Build Your Own MMOG

yebb writes "CNet reports about a company called Multiverse that has just begun beta testing of its platform for creating and integrating online virtual worlds. They are allowing developers and companies to use their online framework to expedite development of online games. Their network is free to use as long as you don't make any money from it's usage, but they also provide open source client applications to use or modify as you see fit." From the article: "'The business model is long-term,' said Richard Bartle, one of the pioneers of online games and an editor of Terra Nova, a leading Web site about virtual worlds. 'Although Multiverse's software will help speed up the to-market time for companies, it's still going to take developers ages to create content.' While Bartle is cautious about Multiverse's business model, he's fascinated by its potential."

Comments

Distributed MMORPGs

[VGPowerlord]

I wonder if anyone will ever get around to making a distributed MMORPG. Something where different parts of the game world would be hosted by different operators.

I suppose that would open a whole new slew of issues, though.

Re:Distributed MMORPGs

[cmstremi]

The overhead and lag would be unbearable using something like BT to keep the world in synch.

A 'distributed world' would make more sense (and probably what the parent post was saying) if 'zones' would be distributed. So running to the edge of one 'zone' might switch you to another one that might live on a different server at a different location. This of course might include a different type of game server if the client would be able to manage both. So you go from a street-based game and hop in a rocket and go to a space-based game.

Re:Distributed MMORPGs

[Siberwulf]

The whole issue with distributed MMORPGs is the lag factor, as the parent said. You need to figure out which content is where, and that requires a centralized node. That node of course, subject to all the lookups and the likes. Clustered boxes are the closest you can really come to this. Certain zones are on certain machines. If you want to switch zones, you go into a load screen, and then your server connects to the new zone, you get your characater serialized and moved to the new box, and then your client sees the new zone.

Re:Distributed MMORPGs

[bigpat]

I started working on the Varium MUD server with a couple other guys over 5 years ago, but frankly I wasn't a very good programmer at the time and I was supposed to implement the distributed stuff. It was written in Python, but we never got the distributed part working (my fault). But I had thought that we could put a graphical layer on top of the underlying game engine, to make a distributed MMORPG.

The goal was to have a server that could be distributed, so that say you direct your avatar to go through portals which would transfer your avatar and any compatible objects you were carrying to the linked server. Really the idea was to make the equivalent of hyperlinks, but that the servers would negotiate a transfer of the avatar and in game objects.

Some of the issues that came up (all solvable)

*security - ensuring that game objects could be transferred from one server to another without violating any rules of the local server.
      - we looked at parsing the incoming python for undesirable code in the same way that "Wizard" created code would be validated on the local server.
      - also having several levels of trust available between servers. (ie peered servers, trusted servers, unknown servers, banned servers and the ability to specify what individual privileges would transfer and correlate between servers)

*saving game state - assuming what if the remote server crashes, would we save any game state in the client (but what about cheats?) Probably just leave reliability up to the local admin and have a inactive copy of the avatar left on recently visited world servers so that you could just reconnect and pick up where you left off (more or less)

*also, game experience. What would you do about themed world servers? So, that if I was say on a Star Wars themed virtual world, and then went through a portal to a Barney themed world server, would my Chewy avatar suddenly turn into a Purple dinosour or would there be some sort of "Customs" border process where by you went through and specified preferences the first time you visited someplace. But that might really interfere with game play...

Anyway, those are a few of the things we talked about. The project is dead, codebase didn't include any distributable elements. Seems like distributed MMORPGs are wide open for an open licensed and non proprietary standard for connecting MMORPGs together to really create a workable metaverse.

Re:Distributed MMORPGs

[ciroknight]

If you've looked around any game creation forums, the idea has been mentioned at least a dozen times, but it ends up being too hackable, too bandwidth limited, and leads to fragmentation worse than a server split in IRC.

So no, I doubt anyone will ever get around to making a truly distributed MMORPG. Keep in mind that all MMORPGs are distributed by nature (as the client and the server both have a set of responsibilies), but as far as distributing the server part, it just gets too complicated.

