Liberation Source Code

Wired (Gamers Eye Open Virtual Worlds) features Peter Ludlow beating the drum for Open Source virtual worlds.  The claim that "(his) expulsion from The Sims Online may well be the highest-profile player ban in the history of video games" only serves as fuel for this metaverse fire: the proletariat want satisfaction.   The real yarn, however, lies with whether or not this can be part of a calculus of virtual world liberation.  After all, game development is hard and perhaps Open Source would be a poor advantage.  While we've discussed Open Source and virtual worlds in the past, another question (beyond technology and its creation) is this one:

...the dream is that one large open-source virtual world could be created by stitching together work created in environments like MUPPETS, the Croquet Project and the OSMP.

What are the issues.  Avatar transfer protocols?  Insurance for virtual property?  Yellow Pages for virtual identiy?  Take your pick.

See also:

Open Gaming Uber-Alles.

State of Play 2 (videos) - Ludlow presents.

Comments

[Matt Mihaly]

1.

If you want true freedom without compromise, your only choice is to run your own, probably non-commercial, virtual world, as thousands of people have chosen to do by downloading open sourced text mud codebases.

[mike]

2.

I'm with Matt. It seems that in other contexts the drive for freedom causes splintering into smaller communities; my hunch is a good open-source virtual world codebase wouldn't result in one enormous world but a zillion tiny ones.

Standards would go a long way, though, allowing people to quickly create worlds from desirable pieces they've seen elsewhere or to recreate their avatars to the extent that the world allows such customization.

[Peter Ludlow]

3.

I agree/fear that splintering may be the result. All the big game development projects in SL seem to disintigrate on the interpersonal level. Maybe that would happen in an open source world as well. ugh, now I'm depressed.

[Peter S. Jenkins]

4.

Peter>I agree/fear that splintering may be the result.

This means that if there is to be a non-fragmented metaverse, it will likely be owned by corporate interests, since it would be extremely expensive to implement and operate. This of course leads to concerns about freedom of speech. As I noted in my paper, we need to recognize that not granting freedom of speech in MMOG's which are already more than games due to commodification, will set a dangerous precedent for the coming corporately owned metaverse.

Peter

[Matt Mihaly]

5.

Peter Ludlow wrote:
I agree/fear that splintering may be the result. All the big game development projects in SL seem to disintigrate on the interpersonal level. Maybe that would happen in an open source world as well. ugh, now I'm depressed.

I know some of you folks who didn't get into MUDs until the advent of mainstream graphical ones probably get tired of hearing this, but you just have to look at text MUDs to see that this is exactly what is likely (though not certain) to happen.

You wouldn't believe how many times people try putting together collaborative standards or projects in the text MUD world that fall apart because nobody wants to compromise on 'their vision' (which usually means 'no vision'). The shot the graphical world has is that there's a much larger pool of people and thus potential projects to choose from. Of course, it's a lot harder to create new art than it is to write new descriptions too.

Anyway, why is that so depressing? What's the real attraction in having Everquest-style player numbers unless you're trying to make money? You never interact with most of them anyway.

--matt

[Matt Mihaly]

6.

Peter S. Jenkins wrote:
This means that if there is to be a non-fragmented metaverse, it will likely be owned by corporate interests, since it would be extremely expensive to implement and operate. This of course leads to concerns about freedom of speech. As I noted in my paper, we need to recognize that not granting freedom of speech in MMOG's which are already more than games due to commodification, will set a dangerous precedent for the coming corporately owned metaverse.

Come on, what is this? Some of you guys have read Snow Crash way too many times I think. I see absolutely no reason to posit some single all-encompassing metaverse. Has there been a single trend towards it? I've not noticed one myself.

--matt

[Scott Jennings]

7.

I gotta go with Matt here, it's somewhat amusing to see open source MMOs touted as some kind of new, radical concept when they've been around for, um, what, 25 years now? All of which either (a) never got more than 10 users or (b) got more than 10 users and then immediately developed "freedom of speech issues".

