He had plenty of time to do so: He was on planned sabbatical from the commercial refrigeration business and was working on a framing crew building houses. “It was a straight 30 hours a week. It’s all I did. I was bored,” he says. He found his intellectual stimulation and challenge at night, when he logged in to play Galaxies and delved into the game’s complex world. He could spend hours roaming harsh worlds like the barren wastes of Lok or traveling between the Jedi temples in the forests of Yavin 4.
“Exploring the unknown parts of the galaxy that felt endless,” Rush says, “the beauty of an alien planet, breathtaking compared to the harsh wind of reality.” He watched the twin suns of Tatooine setting over a graveyard filled with the gigantic bones of Krayt dragons, and stared out in awe at the Jiberah Plains on the forest moon of Endor.
A single character in Galaxies required a subscription fee of $15 a month. By April 2005, Rush was paying Sony Online Entertainment $120 a month for a bevy of different characters. “I found it such an in-depth and complicated game that I ended up just buying more accounts,” he says. “I had eight accounts at one time, and I would keep them all running all day long.”
Galaxies was not friendly to newcomers. In most massively multiplayer online games, you choose a discrete role. Warriors swing swords, mages throw fireballs, clerics heal the warriors and the mages. Identities, once chosen, are set in stone.
In Star Wars Galaxies, you could be a Scout/Entertainer/Medic/Squad Leader, and then change your mind a week later and become a Brawler/Marksman/Pistoleer/Politician. Without a veteran Galaxies player to guide them, fresh players could easily get lost and lose interest in the game.
In 2005, LucasArts, the video game arm of Lucasfilm and the company from which Sony Online Entertainment (SOE) licensed the Star Wars intellectual property, wasn’t happy with the subscriber numbers for Star Wars Galaxies. Speaking to the New York Times in December 2005, Nancy MacIntyre, then-senior director of Star Wars Galaxies at LucasArts, said Galaxies was bleeding players more quickly than most online games. By then, World of Warcraft had launched; designed to reach the largest possible audience, its runaway success would ultimately cement the “theme park” template Raph Koster had tried to avoid.
“We really needed to give people the experience of being Han Solo or Luke Skywalker rather than being Uncle Owen, the moisture farmer,” MacIntyre told the New York Times. “We wanted more instant gratification: kill, get treasure, repeat. We needed to give people more of an opportunity to be a part of what they have seen in the movies rather than something they had created themselves.”
“We wanted more instant gratification: kill, get treasure, repeat.”
LucasArts and SOE implemented two major sets of changes to Galaxies in 2005. The first set of changes, called the Combat Upgrade, landed in April. It substantially simplified the game’s combat system, which set off a cascade of other related changes. Weapons and armor were different; items that were previously valuable became worthless. Players focused on item crafting had to update their entire stock of goods.
The Combat Upgrade forced some veteran Galaxies players, like Doug Rush, to abandon the game. “Things had changed quite drastically on the surface. It played differently,” he said. “On the surface, they reduced it down to common denominators that just removed a lot of the complexity from the combat side. I didn’t like it. I left.”
The second set of changes, the New Game Experience (NGE), was implemented in November 2005 and was far more sweeping. The entire professions system, with all its complexity and freedom it granted players in customizing their play styles, was wiped out overnight.
“They reduced it down to common denominators that just removed a lot of the complexity from the combat side. I didn’t like it. I left.”
Post-NGE, Galaxies players chose from one of a limited number of cookie-cutter character types like Smuggler and Jedi, which mimicked the way players chose roles in games like World of Warcraft. Disgust from veteran Galaxies players flooded the official forums, and a mass exodus from the game began.
Mention of the NGE in online communities dedicated to MMOs still rankles former players of Galaxies. The game limped along for six years, a shadow of its former self, until it was shut down on Dec. 15, 2011.
Victor Popovici, the project lead of SWGEmu, was a student pursuing a degree in computer science and mathematics at the University of Barcelona in Spain when his friends introduced him to Star Wars Galaxies in early 2005. It grabbed him immediately.
“I slept maybe four or five hours a day,” Popovici says via Skype. Most of his prior experience in gaming revolved around Quake and other shooters. He fell in love with the player-versus-player aspects of Galaxies, waking up at 6am to meet friends for in-game activities, like attacking enemy bases, while most everyone was asleep in the real world and thus, fewer people were on guard in the game.
More than the thrill of early-morning raids, Popovici enjoyed the game’s complexity. “The depth that it had, no other game had it, and I think no other game has had it, still, since Galaxies,” Popovici says. And, like most veterans of the early days of Galaxies, Popovici eventually fell in love with the social scene, especially hanging out for hours in cantinas after a full evening of hunting other players. Galaxies was way more immersive for Popovici than hanging out with friends in the real world. College was about going out drinking. Galaxies was about socializing.
Popovici only had a few months to enjoy Galaxies before the Combat Upgrade utterly transformed the game, and his friends began leaving. “Players had to restart and relearn everything they could do,” he says. “A lot of my friends didn’t want to spend months to learn the new game.” He stuck with the game for two weeks after the Combat Upgrade, but with most of his friends gone, he soon had no reason to stay either. He walked away from that virtual world.
He soon discovered the Star Wars Galaxies Emulator project and volunteered his programming skills to help. The project members knew it was impossible to fix the world, post-Combat Upgrade. But it might be possible to rebuild the original Star Wars Galaxies—resurrect the world that had been erased.
