Open Source Worlds

Some of the recent discussion on the EA Growth reminded me of a question I've been meaning to ask the hivemind: what impact is the open source movement likely to have on virtual worlds?

There are obviously some elements of this movement that one can find in VWs: the user-developed content in Second Life, the open source framework of Beyond 2. But as others have noted, it's hard to build a world, it takes a lot of artistic talent, programming talent, etc etc. And some have tried and failed.

Yet it works in lots of other contexts. The only thing that keeps Bill Gates awake at night is Linux.

Is there any reason that it can't/won't work in virtual worlds?

Comments

[Toby Hede]

1.

It seems that a great deal of Open Source development is on reasonably low-level and infrastrucutre projects. A distributed, decentralised OS model seems to work really well for operating systems, servers, and other such applications but starts to struggle when the user-interface is involved. The most succesful OS userland apps have been those driven by commercial interest, such as Mozilla, Eclipse and Open Office.

Games are a notable absence from the Open Source world, and this is not for lack of trying - there are any number of game projects on Sourceforge (http://sourceforge.net/) and other OS repositories. I think that lack of art is one of the largest barriers to OS game development. I can point you to literally dozens of game engines but would be hard pressed to find one comprehensive library of artistic resources.

I think that this will remain the case for the forseeable future. However, it is possible to envisage an Open Source architecture providing the "operating system" for virtual worlds, a common platform that would allow for both commercial and non-commercial VWs.

The current game technology world would however suggest that the nature of this infrastructure is specialised and difficult enough that only commerical products will realistically be viable (think about the current crop of "real" game engines (Unreal etc) - these are essentially commercial with controlled usage for non-commerical and hobbyists).

[Factory]

2.

Hmm are you excluding MUDs?
Most of them are open source in one way or the other, although I don't know if any MUD codebase has ever had a bazaar form of development. (Then again I've been out of the loop for a while)

[Ricky Justus]

3.

Toby> I think that lack of art is one of the largest barriers to OS game development.

Certainly, but what about specifically for Virtual Worlds, the fact that they need to be hosted by powerful servers to accomodate for the massively multiplayer element. Is that not even more costly than artwork?

[Mirjam Eladhari]

4.

Just a few urls relevant to the subject:

/one of the OS engines/
There is the NeL engine, a platform for persitent worlds, developed by Nevrax. The game Ryzome which currently is in beta testing is using the engine, and there is a open source dev community using the engine. A well functioning network layer.

/For research projects/
the Grid that Celia Pearce is working with, is a great initiative, connecting resourses from Butterfly and San Diego Supercomputer Center.

/A paper to check out/
is Julian Olivers Developers In Exile: Why Independent Game Development Needs an Island

[Jayce]

5.

One thing that must be mentioned is the fact that VWs are usually a service more than they are a product. It's a time-consuming hobby when discussing not-for-pay VWs (like many MUDs) and not for the faint of heart.

When discussing for-pay VWs, it seems to me that it may in some part be a NIH syndrome among the money people.

[Richard Bartle]

6.

Jayce>VWs are usually a service more than they are a product

I agree, but if open source graphic engines and free art assets become commonplace enough then the problem could become manageable. We might see graphical virtual worlds on the same scale as the smaller textual ones. If that happened, the service aspect need not be any more onerous than for a text MUD.

I'm looking forward to when this happens (assuming it does), because I think we'll get a lot more innovation than we're seeing at present.

Richard

[Euphrosyne]

7.

Art is definitely a primary barrier to creating graphical VWs. I'm familiar with a local project group of a couple score talented and dedicated people--coders, game designers, businessmen, modelers, etc. The one position they constantly struggled to fill? Artists. And this was a (modestly) paying project.

There seems to be some fundamental lack of the geek/GPL ethic among talented artists; skilled programmers will often spend hours a day, without compensation, on projects they would like to see come to fruition. In my experience, the first thing an artist asks even when he likes the subject matter, is what it pays. This is a reasonable enough position, but not conducive to creating lots of open-source worlds.

