Udo's Techblog

Where to Go?


So after thinking about this whole Dynamic Premise thing for a while, it's time to restate what we're trying to say here. The general idea is that it should be possible to bridge the devide between "game" and "simulation" by introducing the concept that a certain type of simulation can provide content and fun for a "game". While this method (if we can make it work) would theoretically obsolete the instancing of game worlds, this isn't the main reason for trying in my opinion. The most compelling aspect is that when you can create compelling content via simulation, it means that you create a dynamic world that is much more consistent and interesting than the carefully designed MMORPGs that we have today. It means that we create a world where the single player can and will make a difference just by influencing the simulation. If we can make this work, it would also solve the perceived "threadmill" effects that most games have ('been there, done that').





Okay, so let's talk about some parameters and design decisions that go into a MMORPG, and which ones would fit best for the implementation of the Dynamic Premise. First, let's cover all the aspects of the world that need to be simulated.





Aspects of Simulation


Before anything else, we'll need to define the aspects and layers of the simulation. This is going to be our set of goals and requirements against which we can later measure our implementation ideas.





Terrain


The terrain and inanimate environment of the world needs to be adequately simulated for the game to be effective. First, terrain needs to be generated. This is how the planet(s) of our simulated world are built. Let's have a look a different terrain models.





Height Maps


The simplest form to represent a terrain is by producing a height map. Because this is essentially a flat map with altitude data in fixed spacial intervals, a simple height map cannot be used to describe real 3 dimensional geographical information like, for example, a cave under the surface. The big benefit of height maps is their computational simplicity. It is very simple, to implement physics effects like gravity and game algorithms like realistic path finding. Height maps can also be used to simulate vegetation and growth patterns very easily.





Meshes


The most complex form to represent terrain is essentially to treat the whole body of the earth like a giant 3D object that is defined as a set interconnected points and surfaces in space. Because it can have any shape or form, a mesh can be used to easily build structures like caves, buildings, and multi-layered terrain. On the downside, meshes carry a certain computational complexity and do not lend themselves well to algorithmic interpretation. In comparison to height maps, mesh environments are difficult to work with, especially when other modules of our world have to interpret them extensively.





Next, we'll look at how terrain data is created and manipulated, to better judge what model best fits our needs.