The adventures of some ex-UO players who thought nirvana was achievable if only they ran their own server is illustrative: http://ipyuo.com/

[Scott Jennings]

8.

Also, I'll be more inclined to believe in the coming of the "metaverse" when the AOL and MSN IM clients can work out talking to each other.

[Cory Ondrejka]

9.

There is a common confusion when discussions combine "open source" and "interoperability." The two are related but neither is dependent upon the other. Interoperability requires shared standards, federation, identity management, etc. Open source is simply one methodology for creating the systems that then may or may not interoperate.

The web is the most obvious example of this. IP, SSL, HTML, XML, etc, are all shared protocols that allow a wide range of products to communicate with each other. Web pages served by Apache (open source) can (generally) be read by Internet Explorer (closed source). Firefox (open source) can (generally) read pages served by Microsoft Web Server (closed source).

Personally, I agree with Peter that the long term dream is something very much like the World Wide Web where many (most) digital worlds choose to utilize shared standards. Some of those worlds will be open source, some will be closed source, but the standards will provide Peter with the options and choices that he desires.

As for the test case of what happens when incredibly diverse endeavors coexist in a shared space, Second Life does provide a lot of examples. We've seen those interested in cohesive experiences moving into estate ownership, although this is driven as much by the desire for more control over performance as the desire for separation. Many large scale endeavors have been successfully carried out by multi-person teams in SL, although Peter is correct that games have lagged some of the other group projects -- although the folks at Bedazzle seem to be making excellent progress.

We've certainly talked about this issue before. Making games is a difficult process and Open Source methods seem ill suited to making large scale games. Economic motivations help keep large teams together over long periods of time. Oh the other hand, Bedazzle's plan is make their game free (although they'll still get dwell awards, sell their really great builds, etc). So, we'll see.

[Ola Fosheim Grøstad]

10.

Jennings> Also, I'll be more inclined to believe in the coming of the "metaverse" when the AOL and MSN IM clients can work out talking to each other.

Oh, but Metaverse will not be an application, it will be the OS. And you won't be able to replace it with Linux. Direct-3D is just the first step... :)

[Scott]

11.

I agree, and would actually argue that the Web is the penultimate triumph of the Open Source philosophy. I mean, here's something that quite literally changed the world, and no one company or government owns it.

But for a "shared multiverse", there has to be interoperability. I can't send /tells to Second Life from World of Warcraft, for example. There's no technical reason why I can't, but there are very powerful business case reasons why. Those reasons will have to be trumped before we truly see a shared online MMO space.

(And I do believe it will eventually happen, just not for 50 years or so. It will probably come from something totally out of left field. Like, you know, HTTP!)

[Peter Ludlow]

12.

>interoperability

Well, how plausible is this. Protocols are established for interoperable game environments, allowing a single game tag to work in all the games, keeping track of your skill level in each game (with some games allowing levelling credits from other games). Protocols also enable travel between 2 games if both game operators agree to it, so intergame portals are established. Then if game development splinters into a million shards due to creative differences etc., it would still be possible to travel freely among the shards, participate in combat, possibly level in foreign shards etc.

[Cory Ondrejka]

13.

Peter,
Quite plausible, I thing. The obvious concerns are similar to web security issues: what about hostile shards that drop you 20 levels just for visiting them? or, steal your gold? allow free duplication of anything? etc, etc. The good thing is that smart people are working on these problems :-)!

[Scott]

14.

There's something similar to this developing in the pen and paper space: the D20 license. Basically, pen and paper game systems are agreeing on core game rules, which Wizards selflessly or selfishly (depending on whom you talk to) made into a form of open source.

Another development in this sphere is Neverwinter Nights, which has multiple game servers operated by anyone who cares to. Most of them require "server vault" characters, or characters stored at that particular server instead of on the user's machine. The original plan was for Bioware to actually host a "vault" of characters, and those characters would be able to journey between different player-run servers. The obvious reasons (most of which Cory listed) torpedoed that.

[Andrew Burton]

15.

Two quick comments:

> ...my hunch is a good open-source virtual world
> codebase wouldn't result in one enormous world but
> a zillion tiny ones.