I don't see the game engine issue as being intractable--id Software in particular has been very generous with access and has released the full source of Quake 1 and 2 under GPL. There are also many good (smallish) texture and 3D model libraries out there. The problem of overall artistic direction remains, but perhaps we just haven't yet reached a point where enough people have access to the tools (which are usually expensive and specialized) to achieve the critical mass necessary for self-sustaining open-source VW projects. Nearly anyone can spend a little time learning, then sit down and write a MUD expansion in a text editor; spend enough time playing and the mechanics nearly write themselves. The access and learning curve for a full graphical/modelling suite is somewhat larger. Those with the preexisting skills don't form as tightly-knit a subculture as programmers tend to. [solution: teach these skills in required high school courses and in 10 years we'll be swimming in quality games]

If and when we do get there, "more innovation" will be a given.

[Sourtone]

8.

If you really look hard VW's are a dime a dozen. Open sourcing may introduce more 3d environments and perhaps a slightly better scale of small indie games but the overall effect will be minimal.

To offer the kind of quality and level of content most games provide requires a large team. As well, to maintain enough physical servers and connectivity to deliver that content to any reasonable sized player base requires an outlay of cash most opensource developers can't muster.

The limiting factor in who can roll out an MMOG isn't the cost of the engine.

[Toby Hede]

9.

I guess I am thinking that the next generation of VRs will be based on P2P technology, which would remove much of the need for centralised servers. Think that a distributed network and Open Source go hand in hand.

[Nathan Combs]

10.

Toby writes:

"I guess I am thinking that the next generation of VRs will be based on P2P technology,... Think that a distributed network and Open Source go hand in hand."

consistent with this: how to manage content (or optimistically, generate content) in distributed social systems (reputations etc). Such, i think, would be an important element in such an evolution.

-nathan

[Richard Bartle]

11.

Toby Hede>I am thinking that the next generation of VRs will be based on P2P technology, which would remove much of the need for centralised servers

"Much of the need" isn't "all of the need". In particular, P2P has issues concerning security.

Richard

[Jos Yule]

12.

I enjoyed reading Crosbie Fitch articles on Gamasutra about these issues:

http://www.gamasutra.com/features/20000120/fitch_01.htm

He tackles the security issue interestingly i thought - the whole series he wrote is quite interesting. He also started Cyberspace Engineers, tho i don't know where it has gone since i looked at it some time ago - http://www.cyberspaceengineers.org/

[Cory Ondrejka]

13.

Nathan> consistent with this: how to manage content (or optimistically, generate content) in distributed social systems (reputations etc). Such, i think, would be an important element in such an evolution.

Although the central SL machines are running inside one colo facility, we are absolutely generating content in a distributed fashion (in excess of 99% of the content in the world is user generated) by our users. In addition, the fact that SL runs on a distributed grid of machines was specifically chosen to allow, in the long run, for SL to run in a distributed fashion. As Richard points out, that raises some interesting security and technical issues, so in the near future we'll stay somewhat centralized. However, I firmly believe that in order for something like SL to spread to a web-like scale, that end users will have to contribute machines to the grid.

[Dan Hunter]

14.

One of the things that I meant to mention in the original posting was the way that mods work so effectively in the FPS and other single player (or non-persistent multi-player) environment. It's interesting that this is a huge deal in that part of the gaming world, but not in the MMOG/VW environment.

[Sourtone]

15.

Cory writes: "in excess of 99% of the content in the world is user generated"

I think that is a niche market that doesn't translate to most MMORPG's. I mean it works fine in a "life sim" game, but not in the adventure/hack & slash genre. The difference is in the make-up of your playerbase. Most hack&slash players play games to be entertained and to take part in another world. Not many want to pay you to have to create their own world. AC2 is a good example of that not working. The original concept behind it was to have zero NPC's.. everything was player driven. The result: Players got bored and left. In many cases where players have been counted on to create the content it's generally poor quality. Look at the vast majority of Neverwinter Nights games out there.

Dan said: "One of the things that I meant to mention in the original posting was the way that mods work so effectively in the FPS and other single player (or non-persistent multi-player) environment. It's interesting that this is a huge deal in that part of the gaming world, but not in the MMOG/VW environment."

Multi-player FPS games and MMOG's are very different environments. Mods work in FPS games because the content is provided in bite size (single map) chunks. Anyone can create one. Some are good.. some are bad. Anyone can create a map. Some are good.. some are bad. Each new map is a new game. It's the lack of persistence that makes the Mod possible.

