whisperstorm
09-03-2006, 03:47 PM
Warning, this is just idle chatter about an idea I've been kicking around. If you want runnable code please jump to the next posting... :)
The idea is this. Many folks create games and get them fairly far along and simply stop working on them. They might even get 90% of the way and give up.
Imagine if you had a game building tool which allowed you to develop a game that is "always" finished? As a matter of fact it's "finished" as soon as you begin to work on it! Hear me out.
I've been reading about the "Snowflake Method for Writing a Novel (http://www.rsingermanson.com/html/the_snowflake.html)" Where you start with a very basic statement of your novel and you make iterations over that where the story gets more and more complex and fleshed out until its a full novel.
Imagine if you could apply this technique to game building. You start off with a framework of a game that has all of the "basics" of a game. A player icon, a score, lives, background, etc. If you just hit "new game" then chose "play" you'd have a very basic, boring, bland game, but it would be "complete".
Using some kind of refinement strategy you make multple passes through this game -- adding an enemy here, a menu there, some weapons, or inventory, etc etc. The idea being that no matter what you add to the game, you are always in a state where the game would run from start to finish and be "complete"
I know that many folks will be saying things like "well, wont all the games be hopelessly homogeneous and boring?" or "wouldnt this severely limit what you could create?" And I agree with that. I think the genious of the game tool would be to find a way to allow a game to evolve into say an rpg, or shmup, or match - 3 or whatever you want using some kind of iterative technique which modifies the existing game "world" but always keeps the game complete.
Perhaps the games that would be created with this tool might not be considered "pure" games but more like "generative" or "evolved" games. The IDE would enforce many limitations on what objects to add and interfaces to display, but would be flexible enough to allow for someone with persistence to continue to work at their game until it was sufficiently like thier vision that they felt it was ready for release.
I dunno. Just an idea - but I hear about folks who create very involved engines, or get 1/2 through some platformer they created with great ideas, only to give out when it was no longer fun to work at the game. If the tool itself was "the game" where actually working at and evolving your game idea was the fun part (for the developer) and that at the end of the day you could always say you had a "complete" game.
Another way to look at it might be from a modder perspective. You start out with a framework that meets some genre of game. You continually mod it until the graphics, sounds, gameplay, etc. meet the ideas you had. You dont spend weeks or months building an engine or figuring out collision detection, you instead build a game more like an author writes a story.
A tool like this already exists in a way for folks who write interactive fiction: Inform7 (http://www.inform-fiction.org/I7/Inform%207.html). No matter what you do to your "plot" you always have a runnable game.
Is this just completely foolish wishful thinking? I see games like Spore offering something of what I mean. You are given all the tools you need "in game" and you spend your time on the fun parts of making a game (creature) and let the engine deal with the complicated behind the scenes stuff.
Another aspect of Inform7 that is intriguing is the plugins ability. If someone else creates some car object, you can simply import it and it works well within the existing framework. My idea for a generative / evolutionary game engine would follow the same concept. You might spend alot of time working on a really great "starship" object. Instead of me having to build one myself I might go online through the ide and discover you have created this starship. Given the restrictions of the game tool I would be guaranteed that it would "work" in my own game given a common set of interfaces.
This way, folks who just like writing cool algorithms can help others who just want to create a fun shmup.
Anyway, just a pie in the sky idea, but with things like Spore, Inform7, Second Life, and others, it seems like having a tool like this is not an impossible dream.
The idea is this. Many folks create games and get them fairly far along and simply stop working on them. They might even get 90% of the way and give up.
Imagine if you had a game building tool which allowed you to develop a game that is "always" finished? As a matter of fact it's "finished" as soon as you begin to work on it! Hear me out.
I've been reading about the "Snowflake Method for Writing a Novel (http://www.rsingermanson.com/html/the_snowflake.html)" Where you start with a very basic statement of your novel and you make iterations over that where the story gets more and more complex and fleshed out until its a full novel.
Imagine if you could apply this technique to game building. You start off with a framework of a game that has all of the "basics" of a game. A player icon, a score, lives, background, etc. If you just hit "new game" then chose "play" you'd have a very basic, boring, bland game, but it would be "complete".
Using some kind of refinement strategy you make multple passes through this game -- adding an enemy here, a menu there, some weapons, or inventory, etc etc. The idea being that no matter what you add to the game, you are always in a state where the game would run from start to finish and be "complete"
I know that many folks will be saying things like "well, wont all the games be hopelessly homogeneous and boring?" or "wouldnt this severely limit what you could create?" And I agree with that. I think the genious of the game tool would be to find a way to allow a game to evolve into say an rpg, or shmup, or match - 3 or whatever you want using some kind of iterative technique which modifies the existing game "world" but always keeps the game complete.
Perhaps the games that would be created with this tool might not be considered "pure" games but more like "generative" or "evolved" games. The IDE would enforce many limitations on what objects to add and interfaces to display, but would be flexible enough to allow for someone with persistence to continue to work at their game until it was sufficiently like thier vision that they felt it was ready for release.
I dunno. Just an idea - but I hear about folks who create very involved engines, or get 1/2 through some platformer they created with great ideas, only to give out when it was no longer fun to work at the game. If the tool itself was "the game" where actually working at and evolving your game idea was the fun part (for the developer) and that at the end of the day you could always say you had a "complete" game.
Another way to look at it might be from a modder perspective. You start out with a framework that meets some genre of game. You continually mod it until the graphics, sounds, gameplay, etc. meet the ideas you had. You dont spend weeks or months building an engine or figuring out collision detection, you instead build a game more like an author writes a story.
A tool like this already exists in a way for folks who write interactive fiction: Inform7 (http://www.inform-fiction.org/I7/Inform%207.html). No matter what you do to your "plot" you always have a runnable game.
Is this just completely foolish wishful thinking? I see games like Spore offering something of what I mean. You are given all the tools you need "in game" and you spend your time on the fun parts of making a game (creature) and let the engine deal with the complicated behind the scenes stuff.
Another aspect of Inform7 that is intriguing is the plugins ability. If someone else creates some car object, you can simply import it and it works well within the existing framework. My idea for a generative / evolutionary game engine would follow the same concept. You might spend alot of time working on a really great "starship" object. Instead of me having to build one myself I might go online through the ide and discover you have created this starship. Given the restrictions of the game tool I would be guaranteed that it would "work" in my own game given a common set of interfaces.
This way, folks who just like writing cool algorithms can help others who just want to create a fun shmup.
Anyway, just a pie in the sky idea, but with things like Spore, Inform7, Second Life, and others, it seems like having a tool like this is not an impossible dream.