It wouldn’t be an easy task. As players, they had game discs containing the files that described what things looked and sounded like in Galaxies. But they didn’t have the server code, the software that brought all the players together in a virtual world. That, they would have to re-create from scratch.
Here’s how it works: Every action in an MMO is transmitted to the server, which transmits the results back to the player’s computer—a single player in an hour of play might require hundreds of thousands of these transmissions to keep everything in sync. And then you multiply that by hundreds of players, all needing to be in sync with the server and with each other. If the transmissions fail or lag, the game doesn’t work.
If a character takes a step forward, raises a weapon, and fires a shot, each of those three actions generates its own transmission to and response from the server. A programmer can observe these transmissions and figure out precisely which pieces of data are attached to which specific action. What a programmer cannot see in those recordings is the exact process by which the game server processes the data.
To successfully emulate Star Wars Galaxies, the SWGEmu developers have to write server code that is able to process every single thing a Galaxies player could possibly do. Or, to put it another way, they have to write from scratch a program that can flawlessly handle the input from thousands of people playing one of the most complex MMOs ever designed.
The SWGEmu project began in December 2004, with a team of two developers. By the end of 2006, the team had grown to 15 members. Their first attempt to program a server had been an unmitigated failure. The unstable code couldn’t even support 20 players, when a healthy Galaxies community might require hundreds or thousands or of players to function.
Success, some of the developers realized, would depend on taking more time to carefully unpack the game files on the retail discs. In 2007, the SWGEmu developers began recruiting new volunteer developers, organizing a community management team and the establishment of forums for the project, and coding a game engine. Popovici continued researching how best to code a server. By December of that year, the team had realized the project was much larger than they had originally thought, and SWGEmu went open source.
In 2009, SWGEmu launched a phase testing and development program through which community members were asked to log in to the SWGEmu Test Center server and test specifics of the emulation software. 2009 was also the year that Popovici realized he had to begin writing his server code all over again, this time from scratch.
“It was the first time, for me at least, when the whole system clicked in my head,” Popovici says. “It was way easier to start over.” Popovici and the other developers didn’t see this as a loss of five years’ worth of work. It was a development met with tremendous excitement. A restart meant better tools to work with and easier ways to alter and add content.
Popovici is now 27 years old. He still lives in Barcelona and works on distributed systems and servers. He is currently employed by a gaming startup that is trying to build a platform for mobile devices.
After 11 years’ worth of work, Popovici still isn’t comfortable giving an estimate as to when the project will be finished. Once upon a time, he thought the emulator would be done by 2009, but that was actually when he realized he needed to start the server code all over again. There’s still an entire space expansion called Jump to Lightspeed that needs to be added.
Whenever the SWGEmu project is completed, that’s the end of the road for Popovici. “When I start the game and try to play, I can’t play it anymore,” he says. “I only see how the functions work, how there’s a bug or something. So I’m not planning on playing the game again.”
When Doug Rush discovered SWGEmu in 2012, he began as just another player exploring the revival. Then, his interest in the game’s engineering was rekindled. “The game was broken,” he says. “I wanted to give feedback to what I saw broken.”
He logged into SWGEmu’s Test Center server, where all new game code is tested for stability before being published to a larger server called Basilisk, on which usually between 1,200 and 1,500 people are signed in and playing Galaxies at any given moment. Crafting, something Rush had enjoyed so much in Galaxies before he left the game, was in a particularly fragile state.
Three months later, he noticed a forum post advertising an opening on the SWGEmu team for a formal quality assurance tester position. Rush applied and was accepted, based on his work on Test Center. And when the quality assurance leader left SWGEmu, Rush stepped up to take the position.
“They gave me the freedom to use tools that I have, transferable from the rest of my life, that I can apply here,” he says, “and once they gave me the freedom to do that, then I have just had a blast, having fun doing this for the last three years.”
“There’s something about being on this side of the code that kind of ruins playing the game. I know too much about how things work to actually play and be mystified by it again.”
As the QA lead, Rush is in charge of testing the incremental changes to the emulator constantly flowing from the dev team’s efforts. QA work, which often means testing minute aspects of game code, can be monotonous and taxing in the best of circumstances, and Rush is not equipped with the resources of a well-staffed game studio. He also spends a lot of time trying to recruit new developers for the project.
Yet, even after the tremendous amount of volunteer work he has invested in SWGEmu since 2012, he has no idea if he’ll actually play Star Wars Galaxies when the SWGEmu project is over.
“There’s something about being on this side of the code that kind of ruins playing the game,” he says. “I know too much about how things work to actually play and be mystified by it again.” But whether he has the chance to play Galaxies again is no longer the point.
There are currently eight servers being run by Star Wars Galaxies fans, using the code developed by SWGEmu. Some of them are adding entirely new content to the game, using an API that was also developed by SWGEmu.
“We also have client devs that are working on improving the client with mods, with new planets, with new models, cars, houses,” Popovici says. “So there’s a lot of talk and work being done about new content, original content.”
Rush keeps an eye on those eight communities, watching how they manage their servers. Their server admins have been given space on the SWGEmu forums, so that all the developers can exchange notes and learn from one another’s efforts.
“Those eight other servers are the future,” Rush says. “They are already making their changes they want to make, the way they play the game. They’re welcome to do it and have it run stable on our code. The life cycle of this has pretty much been made endless, as long as people are interested in playing.”
Illustration via Max Fleishman. Screengrab via Star Wars Galaxies.