Though it you want to look at it another way, there are examples of MMOG mods out there. They are just controlled by the companies controlling the content and not by the end-user mod builder. These "mods" are the special rules servers... the PvP, the RP. Anything where the gameplay is altered. I think EQ's "Premium Server" is the most extreme example of the MMOG Mod so far. However you will note that in each of these cases you, generally, can't take a persistent character back and forth between these special rules servers and the normal servers. Much like the FPS Mod, they are their own game.

Basically what I'm saying is don't count on your players. Players as a whole are an unpredictable mob.. the more you try to second guess them the more you will end up disappointing them. Players really don't know what they want - part of what's wrong with many of today's games is the designers are listening to the players to much and forgetting to consider if what they are hearing will actually work with the rest of the system they are creating. Poll players for what they like or don't like and use that as general guidelines... but don't build your game around it.

[Euphrosyne]

16.

Dan,
Maybe this is because the big FPS games provide the tools to make mods, and games like EQ don't (or can't?). Neverwinter Nights is the major exception I can think of.

Related to this is the content factor. In a MMOG or even NWN, creating a mod that goes beyond changing the appearance of preexisting items requires the generation of content though items and storylines--and this content tends to have a very limited lifespan of repeat playability. On the other hand, modding an FPS requires less content generation and has near limitless replayability, because in a PvP situation, the other players are generating the relevant content fresh with each game. Which is why Chess and Go have survived thousands of years, but a game of Trivial Pursuit has to be abandoned after a point.

If there's going to be a new paradigm that evolves for VWs, it might well involve this aspect of realtime player-generated content, although hopefully not exclusively in the explicitly antagonistic models of PvP that we have today. Unlike Sourtone, I'm saying do rely on your players, but don't expect them to create content the same way developers (or modders) do.

Short n' sweet: FPS mods work well because the content problem is already solved. Doing the same for MMOGs will take a conceptual leap.

[Avi Bar-Zeev]

17.

Keep in mind "distributed" and "P2P" are often very different things. Even in a distributed system, core attributes like money and abilities need a central authority and servers presume a high level of trust. That's true even if you isolate the problem to a central issuer of "trust" certificates or encryption keys.

Once you allow an unknown person access to the hardware that is authoritative for any piece of a VW (including player state and abilities), you require every transaction to be subject to inordinate verification, possibly decreasing as specific trust builds, but a huge tax on the system nonetheless.

For the sake of latency, the policy might be "act now, verify later," which means a system might have to handle a large amount of rollback and/or user-filtering when multiple clients or servers start to disagree on the facts. But if the system is decentralized, who is to say which rules for the world are valid?

In this scenario, a technical solution might see two players in the same virtual space but "disagreeing" about the world events that flow between them (e.g., the one with the "customized" client could have infinite bullets, infinite money, etc..). Ideally, you'd boot a detected "offender." But if the system is truly open, you might have to actually solve this problem--something along the lines of "each player or involved server does the same computation, and then they vote on what actually happened." But even that leads to some very interesting/strange scenarios: forced invisibility (leading to solipsism), forked worlds, strange paradoxes...

All in all, centralization is much easier.

[Richard Bartle]

18.

Coming back to Dan's original question about open source and virtual world development, by coincidence there was a posting on the GAMESNETWORK list today about a PhD scholarship in "Open source MMOG development" at the IT University of Copenhagen (where else).

It'll be interesting to find out how they get on.

Richard

[ren]

19.

I've been wondering if there is anything structual about the nature of game design and specfically MMORPGs that some how does not work with OpenSource generally or the particular cultural practice of it we see right now.

OK, OS's are pretty complex things, but I think they are also well understood, they have a kernal and varous layers that interact, at a high level its easy to abstract these and seperate out work, its not easy but it is possible then to look at how to optimise accross layers etc - and to me that's pretty much what your doing.

But with games, sure there is all that - the core logic, physics engines etc, but with games it feels like there are a other layers that are harder to hold together in even a big group let alone an open source one.

I guess you could say that games have things like an over arching metaphore or ethos which in very suble ways determines how small design decisoins are made to make a whole - its hard to see when this works, but easy to spot when it doesnt, just think when a game suddenly jars' (not breaks) but something just does not fit right either the way an object seems to work, a sub-plot line, a challenge that is just a bit too hard at that stage.

