Question RE: Discouragement vs. OSS

[Bear]

I was revisiting the Crystal Space project last night ( http://www.crystal.sourceforge.net ) and I was impressed by the degree of improvement they have made recently. I also noticed that several projects for new games have been started based on the CS engine for modeling.

Now, granted, this critter is still very much in beta. But I am wondering what effect, if any, this will have on the commercial developers. I have often made a pest of myself ranting about "re-inventing the wheel" when various game companies spend untold dollars and hours to develop a game engine that is functionaly equivalent to everyone else's anyway.

Is there any chance that the game industry will follow the lead of IBM and RedHat? By which I mean, is it possible that some commercial developers might choose to adopt an Open Source approach to the nuts 'n bolts of the game engine, and choose to focus their creativity on the actual gameplay?

One game project in particular called Planeshift ( www.planeshift.it ) looks like they actually intend to follow through with it. The stated objective is to produce an open-ended, player driven, completely FREE mmo-RPG. Whether or not they suceed, someone is sure to put out a game on this basis eventually.

What kind of effect is this going to have on the commercial releases? Do you anticipate any significant economic impact?

I remember what happened with NWN. Once a reasonably effective set of tools hit the masses persistent little game worlds started popping up like dandelions. And frankly, many of them are just more *fun* than most commercial releases.

What attracts me to this kind of thing is Free access (I am both poverty stricken and a miser), and freedom to innovate in any way that pleases me without worrying about constraints of time or money. These are OSS projects. They will never go away, they are unlikely to ever be totally abandoned, and this kind of amateur involvement breeds a lot of interest.

I recall Doom, and I reflect on how many player made levels are out there, many still being played. Also Heretic, Hexen, Hexen II, Heretic II, Quake 1,2,3,4,5,6,7,8,9,....... and all the others.

The thing is, once something is made and released in the public domain, it sits there forever. Even something that hasn't been finished is fair game for someone to pick up next year and resume.

Crystal Space is currently about as technically sophisticated (to my untrained eyes) as the early Quake II engine. But that may be good enough to build on. Art work is expensive, but there are, even now, a lot of leftover public domain and OS artwork out there. And anyone can take bits and pieces and build level maps. Even I used to do it. And OS tools for 3D modeling, scripting, etc. are becoming increasingly sophisticated as well.

What kind of impact does this have (if any) on the commercial gaming industry's planning and preparation?

Bear

[Vengeance]

Hi,

I am actively involved in the Planeshift project and I appreciate the mention here. It is always a little surprising when I'm reading some website and run into a Planeshift mention. :-)

One thing to know is that the Crystal Space renderer is being completely rewritten to take advantage of c. 2003 technology instead of c. 1998 technology. Bumpmapping, dynamic lighting, realtime shadows, atmospheric diffusion, specular highlighting, etc. are all built in. It is about 75% done right now and should take Planeshift, as well as other projects using it, to a whole new level of graphical polish.

As far as impact on the market from free software, I think Planeshift has real potential to do that. The current technology preview, Molecular Blue, is free and publicly available to download or build from scratch. We currently have >90,000 player accounts of people who have tried it out.

Planeshift's next release, Crystal Blue, will be around the first of the year. It will include combat, NPC AI, progression, magic system, and should be a very basic but playable game at this point.

Thanks,

Keith Fulton

[MahrinSkel]

Well, that's really two different questions:

1) Is there a place for OSS software in bringing down the costs of developing and fielding MMO games?

Definitely, in fact it's already being used for exactly that purpose (both Camelot and Wish run on Linux, saving quite considerable amounts of money compared to liscensing an OS from Sun or Microsoft). Wish will be using the commercial version of the OSS UI technology Qt from Trolltech, and I've been investigating other OSS technology pieces like PostgreSQL for years. As an LGPL package, Crystal Space certainly has potential for teams on too much of a shoestring for the NDL, iD, or LithTech engines that don't want to bind themselves to the Torque engine liscense requirements, whether those projects are non-commercial or not.

2) Are non-commercial MMO projects likely to "poison the well" for the subscription-based business model?

Unlikely in the extreme. OSS can do a fantastic job of building tools and building blocks where the design requirements are fairly cut and dried, but whenever arbitrary aesthetics get involved they break down, with every difference of opinion leading to a split team and a forked tree.