But then again, there are different attacks to every problem, and distributing is just one way to deal with it. Others include clustering, and centralization, and both of the latter have proven to work better in the MMORPG case. Just because it works really well for sharing movies and high bandwidth items doesn't mean it'll work great for sharing game objects and such.

Re:Distributed MMORPGs

[Lonath]

Blizzard sort of did this with Diablo 1. They allowed people load characters from their home computers to play against each other online. I don't recall if there were any issues with exploits or cheats, however.

Re:Distributed MMORPGs

[MindStalker]

Just felt the desire to comment on the theoretics. Regarding Themed World, you simply wouldn't connect them, obviously one would only want to connect world that have similar rules. Sure I guess you could setup portals to other worlds more as a universal login/registration system. But gamestates would definatly need to be completly seperate per conceptual world. Think of it this way, say I travel to another country, I can probably bring with me something of value to trade with, but say I slip into a parrellel dimention not only would my gold be worthless but I could cheat and take thousands of pink tulips (their trade standard) with me.

Same with Security if you decided to allow untrusted servers, to connect you could theoretially send all their data with them when they leave, it would then be up to that server to decide if that data was worth something. Then when you returned said data can be accepted or dumped by you in which you would return them to their previous state.

You could even then setup an escrow service in which items of value can be traded into a central value store then reused in another service, any value at the new world would soly depend upon its creators and what it felt this central value system was worth.

Re:Distributed MMORPGs

[Phae]

MMOGs already have huge issues with synchronization and lag even with servers directly connected to each other, it would just add a huge level of overhead for keeping everything together. Also, as soon as they don't have direct control over servers and all game information, the potential for cheating explodes. As it is, the servers only have to validate input from clients; if it became distributed, then calculations would have to be validated as well. Currently most games have worlds in multiple states/countries, so you can find a server with optimal ping. It makes the most sense to keep an entire world in the same area, so everyone on it has a consistent connection.

Re:Distributed MMORPGs

[C0rinthian]

Yup, this wouldn't be exploited at all. First rule of online play: Never trust the client software with ANYTHING important.

Re:Distributed MMORPGs

[bigpat]

Just felt the desire to comment on the theoretics. Regarding Themed World, you simply wouldn't connect them, obviously one would only want to connect world that have similar rules.

Actually I was thinking it would be more fun and interesting the more exotic the differences and if sometimes you could have arrangements between differently themed worlds to allow migration of items and characters with all their traits across worlds. Likely such linking could only be done across trusted servers, so that you would likely start seeing networks of trusted servers evolve. But ideally you could retain your original avatars identity similar to email you could just be known by your "full name" when you leave your home world. Really, there is a lot that can be drawn from the way that people travel from one country to another with different systems of laws and currency.

But the key is to make it easy enough to set up that people with very little prior knowledge can set up servers. Say on the level of setting up apache with PHP or some other type of application server.

Re:Distributed MMORPGs

[_ph1ux_]

The idea of the actual game world being distributed is a tough one sure, but the idea that the characters can be tranferable is not. I had come up with an idea whereby one could build an avatar and use that model in any game. I called the idea BinaryBeing - basically you would have a site where you have an account which stores your various avatars. You could have one that was a general avatar used in any games that allow your fully customized char. but you could also have theme-compliant avatars where a certain world/fantasy archetype is employed.

The ultimate goal is to allow users to build a character which they can carry with them in the long term. and be able to unlock certain character atributes based on the games they use that character in. For example, you might receive the ability to display certain physical atributes in your character or tatoos on your skins based on what games you have accessed with the avatar. Aditionally - if you and your peers guild together in one game, you could develop guild markings/styles/tatoos/apearances which could remain consistant in all games.

The development would be done in cooporation with the various game devs in an effort to ensure that the characters could comply with whatever base rules the developers designate.

All avatars would be hosted on a site and an in-game connection to your binaryBeing account would be made which would grab the avatar model to the game you are playing - or would show you the useable models to pick from...

One could provide a much more in depth character development UI - and the game devs wouldnt have to focus on that as much.

This is nothing new

[Ninj0r]

This is nothing new. http://www.byond.com/

Umm...

[ndansmith]

I guess by Massively Multiplayer they mean "like 5 or 6 of your buddies."

