WOW, a new way to program large games quickly

[TheDarkJay]

this is quite a break through for "Proffesional" Programmers, those working on games for the Next-Gen consoles are most likely gonna have to use this method a lot.

[CrazyAlabamianLlama]

i have an idea on a game thing like spore...if anyone would like to help i am gonna try to make a 2d version of it with QB and probably VB6 also...if anyone is interested in it please email me CraZyJuGgaLo3088@aol.com

[Rattrapmax6]

:???: If they are doing it, then it can be done... how, I'd have to ponder on that for awile.. :roll: :wink:

[marzecTM]

darkjay, beside the fact that this is really incredible the math part might not be too complicated. from my humble point of view the game is seperated in different parts each having a basic mechanism or better goal. the connections between those parts are the creatures/buildings/vehicles from the last stage. granted the universe UFO thing is incredible but with a procedural approach possible ( there's an article series on gamedev describing how to create a procedural universe afair).

on to the maths. the animation part is probably done with inverse kinematics. this type of animation allows you to automatically generated positions of different parts of your model depending on the state of one of the parts. so for example if you lift a foot to some position your inverse kinematics driven character will automagically lift the leg and pevils accordingly ( given that you limit the movement range of the different parts according to known anatomy ).

the really hard part imo is the calculation of this movement limits. based on what mankind knows about anatomy and physics one can derive basic rules to calculate this limits. with enough abstraction you get a nice small rule set that is universally applicable to a large set of body parts ( like: what would a three feeded thing walk like. in what range can his feed operate etc. ). also the learning process of what limits combination is the most efficient for the creature to move is a hard one, but again with a nice set of limiter rules ( yeah my word invention sorry my english is bad ) this is not an impossible task.

to strap things down: you don't need a degree in rocket science to do something like this. check out articles on inverse kinematics and procedural animation creation ( i remember having seen such an article on gamasutra not sure though )

uhm, my 2.4332 cents

[Atom Ant]

That is absolutely amazing. I see a new genre on the horizon...

[marzecTM]

just revisited that and stumbled over the title of this thread. it is NOT a new way of coding a game faster. it will eat some jobs of some artists but it will not improve the speed of actually coding and designing the game ( actually it's exactly the oposite of it ). also this is not a new approach ( procedural content creation was used in a lot of early days games, think of random level generators etc. ) nor a new genre in fact it's a mixture of a few genres, that is even cleanly seperated in 5 stages so it might not be as superb and all new as the creators state. still a nice idea.