It's hard enough to deal with the feedback and create a cohesive development effort for a large-population MMO when there's someone signing the checks to call the shots and end the arguments. Truly free OSS-based projects *might* be a threat to the business model for very small games like eGenesis and Near Death Studios, but I think it's more likely those tools will become the enablers of that business model. I think that's a great thing, a source of innovation the industry absolutely needs. But as much as I wish the people working on Planeshift and similar projects well, I doubt they'll succeed.

There's been an Open Source MMO package for years: Sphere (now known as WorldForge). There have been worlds running on it, none has managed to scale much past the free-MUD level. An MMO is more than rules and content and the talent to create them, it's all the annoying administrative details, the mind-numbing, grinding, even soul-destroying work of *running* the damned things.

--Dave

[Bear]

I have been considering the implications of OSS a lot lately. I became an adult about the same time that IBM was laughing their butts off at this crazy Bill Gates kid, and the news commentators were mocking the very idea of a computer small enough to fit on a desk. I have purchased a godawful pile of software in my time, and spent many man-hours checking out shareware and freeware packages.

Now we have OSS. Which is basically the academic model of scientific research adapted to the software industry. The general mainstream reaction appears to be about the same as it was when someone suggested that this dinky upstart called Microsoft might someday take a substantial share of the market away from IBM.

I am not trying ot be argumentative for a change. But *I* am not a programmer, I just use the things. And some parts of your response are unclear to me.

Originally posted by MahrinSkel
Well, that's really two different questions:

1) Is there a place for OSS software in bringing down the costs of developing and fielding MMO games?

Definitely, in fact it's already being used for exactly that purpose (both Camelot and Wish run on Linux, saving quite considerable amounts of money compared to liscensing an OS from Sun or Microsoft). Wish will be using the commercial version of the OSS UI technology Qt .........As an LGPL package, Crystal Space certainly has potential for teams on too much of a shoestring for the NDL, iD, or LithTech engines that don't want to bind themselves to the Torque engine liscense requirements, whether those projects are non-commercial or not.

This reads to me as if you believe that OSS is entirely appropriate as a cost saving measure, but that in terms of quality it is inferior by definition. Am I mis-reading this? I get the idea that you consider Crystalspace, or any other OSS game engine, to be a good option for those who can't afford a "real" software package.

Again, I am not trying to argue, I just want to clarify. If I am reading it correctly, then you might want to re-examine some assumptions.

I am writing this using Konquerer running on Knoppix. I have taken to doing all my web surfing in Linux. Not just because it is safer, but because the internet connection is less likely to collapse on me, I get faster downloads from the same ISP, and the web pages load faster.

Earlier I was using Abiword to compose a legal document. It struck me that the copy of MS Word 2001 that I bought not long ago was slower, much more crash prone, and provided fewer useful tools for what I was doing. Not to speak of OpenOffice.org, which is improving rapidly.

Considering that several cities in Germany, at least one South American country, and the state of Mass. are all in the process of switching to OSS it makes one think. Perhaps right now the OSS game software is inferior (I would not know about that) but how will it compare next year?

Before writing this I re-checked some sample demos from crystalspace. I noted the real time reflective surfaces, projectiles that act as light sources, slippery surfaces, moving fog, realistic flames, etc. And Tuxracer, the greatest game ever made in the history of computers, uses a different OSS engine that produces particulate snow that leaves real tracks, along with some *very* nice inertial physics.

Are the commercial, proprietary packages that much superior? Inherently? And will they still be better next year?


2) Are non-commercial MMO projects likely to "poison the well" for the subscription-based business model?

Unlikely in the extreme. OSS can do a fantastic job of building tools and building blocks where the design requirements are fairly cut and dried, but whenever arbitrary aesthetics get involved they break down, with every difference of opinion leading to a split team and a forked tree.

Which I guess is my whole point in a nutshell. When the software is free, and you are not under any deadlines, and people are doing it because they *want* to do it, not because they need to.....what will happen to the entire approach?

I haven't tried any of the games based on Worldforge, or Crystalspace, or any others because I simply have been too busy for in-depth gaming lately. But I am reading that there are a lot of small, persistent worlds out there running NWN. For free. Just for the fun of it. Individual people expressing their creativity.

Shortly after Doom came out someone made a RPG wannabe called Amulets & Armor. It was a noble effort, but since it was based on a (Doom) engine that wasn't designed for roleplaying it never went very far. What would happen if Bioware were to Open Source the engine for NWN? Or more likely, what would happen if someone took the engine for Falcon's Eye and turned it into a decent NWN clone, with equivalent tools?