I just cannot see anyone getting a large community of players together without a large development and advertising budget. Large development budgets also lead to independent, customized systems with total control, which is the opposite of the Multiverse concept.

Re:Umm...

[swilde23]

I just cannot see anyone getting a large community of players together without a large development and advertising budget. Large development budgets also lead to independent, customized systems with total control, which is the opposite of the Multiverse concept.

Didn't people say the same thing about Linux (and its variations)? Look at all the different flavors, and then consider the select few that have a reasonable following. I imagine that's what will happen here, assuming the base code is worthwhile. We'll see a handful of good games emerge from a cool idea, but we'll also see several thousand others that die.

Re:Umm...

[podperson]

Plenty of NeverWinter Nights servers attract larger consistent followings with no advertising.

Re:Umm...

[_Sharp'r_]

Apparently I'm starting to get too old for the /. crowd.

I made it through the entire page of comments (ok, only 2+, didn't read the others) and not a single person mentioned where this has all been done before in the MU* (MUSH/MUX/MUD/etc..) realm. You know, hundreds of players at a time in a multi-user game. Game services that did nothing but rent game accounts for you to build and run your own. All of the software highly customizable, with the better ones (Like PennMush, TinyMU*, etc..) having an in-game programming language that still has list and string functions easier to use than most high-level languages today have.

You know, the things that today's MMOG are built on top of? Heck, if you looked at Everquest, you could tell it was just a MUD with a GUI thrown on top.

This "Multiverse" is simply bringing the current crop of graphical MUDs back to the previous realm of how things were before some big budget guys starting spending the cash necessary to stick a GUI on top of some MUD software.

Oh great

[jackcarter]

Just what I need: joining some guy's MMOG and start standing inside the goatse guy's ass.

Let Users create content

[glengineer]

How about an MMORPG wherein the users can create content? So, after I play a while and become famous, rich whatever, I can build my own dungeon, castle, create my own army, my own artifact, create my own quests that other players undertake... There are a LOT of users out there with good content ideas who cannot get past the large undertaking to create their own game system, but who would love to create their own content in an existing one.

Re:Let Users create content

[Dukael_Mikakis]

True, the EXP model is perhaps the most outdated aspect in all of MMOGs, and user created content might just be the thing the push it forward.

If a certain MMOG consisted of a patchwork of user modules conventional EXP and loot wouldn't work as it is. To build off of my last post, maybe there could be some sort of economic or regressive reward system. Somebody creates their instance or whatever and a formula approximates the reward that each mob gives, or the reward for completing an instance or a quest in the instance, and this estimation would modulate as people run the instance. If people are having no trouble running the instance then the rewards and requirements become weaker, and vice versa.

This would also modulate things like Molten Core in WoW where people have it mapped out to be just loot farms. When an instance becomes formulaic like this, under my model, the reward formula makes the reward less and less appealing (as it is, Blizzard has to continue introducing harder mobs and better loot to deal with this).

And of course, it would help (in a way) to make all the modules "open" so obvious bugs or problems or impossible encounters could be fixed.

Re:Let Users create content

[John Hurliman]

Second Life

Re:Let Users create content

[neovoxx]

There is an MMOG somewhat like that already: <URL:http://www.there.com/>.
If I recall, it's something like $12/year.

Re:Let Users create content

[zoomba]

Sounds sorta like the dream behind Neverwinter Nights multiplayer where you could build your world and link it to other servers via portals. In theory it would create a zoned MMORPG if you had enough people working together to build and host content. In reality though, it never came to be since it required more work than the average user wanted to put in.

User generated content is only good if it's made so insanely easy to do that it's a no-brainer. As soon as you make it so users "design" the content, you're toast.

The general solution is to have pre-fab structures and templates for users to plop down and then to some limited extent fill in *some* custom content (usually smaller prebuilt objects the devs provide).

Player generated content is the holy grail of a lot of MMORPGs, but the trade off in a system like that is so much work has to go into making a workable world for players to create within, and a system robust and simple enough to make creation accessable to players, that there is almost always very little GAME built. Look at Star Wars Galaxies for an example of that. It's a beautiful sandbox system (prior to NGE) that allowed the users to define a vast majority of the game world through the economy, player towns etc... But there wasn't any gameplay or real content in there.