A tiny zillion ones that use TCP/IP, Telnet, and other standards for their protocol. That's the big difference, I see, between MUD's and and MMOG's. You can get into MUD via Telnet, but each MMOG requires special software. (Probably parroting Cory here, apologies.)

Think about it this way, though. Prior to Apple and Microsoft, every PC had its own special operating system. There was Commodore, Tandy, Apple, Mac, IBM, etc. As thing started to shuffle about, and more people used PC's, leaders emerged. Now we're really down to three operating systems: Windows, Mac, and Unix -- and these days, there not much diffeence in Mac and Unix.

Let a thousand worlds form with a thousand code bases. In a few years they'll settle out, standards will arise, and some keen hackers will create the MMOG equivalent to gAIM and Trillian.

> The obvious concerns are similar to web security
> issues...

Too true, but a lot of those security issues can be overcome with social, not technological, solutions. Ebay, for instance, has no way to make sure a Paypal order goes through, but they provide a means for people to know who usually is honest and who's not. I can easily see trade guilds, store hubs (ala Yahoo!Store, zShops, etc.) forming for virtual worlds.

In the absense of a MMOG Guild to regulate, I think the Blogosphere and Google will do well to let people know which worlds are honest, and which ones you should stay away from.

[Andrew Burton]

16.

That should be Bidpay, not Paypal. Ebay owns Paypal, so of course they can track that. :)

[AFFA]

17.

How would this be different from the "emulators" for UO, EQ, and DAoC that are already out there and already open source?

Even though they're open source and have no monthly fee, there's been no mass exodus to "free" servers (at least none that I'm aware of). If even making a server "free" hasn't attracted most players, what would they have to provide to broaden their appeal? I doubt official sanction would suffice.

[AFFA]

18.

Note also that there are plenty of open-source MUDs. MUDs certainly have more variety than graphical on-line games so far, but even the msot successful aren't pushing anywhere near 200k players, even when they're free to play... You could argue that the total MMORPG market would widen, but I can't immediately see how anyone would profit from it.

The main problem with trying to make an open source game is the art. I know plenty of people who will Code For Free (and even more weekend game designers). But I don't know any good artists who will do 3d modelling for free. And public domain digital art tends to be... substandard.

[Matt Mihaly]

19.

Andrew Burton wrote:
You can get into MUD via Telnet, but each MMOG requires special software.

That doesn't seem to make a lot of sense as a distinction. Are you telling me that if we wrote our own simple protocol and had our java client use that instead of telnet, keeping everything exactly the same from the users point of view, you'd then call our games MMOGs instead of MUDs? (Not that they aren't synonyms anyway...)

--matt

[Nathan Combs]

20.

Peter>

Protocols are established for interoperable game environments, allowing a single game tag to work in all the games, keeping track of your skill level in each game (with some games allowing levelling credits from other games). Protocols also enable travel between 2 games if both game operators agree to it, so intergame portals are established. Then if game development splinters into a million shards due to creative differences etc., it would still be possible to travel freely among the shards, participate in combat, possibly level in foreign shards etc.

There was discussion on aspects of this here:

http://terranova.blogs.com/terra_nova/2004/03/should_mmog_pla.html

Where Ted coined the "avatar transfer protocol" phrase.

IMO, interoperability on some levels seem plausible, howerver, outright transfer of characters with stature and assets between worlds seems problematic... Perhaps, I'm a bit too jaded from having sat through an unfair share of AD&D multiverse negotiations back when I played D&D a very long time ago.

There needs to be an authoritative (and trusted) broker to facilitate it. A few observations I can recall:

1.) folks hate giving up something (stature, assets) unless they think its going to be worth it. Implications for target world (going to) stability, "fun". Contrary to first assumptions, however, players quite often don't like moving to worlds perceived as "circus worlds" (GM's offer too much goodies and XP) - they know it would be a bear getting 'em off of 'em into other worlds.

2.) GM's instinctively hate accepting powerful magic/items from other worlds. And though they can usually adapt encounters to these changes, there is a cost to them in re-biasing the world.