It's hard enough to deal with the feedback and create a cohesive development effort for a large-population MMO when there's someone signing the checks to call the shots and end the arguments. Truly free OSS-based projects *might* be a threat to the business model for very small games like eGenesis and Near Death Studios, but I think it's more likely those tools will become the enablers of that business model. I think that's a great thing, a source of innovation the industry absolutely needs. But as much as I wish the people working on Planeshift and similar projects well, I doubt they'll succeed.

There's been an Open Source MMO package for years: Sphere (now known as WorldForge). There have been worlds running on it, none has managed to scale much past the free-MUD level. An MMO is more than rules and content and the talent to create them, it's all the annoying administrative details, the mind-numbing, grinding, even soul-destroying work of *running* the damned things.

On the other hand, when no one is signing the checks, because there are no checks and no fees, how hard is it to simply say "This is the way it is gonna be, if you don't like it, go play somewhere else for free"?

How big is big enough? Does 500 count as massive? Is massive even necessarily better? These questions keep coming up over and over. But the fact is, when you give the public tools that let them design their own game, or their own word processor, or whatever, then someone is going to make one.

Right now people are paying monthly fees in order to experience online games that are run by arbitrary fiat. How much convenience, and how much of the "massiveness" would they be willing to sacrifice in return for a game world that allows the players a real voice in shaping how things happen? If they could get it for free?

Bear

[Vengeance]

But as much as I wish the people working on Planeshift and similar projects well, I doubt they'll succeed. Without sounding too defensive here, what is your definition of success?

If it is defined in terms of user acceptance or popularity, I'd point to our numbers: We have 92,000+ accounts in our pre-alpha demo by now; we have a very active guild setup with more than 600 guilds formed so far; for months we were in the top 10 most downloaded SF projects along with all those P2P network clients; searching on "free mmorpg" in Google returns Planeshift as the first entry.

If you are defining success in terms of ever finishing such a huge project, I'd point to our team: We have a very active dev team, including people from 11 countries. Most of those people are pros in the computer or gaming industry, and our average age is about 30. The core team of 6 devs has worked together on this project for two straight years and know we have another 2+ years to go.

Of course, with the vagaries of free software projects--wives, salaries, careers, etc.--momentum is an ephemeral thing. Popularity is fleeting. Having spent no money on PR and nothing but time in development, Planeshift is just a bunch of guys around the world, coding by themselves in the dark, connected to each other only by irc and email and a common vision of what *could* be.

So what is success and what can the PS team do to maximize our chances?

- Keith

[MahrinSkel]

Originally posted by Bear
This reads to me as if you believe that OSS is entirely appropriate as a cost saving measure, but that in terms of quality it is inferior by definition. Am I mis-reading this? I get the idea that you consider Crystalspace, or any other OSS game engine, to be a good option for those who can't afford a "real" software package.

Again, I am not trying to argue, I just want to clarify. If I am reading it correctly, then you might want to re-examine some assumptions.

[...]

Are the commercial, proprietary packages that much superior? Inherently? And will they still be better next year?
If Crystal Space can truly deliver everything that NDL, LithTech, etc., provide, and do it cheaper, it's certainly worth using. But that isn't a game, it's a game engine. A game engine needs to support the capabilities of the hardware 3D cards, provide a UI API, and have hooks for resolving physics. It's a target that OSS can hit.
Which I guess is my whole point in a nutshell. When the software is free, and you are not under any deadlines, and people are doing it because they *want* to do it, not because they need to.....what will happen to the entire approach?
Nothing nearly as shiny and egalitarian as you seem to be expecting. In making a game, you are forced to make aesthetic choices, some of them arbitrary. In a commcercial project, the politics surrounding the choices can get quite intense, but at some point the Golden Rule (he who has the gold, makes the rules) takes over and the decision is made. In an environment where everyone is working on it because they want to, odds are those that don't like the decision will split from the team and start their own effort.
How big is big enough? Does 500 count as massive? Is massive even necessarily better? These questions keep coming up over and over. But the fact is, when you give the public tools that let them design their own game, or their own word processor, or whatever, then someone is going to make one.

Right now people are paying monthly fees in order to experience online games that are run by arbitrary fiat. How much convenience, and how much of the "massiveness" would they be willing to sacrifice in return for a game world that allows the players a real voice in shaping how things happen? If they could get it for free?
Why does this discussion always take on ideological overtones? If I'm wrong, the discussion is pointless, free games will put us all out of business. If I'm right, the discussion is also pointless, the free MMO engines are competing with MUD's, not subscription games.

