Slashdot Mirror
Game Engine Marketing Models Compared
death00 writes: "GameDev has an interesting story about the success of Garage Games Torque engine (the engine behind Tribes 2). I especially find it interesting to see the number of developers working on high-quality games based on the Torque engine. The basic premise is that Garage Games gives a full license of the Torque engine to a team for a project for $100 USD per developer. The only caveat is that you must publish any finished works through Garage Games. Perhaps id software might consider doing this with the Quake III engine once the Doom III engine comes out. From my understanding, the Quake III engine currently licenses for significantly ($250,000 USD) more than that. Instead of waiting 2 more years and GPL'ing the full source, why not license it for cheap after Doom III comes out, then GPL later?"
How is this a good question? It's simple.
why not license it for cheap after Doom III comes out, then GPL later?
Easy, they want to make money from it! If they can charge $250,000 and sell copies of it, there's clearly demand for that product and as a result id gets money (again, this is their goal). It's not like id gets anything from other companies licensing their software OTHER than money.
Of course, the price tag puts it out of hobyist's reach, but we all know that, eventually, we'll be allowed to look at the source for free.
Go Id!
A message from the system administrator: 'I've upped my priority. Now up yours.'
So the chose this alternate model. The $100/developer model will probably not work too well for iD, who offer the services of John Carmack for a day to all lisencees, as well as a lot of email support, etc. Would it be worth it for JC to make a trip to teach 4 developers to use the engine, get $400 in return, and then have the game cancelled in developement?
Besides, iD doesn't really do publishing. They publish through GT Interactive as far as I recall, so this revenue stream is not viable at all.
Ñ'
Um, you know, most businesses bigger than Bobs Lawnmower Repair Shop will use that $1000000 quicker than you can say 'cash flow problems'.
Often the free software development model is criticized for simply rebuilding what has been done already. And I feel that the release of the Quake engines and DooM engines have exemplified this very inadequacy.
I had hoped that we would see some really brilliant things come out of the GPL releases of these codebases, and, in reality some very good, cleaned-up clients have been developed. I certainly enjoy the mouselook, higher resolutions, and enhanced levels that have been developed from the DooM engine (see DooMWorld to see the kind of stuff that's out there). The improved QuakeWorld client I'm aware of is pretty nice. And Q^2 has a good Quake 2 client.
But these are just the obvious extensions of what was already done. The community now has (for the most part) all the source and tools that went into making Half-Life, the most successful game to come out of all of these codebases. Yet, to my knowledge, no project has arisen from the community to mold the next such game. How about another story-driven game that people would compare to Deus Ex? Or an all-out action game in the same vein as Soldier of Fortune? Or how about a freaking free software teamplay game that we compare to Counterstrike so that Linux users can play a team-oriented online FPS using free software only and not rely on WINE or WINEX? Or meld two free software projects and connect a Z-machine interpreter with the Quake engine and make a text-command driven story with a 3D view of the action?
These are things that would demonstrate just how momentous and visionary the release of the Quake source under the GPL was. Yet, all the community has managed to come up with is Quake++.
People slam my posts for being negative lately. That I'm ripping on people that have done good work. That's fine, I've got the skin for it. (Try USENET...) I admit that some really find refactoring and coding has gone into redoing the Linux Quake clients. But really, I hear plenty of bitching about how Linux (and other free OS) don't have good games and don't get the attention of the big game companies. Yet, when empowered to do new and exciting things and to make your own games, the group is content to simply recompile Quake for the Zaurus and call it a day. That's good work, for sure, but it's not the kind of work that's going to move free software forward and make it the kind of interesting world that non-free software people take a real interest in.
Again, I'm not making a judgment about the quality of the work that has been done. It's great. But now that you have the best raw materials from John Carmack, can we see real creativity out of the free software gaming world? (FWIW, I think CrystalSpace has done a good job of attracting some interesting new development.)
