Slashdot Mirror

It's now used not only in Computational Science, but also in business. Butterfly http://www.butterfly.net/ uses Grid as the online game solution. And as I know, SUN comes out with SUN Grid Engine, and Oracle also has its Grid solution for business customers.

As we worked on both intranet and internet a few decades before, we now have the grid for enterprise use and also the so-called Global Grid http://www.ggf.org/.

For one thing, centralized MMOG developers have to face a network scalability problem. Each connected player must be updated in real time, and serving thousands (or millions?) of players can be very network intensive. Your typical, optimized 3D shooter protocol uses something like 2K bytes per second. Serving five thousand players, in this case, would require an 80 MBPS link, which I imagine is not really cheap, at least not yet.

And of course the server is also a processing bottleneck. Running the simulation in the clients saves you from buying expensive machines or having to pay for the Butterfly Grid.

FreeMMG is a mixed peer-to-peer and client-server MMOG model. It leaves the server doing "lightweight" work such as authentication, session tracking, etc. and uses a P2P protocol to run the actual game simulation. And it offers protection agains player cheating (I hope so, if not, I'm not getting my master's :-).

IBM's last foray into MMO middle-ware: the Butterfly Grid.

Not to cast dispersions on the companies listed as developing games for the grid - but this is not a list of clients looking for middleware that's going to be worth IBM's focus.

Though there could possibly be some fairly interesting games that develop around such a fan-content real-money economy in a massmog - I don't see many games going in that direction, let alone enough to necessitate middleware.

They already did it...

Guess what is coming next?

But I am wait ing for this one.

Sony and IBM are throwing a lot of attention towards butterfly.net.

It's a small company in WV, but with some interesting applications for grid computing with massively parallel games.

For those who are too lazy, here is
Butterfly.net
"Welcome to Butterfly.net! Our fully-distributed server technology is pioneering the use of open grid computing protocols in large-scale immersive game networks that support unlimited numbers of players and require the most demanding levels of service."

If it's different Butterfly, sorry for that, thanks for Karma.

The first hurdle I see to this kind of setup is hardware. I think the guys at Butterfly.net are taking that battle to the streets by providing a game development platform on top of a massive Grid. This promotes seamless worlds that support hundreds of thousands of players.

The second hurdle is monetizing a world like this to provide impetus toward development. You could provide the world itslef as an open source playground, but the really nice artwork (textures, models, etc) and interactive elements (scripts, epic quests, etc) would be in packaged SDKs. You'd have to have some sort of monetizing here in order to support the bandwidth requirements.

The final hurdle is to manage adoption and pricing in such a manner that your pipe dream doesn't fold the minute the first player logs in.

On a small scale, I'm reminded of InterMUD and privatized mudlibs. You can download something like LimaMUD for free, stick it on top of MudOS and start developing. Support is nearly non-existant, and you have to develop 80% of the content yourself before you open the doors. However, you can license one of the old well-developed Mudlibs (like Nightmare, for example) and get yourself the majority of the way to the finish line.

The obvious differences between the above and your pipe dream are scale and graphical implementation.

Well, WV isn't exaclty a "backwater." That's not to say that we're directly competing with the likes of CA, but there *are* some interesting technologies coming out of the state. For example, the company I work for is doing some pretty interesting stuff with Linux, gaming, and Grid computing.

Is Grid Computing (http://www.butterfly.net)) really the foundation of enterprise-based Beowulf technology? If so, what other modernized aspects can this technology be applied to?

It would be interesting to see what this kind of technology could do for Myst as well. Perhaps they could combine randomly-generated Myst-like puzzles with MMORPG gameplay to liven up MMORPG a bit as well?

Unlimited numbers of players within one persistent-state world

Advance, Distributed Artificial Intelligence

Every game genre

Multiple, concurrent games

Any connected device

Hot-swappable components

Shared-source developer sandbox

Butterfly.net and IBM unvailed a gaming Grid build on Open Source software and IBM infrastructure. They claim the Grid can handle up to one million concurrent gamers. I don't see any mention of specifick games out there nor any mention of pricing. You can read the press releases from Butterfly.net, Infoworld and CNNalso vovered the story. And the toolkid can be obtained here if you want to get a headstart.

Butterfly.net and IBM unvailed a gaming Grid build on Open Source software and IBM infrastructure. They claim the Grid can handle up to one million concurrent gamers. I don't see any mention of specifick games out there nor any mention of pricing. You can read the press releases from Butterfly.net, Infoworld and CNNalso vovered the story. And the toolkid can be obtained here if you want to get a headstart.

Butterfly.net and IBM unvailed a gaming Grid build on Open Source software and IBM infrastructure. They claim the Grid can handle up to one million concurrent gamers. I don't see any mention of specifick games out there nor any mention of pricing. You can read the press releases from Butterfly.net, Infoworld and CNNalso vovered the story. And the toolkid can be obtained here if you want to get a headstart.