I *want* to see those projects succeed, it's the only real chance we've got to see a proving ground for innovations. But I also want to work on the big scale games, make games that hundreds of thousands of people will play.

--Dave

[MahrinSkel]

Originally posted by Vengeance
Without sounding too defensive here, what is your definition of success?

If it is defined in terms of user acceptance or popularity, I'd point to our numbers: We have 92,000+ accounts in our pre-alpha demo by now; we have a very active guild setup with more than 600 guilds formed so far; for months we were in the top 10 most downloaded SF projects along with all those P2P network clients; searching on "free mmorpg" in Google returns Planeshift as the first entry.

If you are defining success in terms of ever finishing such a huge project, I'd point to our team: We have a very active dev team, including people from 11 countries. Most of those people are pros in the computer or gaming industry, and our average age is about 30. The core team of 6 devs has worked together on this project for two straight years and know we have another 2+ years to go.

Of course, with the vagaries of free software projects--wives, salaries, careers, etc.--momentum is an ephemeral thing. Popularity is fleeting. Having spent no money on PR and nothing but time in development, Planeshift is just a bunch of guys around the world, coding by themselves in the dark, connected to each other only by irc and email and a common vision of what *could* be.

So what is success and what can the PS team do to maximize our chances?
You're asking the wrong person, I already sold out. I define success as involving hundreds of thousands of players and wheel-barrow loads of money (never mind that I don't actually get the money, I do get to make a living while thinking about nothing but games every waking minute). It's good for the ego, working on something so many people prize so highly, "Online Game Designer" is one of those infinitely cool "Rock Star" jobs that computer geeks all over the world would like to have. But that is *my*definition, how do *you* define success?

If you define it as creating something that puts the commercial efforts out of business, you're probably going to be disappointed. If you define it as making a game that you like to play, and that some number of hundreds or maybe thousands of others also enjoy, and doing it for the sheer love of the work, then that's an admirable and worthwhile goal that I would agree is achievable.

The biggest danger you face, in my opinion, is in trying to do too much, and get too large. Volunteer admin systems never seem to scale past a few dozen part-timers, supporting a few hundred players. Try to do too much for too many, and you're going to have problems. If you want to avoid that problem, then focus on maximizing the efficiency of the tools and minimizing the need to use them.

--Dave

[Firedrake]

Sorry to revive a dead thread, but I ran across it and had a few things to say.

I think there is some confusion here as to what we are talking about with Open Source. People seem to be slipping between the software itself (an engine) and small projects using this engine. I think we can all agree that a small hobbyist project team using an OS engine to build a game will never take on the big players.

However, an OS engine, such as CrystalSpace, given some time and attention has every possibility of being the foundation of the next great MMO (well more likely the one after the one after the next, but you get the idea). Linux was brought up, and it serves as a perfect example. What started off as a hobbyist project, fit the need of a larger player (GNU initially) and gained respect and some acceptance among the techies. Once Apache was paired with Linux (its killer app) a widespread acceptance began. Once some major players took notice, and poured their own resources in, Linux became a billion dollar business.

There are some essential ingrediants to this success story.
One is a financial need. MMOs are incredibly expensive to create from scratch. There is every precedent to point to a future where companies largely license an engine from people who specialize in creating them. There is no doubt that there will be room in this space for both proprietary and OS engines.

Second is the ability to do its job, and I think this is where people get confused. Just because an engine is OS, does not mean it cannot be used and developed by mega corporations. They get a free piece of software that puts them months if not years ahead in their development cycle. They are then required to give back their programatic changes, which helps the underlying engine. They are not however in any way forced to release their content. So their game is still unique. If you want to argue graphics, AI, competative advantage etc. I will only say that this entirely depends on how the engine is designed (are these things part of the engine and therefore subject to its license). Generally this is not the issue people think. For example, I am not required to give the source code for my web application to apache, just because I use their server. I might be in a different situation if I use GPL, but it largely depends on architecture/design. LGPL is a good compromise, but I am getting ot.

Finally, the business model seems to be the point of most contention. Just as several people pointed out MMOs are primarily services. Players pay for not only the right to connect to your servers, but more importantly are paying for your content. Just as you do not reinvent servers or networks, there is no business neccessity in creating your own engine. If you can commoditize that piece as well your cost of entry is greatly reduced. This also has the benefit of allowing smaller players to enter the field. I again do not mean hobbyists. Now you are in a position to hire designers, artists, management, support etc. and your only development costs are game rules related. This is however still a business with many costs, and will live or die based on many conditions, but simply having OS software at its core is not one of them.

Chris Duesing