...

[TL]

21.

History... always repeats...?

Circa mid-late '90s saw a kind of boom in graphical multiuser spaces. I remember sitting in on tech sessions at Earth to Avatars and hearing lots of talk about the metaverse, portability, open platforms, interoperability, etc. Just thought I'd toss in a few tidbits from the past ;) There are several others that read TN regularly that were also there and could probably give some interesting historical insights on this (and why it failed).

"Having a universal avatar would mean that, regardless of the virtual worlds users choose to inhabit, they'd be able to maintain an invariant identity, appearance, and set of belongings and preferences across those worlds. All of this has obvious implications in the marketplace and any other virtual space. Mark Pesce, the father of VRML, explains: 'In the same way that your Social Security number acts as an absolute reference point in so many parts of your physical life, these avatar standards will create a way to establish that reference point in the virtual world.'" - from Wired online, 1997

See also, Universal Avatars Proposal, VRML, Virtual Humans Initiative, Living Worlds

[Peter Ludlow]

22.

so how come that didn't happen. what are the obstacles to it happening now?

[Ola Fosheim Grøstad]

23.

Because nobody, but those who read too much SF wants it? Players create new characters/avatars all the time, and they enjoy playing up new ones.

What you need is an infrastructure for players to keep track of their gaming-friends, so they can find their old buddies when they switch games, and for MMOs to adopt the equivalent of the good old InterMUD protocol. Terranovans have a lot to learn from the 1990s USENET archives...

(Besides, the vision is rather naive too, from a design/technical/commercial/privacy point of view.)

[Mike Rozak]

24.

Metaverse - What do you hope to gain with a metaverse, as an author and/or player? Will it be worth the pain and compromises?

For an extreme example: Can a character and possessions really move between Second Life and EQII? Why would someone want to do this? What APIs/standards would be necessary? What information would be lost as a characer is moved? What would it do to game balance? Etc.

Realistically, the only information you can transfer from SL to EQII is the name and gender. EQII from SL can include the PC model and equipment models; transferring the actual equipment properties and scripts to SL can be done, but with much pain and suffering, and huge consequences to the other players in SL.

Open source - There are several "open-source" virtual world projects on MMORPG.com. They don't seem to go anywhere...

Part of the problem is that most programmers and artists that can work on open source projects can only devote 8 hours a week (or less). I'd rather have 1 programmer working 40 hours, than 20 programmers working 160 hours (8 hours each).

Plus, most of the open source programmers are learning their trade, and not very skilled. I'd rather have 1 really good programmer than 5 mediocre ones.

On top of that, the more people in a group, the less efficient it becomes. A group of 10 programmers are half as efficient as they should be due to their stepping on one anothers toes, waiting for other parts of the project to be completed, management requirements, bickering, etc.

Not to mention that a group of 10 people gets pulled in 20 different directions since everyone has their own somewhat-divergent agenda(s).

Having said this, if you could get a handful (<= 5) of top-notch people working full-time on an open source VW, they could do wonders. They would be worth an army of 1000 part-time open source programmers. The probability of this is EXTREMELY low.

[magicback]

25.

One benefit of a metaverse with a VW "passport": single billing account.

Sure in each VW, the "visa" may determine what is allowed to be transfered across VWs and what is not, but the RL person will get one RL bill (assuming commerical VWs).

I don't like getting one bill each from SOE, Blizzard, etc.

[Jerry P]

26.

There's a good audio interview on IT Conversations with Dick Hardt, CEO of SXIP, that gives food for thought about a future "avatar transfer protocol" (though MMORPGs aren't mentioned). The subject of the interview and the goal of SXIP (Simple eXtensible Identity Protocol; acronym pronounced "skip") is "transferring reputations from one social network to another" and creating a common, verifiable and portable online identity. The difficulties, not just the benefits, of implementing something like this are pretty well addressed.

[Cory Ondrejka]

27.

Mike R raises some good points, although as Jerry points out, being able to move identity is what matters. I also agree with magicback that ease of access, simplified billing, &c, are worthwhile as well.

