Back in 1994 Jay Cotton, an American college student from the University of Georgia, was struggling to find an easier way to play DOOM over the Internet. ID Software’s first phenomenal hit had the capacity for IPX multiplayer — a protocol that operated perfectly fine in LAN environments – but was not compatible with the TCP/IP protocol used for the Internet. After sifting through information on Newsgroups and subsequently discovering a piece of software known as TCPSetup, he and other like-minded players were eventually able to string together a few IP addresses (after a couple of hours of fiddling around) and have a few stable games. It was only after receiving an email from one of his regular opponents, Scott Coleman, that the first step to a solution to millions of frustrated gamers was established.
It was called iDoom.
iDoom was the first streamlined attempt at creating a game browser, improving on TCPsetup with the inclusion of a chat system, a database of players with stored information like IP Address and location, and algorithms to smooth out connection problems and, most importantly, display latency.
Fast forward to a few months later and version 1.0 was running, but it wouldn’t be until 1995 when Scott suggested to Jay that they create a similar piece of software for the popular game Descent, that the idea for a single, unified IPX matchmaking client reared its head.
Kali, named after the Hindu goddess associated with empowerment, did just that — allowed any MS-DOS based software with IPX to be easily and successfully played over the Internet. Shortly after Kali’s inception, however, Scott took a job at Interplay and Jay formed Kali Inc, producing a Windows 95-compatible client that continued support throughout the next five years.
The competition for online matchmaking heated up during the nineties with services ranging from Microsoft’s own Zone.com to competitors like MPlayer, TEN and WON fighting fiercely for control as games began early (and largely terrible) support for the TCP/IP protocol. As most titles still either lacked central servers or the resources to maintain them correctly, players were left to use 3rd party resources to connect them to other players.
Around the end of the decade, a group of 3 programmers created QSpy (QuakeSpy) in a similar manner and for a similar purpose as iDoom — except in the case of Quake, it was to find servers rather than players. This software eventually evolved into GameSpy3D and subsequently GameSpy Arcade, quickly becoming the dominant form of server browser until the mid 2000′s, when Steam exploded onto the scene, coupled with the burgeoning popularity of console-based matchmaking services like XBox LIVE.
But as time rolls on and developers and publishers are now (generally) able to host their own modern servers for matchmaking, and players are locked inside increasingly DRM-heavy systems of centralised control, what happens to those of us who like to play older titles? More importantly, what about the titles that are released with broken matchmaking, forcing players to fall back onto those “ancient” systems of IPX or IP based connection protocols, trawling through forums and IRC for other players?
It’s happened before. Brand-spanking new centralised servers for the 2009 launch of Gas Powered Games’s DemiGod were no match for the barrier much-maligned and now defunct Impulse DRM, forcing most players to find a solution on their own.
That came from a pretty unlikely place — Perth, Western Australia.
Scott Kevill is a strangely humble man for a guy who hosts roughly 60,000~ players on his service, GameRanger, at any given time. “I don’t really like to give out those figures”, Scott mentions bluntly, “But I will say there are about 9 million games played every month, and it’s growing”.
During the DemiGod debacle, Stardock was drawing an enormous amount of heat from players, furious at an inability to play their multiplayer-heavy game as designed. Stardock CEO, Brad Wardell, noticed on forums at the time that most players who were eventually successful were using GameRanger to facilitate the connection, as Scott had impressively anticipated the situation and quickly added support for the game to his software.
“He then changed the DemiGod site to start recommending people use GameRanger,” Scott noted. “At that point things were pretty chaotic and I was just in the right place at the right time”.
He’s not wrong. Membership exploded after the impromptu referral and is now pushing over 3.5 million accounts, with more being created every day.
GameRanger originally started as the first, and most comprehensive, matchmaking system for Macs. On its beta launch in July 1999, it had support for 11 games, including WarCraft 2 and Unreal, expanding to 61 by 2001, 130 by 2003 and 175 (193 in 2012) by 2008 — the same year Scott launched the much anticipated PC version. To date, GameRanger supports matchmaking for over 670 PC games, significantly more than Kali or GameSpy did at their peak.
It’s this level of overwhelming compatibility, and dedication to continuing support for new titles, that provides Scott with an enormous opportunity to take advantage of players who still struggle to navigate the laybrinth of firewalls, NATs and port forwarding. But that’s not the only situation where GameRanger finds itself attaining users: in most cases, GameRanger is the last man standing to support communities when games are orphaned by their developers.
EA is famous for dropping support for older titles once they have reached an internal and arbitrary use-by date, regardless of how many active gamers still play them. In April 2011, EA announced that it would shutdown support for a host of games, including that of Need For Speed: Most Wanted, as it had dropped below “1% of peak online players”.
“We added multiplayer support for Need for Speed: Most Wanted on April 7, in anticipation of those shutdowns,” Scott notes, “Back in August last year, there were about 5,000 multiplayer games of NFS:MW played each day on GameRanger.”