So sure, platforms (and of course app's like Gnome, The Gimp etc) but OpenSource games (of any significnat complexity) look like a major challenge to me.

[Mirjam Eladhari]

20.

Toby wrote:
>I think that lack of art is one of the largest barriers to OS game development. I can point you to literally dozens of game engines but would be hard pressed to find one comprehensive library of artistic resources.

Oh yes! That's a real problem: its so immensly time consuming, but there is maybe also a difference in professoion-culture.- (forgive me now for generalizing) Not that software developmnet is not time consuming, but developers acutally do it - I have sedom met graphical artist dreaming the big dream of open source. Seem to be a programmer thing.
Could it be because of lacking structures. Is there a GPL for meshes and textures in some format that ppl usually write exporters for? Would it even be applicable?

Sourtone wrote:
>part of what's wrong with many of today's games is the designers are listening to the players to much and forgetting to consider if what they are hearing will actually work with the rest of the system they are creating.

I guess that would compromise the integrity of the overall gamedesign to do it in exess... Could you give examples of mmogs where this happns? - i find it very intriguing.
Just remembered an older thread here when someone (i think it was Raph Koster) pointed out the importance of not only listening to players but to actually check the databases for what players do)

Ren wrote:
>I've been wondering if there is anything structual about the nature of game design and specfically MMORPGs that some how does not work with OpenSource generally or the particular cultural practice of it we see right now.

How i hope there isnt! I so much love the romantic grandness of the OS dream and practice. When it comes to mmogs is so complex systems that it becomes really difficult, as some posts here above have summarized. On the other hand we have ATID, that team is really small. (not OS, but team = small)

(and a little bit off-topic: sourtone wrote
> The original concept behind it was to have zero NPC's.. everything was player driven. The result: Players got bored and left.

when i played it the main reason that i heard player state as a reason for quitting was the lvl 50 cap. Players had levelled fast and got bored when they couldnt continue. The game desingn actively did not reward power levelling, the game was very much more fun if you took it in really esasy going pace, at least that is my experience, and also the advice i heard many players give to new players. I dont argue against Sourtones comment, just wanted to add to it.)

Mirjam

[Sourtone]

21.

Mirjam wrote:
"Could you give examples of mmogs where this happns?"

Well it takes me even further off topic but...

I can give you an example where it *seems* as if it was the driving force (not being a member of the dev team I won't claim to know the actual reasons for design). The best example I can give right now is Horizons which seems to have almsot every feature (short of PvP) I've heard players cry for on the various development boards I've been on.

Leveling is fast and easy... a few short months since relase and people are already hitting the soft caps and running out of things to do.

Plenty of ability to solo... grouping is ineffective at best and at times even counter-productive. While I'm all for the ability to solo, prefernetial treatment should be given to grouping. Grouping build community which I think most will agree is the lifeblood of any MMOG. (Poll most EQ players and they'll tell you they stay because of the relationships they've built - not the gameplay).

Crafting is easy and available to everyone with equally accessible resources and an entirely player based economy... which is a totally unviable economic model.

Weekly storyline updates... actually this is their best feature though their patching process is a tad painful.

Very easy death penalty... I'm still not sure if this is good or bad. The fear of death is really based solely on the long run back to where you were.

Playing as a Dragon... while in itself not a bad feature the way they structured it gives dragons their own version of the game (seperate equipment, skills, quests, classes, etc) and thus there are many times when dragons are totally left out of the game which remains mostly bi-ped oriented.

All idea's I've heard from players on various development boards from various games and all decent thoughts to incorporate to a degree... but taken all together like this they make for a clumsy game that I don't think will hold a very large audiance.

As for AC2, it had a myriad of problems... easy leveling and lack of content were just two ;-)

[Olivier Lejade]

22.

Well, as the guy who decided Nevrax would make it's engine (NeL) available as Free Software, I couldn't let this topic pass without commenting. :)

I see no reason why there couldn't be Free Software graphical virtual worlds. Not only that, but I am convinced that it will become the norm at one point or another.

The main reason it is not yet the case, as far as I am concerned, is the way the gaming industry is structured. With a few exception, the gaming industry is *extremely* proprietary. It also suffers from a grave case of Not Invented Here syndrome.