The portion of MMO players who want a toolset to create a universe is vanishingly small. Most people want a game to play, so devs like Blizzard don't bother with the player creation and focus solely on game content.

Re:Let Users create content

[Solder Fumes]

Obviously.

Scrolling through the comments, I see:
1. People who read about a similar universe in a book.
2. People who think this is a new great idea.

Seriously, guys, anyone interested in a collaborative freeform 3D world should have already registered an account at Second Life. It's been out there for a couple years already. The client *is* a 3D modeling tool, everything from clothes to massive Klingon spacecraft are built inworld. You can build anything you want in one of the sandbox areas if you don't own land. You can attach scripts to do almost anything to almost any object; everything from animating a sculpture to running a store or party game.

Second Life is now free. That is, you can register, have an avatar, get a weekly stipend, build anything you want, but you can't own land. No excuse not to try it out. One hint: turn off local lighting to speed up the framerate, Second Life is CPU-bound.

Tech

[kevin_conaway]

What are the technical details? Platforms? What apis are available? In what language?

Re:Tech

[Multiverse Corey]

Our server software is written in Java, so that gives you some flexibility with your server OS.

And as I mentioned in a different thread, the client that's in beta right is WinXP, but other platforms are on the roadmap. We are a bunch of early-days Netscape folks, so we've been pretty committed to the cross-platform thing for a decade or so.

We'll be posting more info on our tech specs at our site in coming weeks and months. You won't have to be an early developer on our beta platform to find out the important details.

--Corey

Corey Bridges
Exec Producer, Multiverse
http://www.multiverse.net/

Too Late! Been done...

[glengineer]

Take a look at this pic: (It's a PG-13 rated pic of a Quake2 arena, goatse style.) http://en.wikipedia.org/wiki/Image:Quake2gloomtemp leofgoatse.JPG

Nice press release.

[Minwee]

This sounds like the same sort of business model that companies like Kaneva are offering. "Play with our tools, build games, host them on our servers and we'll take a cut of the money."

It could be good if it works out. If not, at least it will let more people see that game design isn't about eating donuts and throwing darts at the "Nerf This" board.

Acronym translation

[Otto]

Sadly, I understood that and I've never played that game.

SWG = Star Wars Galaxies
CU = Combat Upgrade, which changed SWG in large ways, making it incompatible with previous versions. This led to pre-CU servers being made for people who preferred the old way
NGE = New Game Enchancements, which did much the same thing in terms of splitting the userbase.

I read too damn much gaming news.

Re:Acronym translation

[cnelzie]

The CU (Combat Upgrade) lead to calls for Pre-CU servers, but that never happened and never will happen as SOE (Sony Online Entertainment) isn't interested in making pre-CU Servers available.

    The rest is accurate.

I suppose it has merit...

[design by michael]

Somehow I think you'd be subject to a lot of loading screens -- like switching from a fantasy theme to say a WWII shoot-em-up setting. I can see how some things might be used centrally, like the engine itself; but other visual and audio resources?

Somehow I find it hard to imagine how they could provide enough universal resources to accommodate many possible different themes and gaming environments -- unless you don't mind using a "thud" sound for your AK-47 as you storm the beaches of Normandy on your shimmering unicorn.

TomeNet engine is open

[Lord Satri]

If you take a look a TomeNet, you get a MMORPG open engine. The sad thing is, you must go to the forums to learn anything about it because the website isn't updated often.

TomeNet is a roguelike multiplayer online rolepaying game based on Tolkien's work. The single player game is also getting closer to version 3.0, with a revamped game engine (open, with source code available). Great game and great community.

Re:Umm... Right

[Esteanil]

Yeah, like for instance Planetarion, a small browser-based game with absolutely no advertising, which grew from 0 to 200 000 users in a couple of months.
Seriously, word of mouth is and remains the main way to recruit new customers.
Just make a good enough product and your customers are more than happy to do the advertising for you.

