From this thread:

I'd like to hear more about this… what do you mean here? Who is "you" (players, imms, …) and what kind of building are we talking about? (game content creation vs. creation as gameplay)

The idea was that everything in the game was essentially just a bag of particles. The game was to be soft coded using python. These particles would be the building blocks of -everything-. They'd react to one another differently depending on how they were combined, or if they were just in the same container as another. You'd have a team of staffers who'd be in charge of creating these particles, setting up cause/effect cases for as many as they can think of. These particles would also container slightly more complex activities, such as events. _Anyone_ with the knowledge on how to use these particles would be able to manipulate them (an example was given that if you control the carbon you control the man.) All particles had an origin (starting time) and a duration. The origin could be in the past, present or future. So say someone picked up a particle that'll blow them up in two years. In two years… kaboom! It was kind of a way to implement forms of fortune telling into the game. A step up from these particles where energy. Energy had a controller (which could be a socket connection or an AI script [heat this, move straight at rapid speeds, bend sheet metal] obviously very basic examples, but you get the idea)

Did any of this actually get beyond the conceptual phase? It's just that implementing an entire causal chain sounds like a huge amount of work, not to mention rather hard work at that. Sure, it sounds pretty cool, but the idea of implementing, say, a sword that whacks things sounds pretty hard in this system. Once you have the model, it'd be fairly easy, but getting that model is another question entirely…

It seems like the trick would be to not try to create a real world physics model, but a simple toy model. You could even base it on something like the physics of classical antiquity, with its objects made up of a few elements and so on. That way you can create a creature without having to worry about the millions of parts and interactions it would contain.

The cool thing about this, and I think I remember a discussion on magic systems at MudLab bringing this up, is that many of your game systems would naturally fall out of the model world you've created. I guess the question would be if those systems were what you wanted ;).

Well, I was already assuming that the model was very simplified: the idea of trying to accurately model the physics of biology is so mind-boggling as to be basically ludicrous. :wink:

The problem here is that you need your system to be sufficiently simple as to be actually workable, but you also need to define enough interaction possibilities (i.e. introduce enough complexity) to be able to actually make your world do interesting things. If everything has to be modeled using some kind of generic causality description, things get very complex very fast. This is why I'm hoping to get more detailed information on what was going on here.