In other words, don't think about this as "bring my Cory Linden character into EQ2" but rather be able to move whatever identity (which doesn't need to be Cory Ondrejka, but could be another verifiable pseudonym) into EQ2 and be able to leverage my reputation, social network, payment options, &c.

Another way to think about this would be if Amazon and eBay started using federated identities, so that reputations and payment options could be maintained between them. They would both benefit from this (even though they are competitors in many ways) and would increase their leads over other sites.

[Peter Ludlow]

28.

In social MMORPGs this kind of federated identity happens anyway -- your reputation follows you around, and this too is an old phenomenon. Remember when SamIam was banned from MediaMOO for alleged crimes on LambdaMOO? That was about 10 years ago.

But I'm imagining something like this...

First, I'm not thinking of portability between EQ and SWG, but rather a consortium of open source mini-platforms that have agreed to certain protocols. So this is not about existing games, but about building a new metaverse from new minigames...

Second, I imagine that the game tags keep track not only of skill levels and wealth, but also where the levelling took place and the origin of the currency (imagine each minigame had its own currency). Skill levels from a game that is overly generous might be devalued based on combat outcome. For example if a level 5 fighter from game alpha defeats a level 5 fighter from sim beta, then the points from game beta are devalued accordingly. Alpha can churn out level 10 fighters, but they will be valued less and less when they travel. Of course it's still possible to game this by holding back you people from battle, but the system could also evaluate your score based on the number of intergame battles you've had, and the rankings of the people you have or haven't defeated. I guess this is like those football power rankings in some ways.

Finally, each minigame would have its own currency, the value of which would fluctuate. minigames with stable economies might form a supercurrency from a basket of their mini-currencies, sort of like the Euro.

[Ola Fosheim Grøstad]

29.

Cory Ondrejka> if Amazon and eBay started using federated identities, so that reputations and payment options could be maintained between them. They would both benefit from this

Exactly how would Amazon benefit from this? They would most likely loose trust by sharing information with a more dubious service such as ebay. Besides, if you start to mix information about payment and reputation you enter a space which is heavily regulated for privacy reasons (outside the US). The same goes for identity information. Another "solution" looking for a problem to solve...

[Ola Fosheim Grøstad]

30.

Peter Ludlow> I imagine that the game tags keep track not only of skill levels and wealth, but also where the levelling took place

This is all well, but why would you want to do it? What problems will it solve? It just creates a lot of problems, and doesn't really solve any.

Problems:
1. assumptions about interoperable mechanics which severely limits game design
2. allows bypassing of content which wrecks game design
3. standards limit technological progress, which makes the system less competitive
Etc.

[Peter Ludlow]

31.

>standards limit technological progress

I guess that's why the internet is a total bust.

[Ola Fosheim Grøstad]

32.

Peter Ludlow> I guess that's why the internet is a total bust.

Eh, yes. Interoperability has been more important than innovation for the core applications which require a more distributed approach. The most widespread core technologies (email/WWW) are very low-tech and primitive.
They also are rather resistant to change, e.g. although MSIE supports XSLT, very few website use it.

[Andrew Burton]

33.

> That doesn't seem to make a lot of sense as a
> distinction.

Forgive, matt, for not being well versed on MUD's, MOO's, and such. I tend to lump MUD's, MOO's, MUCK's, etc. into a "text-based world" category in my mind. What I meant to say was that if I want to play in a FurryMUCK and then in to TransformerMUSH, I don't need to install any large pieces of proprietary software. I just use a telnet client (that was the way it worked last time I played in a MU*).

With the advent of multi-user platform tools like VRML, XML, OpenGL, wxPython, Shockwave/Flash and Mozila, it only seems like a matter of time until someone will come up with an open source client/server package that is as prolific and simple to use for 3D worlds.

It disturbs me though that this conversation has seemed to go in the direction of complete interoperability. Software like browsers, Trillian, gAIM, and syndication readers seem to point in a direction of access ubiquity, not content interoperability.

I think "The Matrix Online" working with AIM shows a trend toward this. People, I doubt, have any desire to import their AIM buddy icons into the game, instead they want to able to speak with people outside the game while they're inside.