WoW

[cmburns69]

How long before somebody creates a knockoff of World of Warcraft (ala the bnetd fiasco), and this company gets sued by [insert Blizzard parent company here]?

But seriously, what's to stop people from implementing their favorite games here, and what kind of liability is assumed by this company for providing the platform? With the current legal climate, services (such as Kazaa, morpheus, even bit-torrent) have been held accountable for copyright violations despite not having any control over the contents.

How long until the same issue effects this system?

Blender

[LetterRip]

I would think that Blender would be an ideal match for this, since they need a way for potential users to create content - models and animations, cheaply and easily.

LetterRip

Hasn't this been done?

[gtshafted]

I could be wrong, but doesn't Neverwinter Nights and Second Life already let you do this?

The whole issue?

[bill_kress]

Maybe a part--but the biggest is sharing an avatar across multiple rulesets.

All you need is one renegade sysop handing out +127 boltlightnings or modifying stat points and the entire site is gone, all users have to reset. Do you trust others with that much control over your system?

I suppose if you really had to do something like that you could introduce realms where attributes didn't carry over (say, air, fire, water, spirit or some such garbage). The problem is that to make it work your stats and items can't carry over, but I'm sure that's exactly what people think they want.

Re:Been Done Before... Sorta

[halber_mensch]

But yeah, when it comes down it, not many blockbuster titles really used somebody else's source. Half-life is the only one I can think of that redefined gaming.

Half life was built on a heavily modified Quake I engine (which still looks gorgeous to me). Medal of Honor: Allied Assault and Jedi Knight II were built with the Quake III engine. Those were fairly popular games, as I recall. A good list of games derived from the engine can be found here.

But that's always how iD has been. John Carmack is a dyed-in-the-wool algorithm wizard. He writes very elegant and optimized code to solve problems, and researches data structures and algorithms heavily to build his tools the 'right' way. If you read Michael Abrash's Grpahics Programming Black book, he notes Carmack's obsession with optimizing his spatial organization routines and data structures - how he stayed up late nights trying out different data structures and algorithms to get the most optimal rendering time out of his engine. Unfortunately, while the engines are superb implementations of advanced concepts, for the most part iD doesn't push out the same kind of content that people like Valve can 'pump' out. They're just more programming-centric. But when a content-centric entity publishes a title built on one of iD's engines, it's usually pretty rad.

Re:Been Done Before... Sorta

[TheZorch]

Half-Life certainly changed things. The game showed that FPS didn't have to be run, shoot and shoot some more type games. It was a thinking man's game.

Another game, though greated underrated, is Deus Ex. It used the Unreal Engoine and introduced gamers to the world of Cyberpunk and to a form of gameplay that didn't revolve around hosing down room after room with bullets. It was also an RPG of sorts, though it wasn't the first RPG built to play like a FPS. That honor goes to The Elder Scrolls: Daggerfall which used the Quake engine.

More innovative use of popular modern day engines can be seen in games like Splinter Cell which used the Unreal Tournament 2003 engine, and the MMORPG Lineage uses a highly modified Unreal Tournament 2003 engine.

Quake 3: Arena wasn't much of a game as compared to Unreal Tournament. UT offered many more gameplay features and better bot AI, but the Q3A engine itself was a decent graphics engine. It was used to make games such as the Medal of Honor series, the Call of Duty series, the critically acclaimed Star Trek Voyager Elite Force and its sequel, and Return to Castle Wolfenstein as well as Enemy Territory.

Now Warren Specter, the mastermind behind Deus Ex, is going to use the Half-Life 2 Source Engine in a new game project. Not to mention all of the real-world training sims that are in development using the Source Engine. One of them trains surgeons. Its the Source Engine's astonishingly good physics modeling engine that makes this possible.

Oh, and lets not forget America's Army, the game that doubles as a recruiting tool for the U.S. Armed Forces. It was built using the Unreal Engine and has been upgraded to the UT2k4 engine recently.

What goes around

[Zatar]

You mean like so many text-based MUDs have done since forever?