Now, i am of course only talking about the software part. The design part (conception and graphics) is another story altogether. I strongly believe we will see more and more user input, but it's going to take time as the tools and the culture necessary for that to happen has yet to be developped.

And just in case anybody is interested, my new company is working on a another framework for persistent worlds. Still very alpha though...

[B. Smith]

23.

RE: Lack of art, etc. in open source projects w.r.t. the scarcity of creatives....

1) Once something becomes open source/public domain, it stays there. Litle by little the amount of free artwork and models available is growing.

2) It doesn't take much technical ability to assemble the functional equivalent of a VR Leggo set. Once the textures, basic wire frames, filler items are made and put out there-followed by easy to use tools-the amount of art generated will skyrocket. Anyone here old enough to remember the first word processors? Took a Master's in electrical engineering and specialized post graduate work in comp sci to write a memo that actually had justified margins. Look at OpenOffice now. The tools are getting progressively better.

3) Creativity is scarce in any area. Look at the music industry. How much of what gets released every year is pure crap? How many movies bomb? Plays go belly up routinely. But people keep creating.

[astalok]

24.

[I'm new here (hi!), please be gentle. Long post]

Above, several posters have lamented about the shortage of artists working for free, and some have offered counterpoints. I offer two ways to get artists to work on a free MMORPG project:

1. The (almost) free approach: As pointed by Dan, mods in the FPS world have been very successful, some of them even going commercial. The wildly popular Counter-Strike for Half-Life and Tactical Ops for Unreal Tournament are good examples. I'm very familiar with the latter and can attest that there were lots of talented artists working on the project well before there were even rumors about commercialization. But one has to remember that most mod artists _do_ copyright their work and are very jealous of its use. There have been bitter fights over re-using textures or ambient sounds. On the other hand, most of the artists are quite tolerant of other people modifying their creations as long as they are credited.

Artist for FPS mods do mostly textures (walls, floors, trees, etc.) or models (cars, buildings) for which there are tools to make things easier - often provided by the game developer -, making entry into artistry easier. As B. Smith comments above, these easy-to-use tools are essential in attracting artists to a project. When you have that, competition, egos and desire for prestige will take care of the rest.

This works well with the mod community. I've seen not only beautiful textures, but also a lot of great original artwork (paintings, posters, graffiti). The problem with MMORPGs will be selling of items. When artists see other people profiting from their hard work, and in this case free work, they will not be happy. Expect rough seas.

2. The let-the-players-create -approach: Players themselves can and do create a lot as we've seen in games which allow it. So what Second Life is doing with giving the players ownership of their own works is another way to draw in artists. While not any more "free" as the previous approach, this does provide the gameworld with original artwork which doesn't have to be created by the developers themselves. Again, this can be encouraged by providing the players with the tools for creation.

This approach works very well with stuff you can sell (garments, houses, carryable items). Where it might fail is with the environment. Or would you want to see lawn texture with "copyright Ormang the Lanscaper" written in bright yellow letters at a corner of each tile, or mountains with logos of guilds on them? Another problem is the "plain world" -problem, ie. if the developers put out a game with not textures, artwork, etc, the players will not come even for the beta.

Maybe the ideal solution is to combine the above two approaches. This might be prohibitively difficult to do due to conflict of interests. Suppose all landscape artists, texture-makers and ambient noise creators work for nothing but recognition, but everyone creating trinkets and paintings own their creations and make seashells. Would the free artists flock towards seashell-making creations? Or would there be a fruitful rivalry between the two groups? I suspect human nature would make free artists extinct in short order.

Or what if the developers announce early on in the project that everything created during beta is free, and everything after that will be owned by the artists? Would we see people creating crappy art in anticipation of getting rights to them after going gold, thus ruining the project from the get-go?

As we can see, artwork for free games is difficult, and artwork for free MMORPGs is even more difficult. I see no way of attracting artists who won't want to be credited for their art. Attracting artists who get credit but no right to their work is slightly easier. But I'm sure it would be not difficult at all to get artists to create if they own rights to their work. Combining that with the unique features of MMORPGs, namely the need for and environment vs. sellable items, will be extremely difficult.

Ladies and gentlemen, what we have here is the classic chicken and egg problem. There are solutions to attracting artists for free MMORPG projects, but encouraging people's creativity at the right stage in the project and harnessing ongoing creation of art need a lot of brain cycles from project coordinators.