Curmudgeon Gamer: Not happy
A typical A-class game costs 3 to 10 million dollars to develop. I'd say that 3 million is too low a figure in today's market, though it was typical several years ago. Let's say 4 million is a base figure. $250,000 is 6% of the total budget. That's it. Six percent.
I'm not saying that the Quake III engine is state of the art, or anywhere near the only choice out there (frankly, it's the only engine that most gamers know about), but in the overall scheme of things, $250,000 isn't that much.
The engine, of course, is only maybe 20% of the work required to make a game. Fanboy-types refuse to believe this, but it's true. Art creation is much more time consuming, for example. And there's lots and lots of coding that has nothing to do with the core engine. I'd estimate that graphics-type code is maybe 10-15% of a typical game. For complex games that are less gameplay-shy than Quake, this could easily be under 5%. The reaction to that is usually "But what else is there besides the graphics?" which is greatly amusing to those of us developing games for a living.
If they can make money off of Quake3, why on earth would they GPL it?
They won't. id is not stupid.
But, based off past procedures, you can expect the Q3 source to be GPL'd within 1-2 years of the Doom3 release. They did it with Wolfenstein, they did it with Doom (sans sound code), and they did it with both Q1 and Q2.
But, like I said, they're not stupid. It's released well after its licensing value is near zero and under GPL. Want to make a commercial game using the old Q2 engine? Sure -- pay id Software $10k (a heavily reduced price) and you can have the exact same code as what's under GPL -- except that you're not bound by the GPL.
It's one of the few ways to make the GPL work for companies to make money. Of course, even then, you'll get rabid idiots decrying the fact that it's not GPL from the start, and that you can actually get a non-GPL license in exchange for cash.
Fortunately most people are sane and simply thank id Software for contributing to the public knowledge base in such a manner.
Well, your figure there puts development of the game engine alone at, what, six months? Now, look at it from the point of view of a company making a game. That's six months where they can't use screenshots to feed a hype machine. That's another six months tacked onto their release date. Another six months for their chosen technology to become obsolete, another six months for someone else to beat them to market.
Don't you think trimming those months off the start of the development cycle is worth $250,000? They apparently do. And that's not even considering the publicity from being able to say "we use the Quake 3 engine!"
Actually, as a Torque licensee, the support has been pretty good. But because of the way the GG team has set things up, they don't have to answer all the questions - instead, you post to the forums online, and sometimes the community addresses the issue, sometimes the GG team does. All depends on who's fastest on the draw it seems. (I have, however, seen some newbie questions go unanswered - mainly because if they would have bothered using the 'search' function, they would have found the answer right off. RTFM & RTFW! Jeez people!)
I've been happy as hell with it all - it's worked great, I've definitely gotten more than $100 in value out of it, and both the community support and GG's support has been great.
Davis Ray Sickmon, Jr - looking for something to read? Check out my three free novels at MidnightRyder.org
You don't need GPL to get creative with gaming by a long shot, and the good "freeware" game writers had already moved on to the latest and greatest by the time Q1 & 2 were released as GPL.
What made Quake so great was the ease with which it could be customized. Take Team Fortress for instance. A better multiplayer fps has yet to be released. Team Fortress Classic for Half-Life and even Return to Castle Wolfenstein are commercial releases that do very little innovation on a theme developed by people that didn't get paid a cent.
The point being that Quake 1-3 were so open to gameplay modification that the GPL doesn't entice people who are donating free time to making games any more than the games did on initial release. Heck, Quake 3 even releases the same tools id created to make the games to end users. Why do I need to look at the code to combine z-Machine interpreters when Quake 3 already has the hooks for my customizations?
The same thing's going to happen with Doom 3, I'd imagine. Quite simply, all the "opening/freeing" of the Quake code did was make it possible for hardcore programmers to bring it to other platforms. The gameplay talent already have all the tools they needed to move to the next level and more recent games. That's part of the beauty of the way id programs.
It's all 0s and 1s. Or it's not.