I remember playing a MUD in the early 90s where when you hit the max level you could plant a dungeon somewhere (just connected your dungeon to an existing room node) and write object oriented code to implement monsters, puzzles and treasure. I always figured it would be a good way to teach OO design since it's a lot easier to think of a monster or a magical scepter as an object with attributes than it is to start with abstract data structures...

Anyway, like all games, MUDs have had a large drop in the ability for users to affect them as the graphics and detail went up. Back in the day everyone could easily make levels for Lode Runner or Wolf 3d without any training or learning curve. Nowdays you need a degree in 4th dimensional geometry and several days of free time to make a good FPS level, never mind a virtual world...

Never heard of mods?

[Fluffy the attack ki]

Because everybody knows that Counter-Strike needed a massive development and marketing budget to become popular...

</sarcasm>

Really this sounds like a Good Thing(TM), and I'm really curious to find out what this is capable of in the long run.

NeverWinter Nights?

[podperson]

The only real difference between this idea and NeverWinter Nights is that:

This product supports ANY setting (although it probably requires a ton of work to make it support anything other than the vanilla fantasy setting they first thought of)

This product supports ANY ruleset (although it probably requires a ton of work to make it not support something other than the default fantasy ruleset)

This product is MASSIVE whereas NWN isn't. Although NWN or a descendant probably will be before they ship anything.

This product provides developers with an SDK. NWN provides developers with a fully functional IDE allowing a person only one skill (e.g. writing / programming / art) to contribute to or create a world.

This product provides a revenue model for content developers. NWN kind of does (they can commercialise a module you develop) but so far this hasn't worked out well for anyone except the developers of NWN.

This product doesn't exist. NWN does.

Mac Support

[Zobeid]

The Multiverse website says nothing about Mac support -- in fact, I couldn't find where it says anything at all about hardware requirements or operating systems. However. . . There is something on the Kothuria website. Their FAQ says:

Q: Will your game appear on other platforms?
A: Currently we aren't developing for other platforms than Windows PCs.

Based on history, I would assume if it's not being developed for both Mac and Windows simultaneously, then it's highly unlikely that it will ever support Macintosh (or Linux for that matter). As a general rule, it's not economically viable to take an already-developed MMOG and make a Mac port of it. (That was the official answer to our question about a Mac version of Star Wars Galaxies, in fact.)

Almost all the MMOGs that have appeared on the Mac platform came out of a combined PC-and-Mac development process: WW2 Online, Shadowbane, Second Life, and World of Warcraft.

Of MMOGs that began on PC and were later ported to Mac, the only one that comes to mind is Everquest. It was so late arriving that EQ2 and WoW were already knocking at the door, and it lacked compatibility to allow Mac and PC users play on the same servers.

Another example is Neverwinter Nights, which is not quite really a MMOG but is similar in some respects. It was ported to Macintosh, but the whole Aurora toolkit was left out -- excuse being that it depended on some developer library provided by a vendor who promised Mac support but never came through with it.

Multiverse are setting themselves up to be the same kind of obstacle. Basically, they're creating an ecosystem that locks out Macintosh not at the consumer level, but at the developer level.

SuperNodes?

[bhsx]

The concept of trusted supernodes comes to mind. Let's say you require one supernode per ten servers. The supernodes update each other as frequently as possible, use secure logins, and store your encumberance, your portable items and your traits. Depending on the style of game, how much you can carry with you could develop with character. These servers would also keep track of themed worlds and out-of-theme worlds and your personal config for each. For those who use the "standard" rule systems and themes, they would all share the default info. Those worlds that defined their own rules sets would set those up on the supernodes and could decide weather or not and with whom to connect.
This also takes care of the intermittant server crash/take-down problems. Just start-off on another server, keeping perpetuance; this could even be automated to be less annoying to the player.
Hmm, I've always thought I'd be good at designing this sort of thing. Too bad noone would listen to me as I can't code to save my life. sigh

Ever hear of Second Life?

[JamesGecko]

It sounds like you might want to check out Secondlife (http://secondlife.com/).
While not exactly an RPG, it is versitile enough to let you create your own little fantasy subworld and invite other people in for some leveling up on orcs and trolls. While I don't play it, I'm thinking of signing up sometime soon, it looks pretty fun.