[Mike Rozak]

34.

Andrew Burton wrote - What I meant to say was that if I want to play in a FurryMUCK and then in to TransformerMUSH, I don't need to install any large pieces of proprietary software. I just use a telnet client (that was the way it worked last time I played in a MU*).

If this is the goal, then what you're really doing is creating a client app with a public communication standard to the server. This client is not necessarily free, but it will be cheap. In other words, the equivalent of Netscape or IE, but for virtual worlds.

The reason why this HASN'T happened with MMORPGs yet is because most of what EQII or WOW installs on your computer are textures, audio files, and 3d models. These models can be downloaded on demand, as Guild Wars does, but if you want GW/EQ/WoW quality graphics, you will eventually need to send many GB down the slow and expensive Internet pipe. 56Kbaud modem owners won't be happy, nor will whomever is paying the Internet bill.

Technically there is nothing stopping this. I am working on something like it EXCEPT it relies on static 3D scenes instead of full animations. (I am targeting the hobbyist who wants to make a VW or IF title, so eliminating the animations makes life much easier for the hobbyist. Think of it as a MUD or text-IF crossed with Myst III's or Zork Nemesis' 360-degree surround images, but with hobbyist-quality graphics.)

For a 3D MMORPG-style client to become a standard requires a bit of political maneuvering. Having worked on an industry-wide API, I can go into excruciating detail about the work involved, but it would inevitably put everyone to sleep.

[Ola Fosheim Grøstad]

35.

Mike Rozak> This client is not necessarily free, but it will be cheap. In other words, the equivalent of Netscape or IE, but for virtual worlds.

Ah, but this exist. Adobe Atmosphere also has an open source server, lets you teleport from world to world etc.

[Mike Rozak]

36.

Ola Fosheim Grøstad wrote - Ah, but this exist. Adobe Atmosphere also has an open source server, lets you teleport from world to world etc.

Forgot about that one. I haven't looked at it for awhile. I don't recall seeing that the communication standard was made public, but it could be. I suspect that the client wouldn't be up to the task for most MMORPGs, but it's certainly on the way.

[David Evans]

37.

This is a great thread. I read the same one every few years and get all excited that something is changing, and nothing ever does. T.L and I used to hang out in MediaMoo. I met Mr. VRML way back in the day and we all got excited about the crappy language, lame tools and terrible user experience. I pitched SGI on a meta-verse concept in 1997. They were very close away from giving me $100,000 to explore the opportunity but the hardware and management wasn't up to the task. In hindsight, they made the right decision, it was too early.

Scott Jennings nailed it. Look at IMVU and Yahoo! Chat. The future of virtual spaces will be based on environments based on real-time communication. 100 million-plus people use chat clients, hows that for an installed base?

Hundreds of thousands of people buy Evercrack, but that's a drop in the bucket. The rest of us wouldn't be caught dead playing those games.

Yahoo! Chat is interesting. It's avatars are quite customizable, and Yahoo! is smart, they are doing deals with consumer brands like The Gap, Coke, etc, and have branded clothing, backgrounds and ties into music industry.

I can't believe no one is talking about hanging a virtual space off of Myspace or Friendster!

Of course it will be owned by corporate interests. Like Mike said, I'm not waiting for you to come home from work to fix a bug in the system. Linux is an example of open source working, but then again they have Linus and OS hackers are fiercely driven. Which of you will lead us forward?

You don't think that Starbucks is thinking about virtual spaces? They have wifi, that is the first step towards connecting their customers.

SXIP is interesting, but Trufina is where it's at, believe me. I just attended a panel on identity verification and management. That's a hot topic being driven by the online dating industry of all places.

Someone said, "The Web is the penultimate triumph of the Open Source philosophy." Huh? We're just getting started.

Adobe Atmosphere? Adobe is a document management and creative tools company. They know nothing about virtual worlds and games.

Anyway, my two cents.

38.

http://craps.cnt-group.com > http://craps.cnt-group.com
http://craps.cnt-group.com > craps