This happens more often than most people think. Large publishers regularly purge older titles from their systems to free up space for their new stables of games, even if a title is barely a year old. In more severe cases, it can be more than just one title, as players of titles that were famously “Powered By GameSpy” have begun to discover.
As part of IGN’s attempt to bring costs down ahead of a possible sale, GameSpy Technologies, the team and servers behind the matchmaking services for GameSpy Arcade and GameSpy Comrade, was sold in August to Glu Mobile, a publisher of games for smartphones and tablets. It’s largely unknown why Glu purchased the tech, but one can assume it would be to acquire either patents, code or both to develop internal matchmaking systems for mobile games. More importantly, the sale included the slew of servers currently hosting an enormous catalogue of PC titles that licensed GameSpy for multiplayer matchmaking.
Over the past few months there have been reports, both on Glu’s Facebook page and GameSpy’s own forums, from players reporting that many titles are no longer receiving any response from GameSpy servers, effectively breaking almost all multiplayer compatibility. Comments from developers have referred to the possibility that extra funds on top of the original agreed amounts were requested from publishers to continue support, and in most cases rejected, thus leaving many of these games completely broken.
I asked Scott for his thoughts regarding the GameSpy situation, and how it effects his company and the community. “Even though it does benefit GameRanger (in the long run), I still hate seeing it happen,” he said.
“I think it’s tragic that it rips the rug out from under players that have been regularly playing for years, leaving a void in their life. The lack of warning is the worst part, because no one can prepare for it to find alternative solutions before the community fragments and scatters. Why do they shutdown servers without notice? Because they hope no one will talk about it. They hope the remaining players will be silenced and not kick up a fuss.”
“Regardless of the size of the communities, that really shouldn’t matter anyway,” he continued. “If you and a few mates get nostalgic and decide you want to play an older game, you don’t always care about how active the community is since you were only playing with friends anyway. Yet you can’t play at all. There’s no reason it needs to be this way.”
Hopefully not for much longer it won’t. GameRanger, as is always the case when these situations occur, has been flooded with requests from both the developers and the community to help support these orphaned titles, in the same manner as previously when master servers mysteriously fall away, or — as is the case with Glu — licensing agreements are no longer cost effective or the company that originally provided the service simply keels over and dies.
Regularly adding compatibility for older games is akin to placing a bet on a wide field of horses. Demand can ebb and flow, so choosing which projects to invest time in can be volatile, particularly when a lot of the work is manual. “I’d love to see every game supported, large or small,” Scott muses, “I don’t think any should be left behind as they are all an important part of history. Even unpopular games have at least someone that loved them.”
“Yet the reality is that each game takes a significant amount of time to add and test to ensure that all games get the individual, first-class treatment they deserve. Given enough time, almost any barrier is surmountable and GameRanger has an ever-expanding toolbox of technical tricks to get these games working.”
I ask Scott if he can elaborate. “Any sufficiently advanced technology is indistinguishable from magic,” he replies, laughing. Fair enough.
So what process is there to choosing new titles? “There are many factors to this,” says Scott. “Is a game developer or publishing wanting GameRanger support in a new game? Which games are having the most problems playing online? How many users want this game right now? Did the servers for a game just get shut down? Does this game fill an under-represented genre? Has a new game just been released with show-stopping multiplayer issues? Which games’ communities are in danger of extinction?” The heavy weight of thousands of gamers, desperate to hold their small, tightknit communities together can, however difficult and stressful — Scott briefly mentions a time when he spent almost a whole day troubleshooting a user from Vietnam, remotely — provide a path to success.
The failures of other similar networks — and there are a lot of them — were usually due to an overwhelming reliance on short term thinking. Providing a service to users who rarely paid while at the same time hosting games for publishers who they didn’t charge licensing for was never going to be a successful, or even remotely viable, long term model.
“Greater efficiency, less overhead, superior tech, and understanding what is most important to online gamers,” Scott explains when I ask him how he’s outlived his competition. “These mean that GameRanger can be successful with revenue from advertising and non-essential premium upgrades and features. Charging game developers and publishers exorbitant yearly fees for the life of the game to cover mediocre, outdated tech is certainly NOT a solution. Those developers and publishers are eventually forced to make the financial decision of ending support for their games, and instead look for alternative service providers without strings attached for future titles.”
So, what’s the most popular game on the service? “Age of Empires II, but it varies greatly. Borderlands was equally huge for a while.” How fitting. As we enter a future where the online capability of our favourite titles face the distinct possibility of eventual abandonment, it’s nice to know there’s still a safety net available. After all, most of us get the urge to play StarCraft every now and again, don’t we?
If you’re keen to join the legions reliving the glory days of old, you can download GameRanger here. Many thanks to Scott for his time and insight.