Sun's chief gaming officer, Chris Melissinos explains, "I argue that we've been the principle architect of the largest massively multiplayer online game in the world. It's Wall Street. If you took a look at all of the mechanics that go in to building an online trading system, they're almost one-for-one, the same functions needed to build an MMOG. Except we've done it with more redundancy, reliability and scalability than pretty much anyone else."
And it's with that experience that Sun has been quietly applying resources toward building what it considers to be the most versatile and scalable MMOG middleware ever conceived. With the imaginative working title of Sun Gaming Server, the first SDK's will be delivered for free to developers in the first half of 2006.
The Utility Model
But there's more to this vision than just simple middleware. Nothing less than rewriting the underlying economics of the MMOG space is at stake.
"We believe in order for the online multiplayer game space to really flourish, we have to start moving toward a utility model for online computing," says Melissinos. "The reason we're able to get the scale that way is because we're built around utilities, which is common feature sets for common use.
"Today, if you talk to a lot of MMOG companies, they're constantly reinventing the wheel, every time they build a new game. Chat is never the same, toolsets for asset management are never the same from game to game. For their own particular title or own product. It's an incredibly expensive proposition for a small developer to get into this space against larger companies that have the resources."
What's the Big Deal?
There have been a number of companies developing MMOG middleware for at least two or three years (Big World, Butterfly.net and at one point, Massive Inc.). So what makes Sun's solution so special?
"The difference you see between a lot of the technologies today (say, for example, Big World) and ours is that not only do we offer the scalability that they do, but I can take multiple games of different types, running across different hardware clients and run them simultaneously on the same stack of hardware."
"So what you can do is rather than building an infrastructure to handle a particular game, I can go to an operator that has built an infrastructure to handle 5 million concurrent players. I don't care if it is one game running 5 million players or it's a 100 games handling 50,000 players each or 1,000 games running 5,000 players each. It's the first solution that's able to do this. We've demonstrated this publicly at GDC this year and we're looking forward to showing much more at next year's GDC."
Content is King
The goal of this whole process is to have a number of operators being able to build sustainable, robust, easily maintained infrastructures (just there are for web hosting, Teamspeak/Ventrilo or other standard online service hosts) and to have enough content being generated of different types to reach entirely different audiences. Because there's no need to separate the hardware stacks per game, it's safer for an operator to invest in such hardware. For the developer, there's theoretically no upfront infrastructure investments required because of operator support and competition from operators keeps costs from spiraling out of control. All the developers have to do is create the content and operators potentially share in the revenue streams from the games.
"If you want to go ahead and deploy it yourself, build your own data center and network operation center, you're more than welcome to do that. We'd be more than happy to sell you the technology to facilitate that," says Melissinos. "But I think the more attractive model will be for a small developer to be able to build a massively multiplayer title and not have to make that investment and take that game that they've built and seamlessly drop it onto any server at a number of providers that are running this solution. Just choose the one that has the best service plan for you. Choose the one that has the best additional feature set for you in terms of services and deploy to that network.
"There's no difference between the developer's technology and the technology used to deploy in a massive user base other than scale. You're not going to house 59 racks of servers in your office. You'll test and develop on one server and then drop it into a network that is built to scale [at an operator]."
"Because you don't have to invest in any of that [hardware], we now providing a viable business model for a small developer and allowing him to take advantage of economies of scale that a utility model can provide. Because I can now build a single infrastructure to handle millions of people, I can now leverage a utility model, exactly like power companies or water companies can.
No Fear of Cannibalization
"People talk about cannibalization in the MMOG space in general, but I don't believe that," says Melissinos. "I believe there's cannibalization of a genre within that space. The market for men-in-tights MMOGs is narrowing. I don't believe cannibalization of something so broad as online multiplayer games is possible. It's like saying television programming cannibalizes each other which is why we only had four television networks. Well cable came on the scene and we now have 250+ channels of programming that blew a hole right into the side of that market space. You could say that most of those channels are crap, but there's an audience for every single one of those channels. I don't think the people running Home & Garden Network are in any way cannibalizing the Sci Fi channel's audience.
"We think this is going to be the kind of model that opens up broader content. There's probably a market for a 5,000 person competitive needlepoint game for grandmothers. No one would build it today because the economics wouldn't work. With our technology, you could. And you and I as EverQuest or WoW or even Halo 2 players, would never touch it, but that doesn't mean there isn't a channel of viewers waiting to get access to that content."