Tutorials

In the previous post, we have seen how to use the Box2D support into the AGE engine. Now I just want to show you how simple it is to add a new framework/engine support.

For example, we are going to add the Nape support.

For those of you who don’t know, Nape is a Haxe/AS3 physics engine. Since Nape is really close to Box2D, it will be simple to understand how box2d has been added to AGE.

The behavior

Since AGE is based on a behaviors system, we are just going to create a new behavior, that we will call NapeMovementBehavior :

class NapeMovementBehavior implements IBehavior
{
    private var _entity : BasicEntity;
    public var enabled(default, null) : Bool;
 
    public function update():Void {}
 
    public function enable()
    {
	enabled = true;
    }
 
    public function disable()
    {
	enabled = false;
    }
 
    public function destroy():Void {}
}

Now we are going to initialize the data needed by Nape for the behavior’s entity :

private var _body : Body;
public static var world : Space;
public function new(pEntity: BasicEntity, pDynamic:Bool)
{
    if(world == null)
        world = new Space( new Vec2(0, 500) );
 
    _body = new Body( (pDynamic ? BodyType.DYNAMIC : BodyType.STATIC), 
                       new Vec2(pEntity.x + pEntity.halfWidth, pEntity.y + pEntity.halfHeight)
                   );
 
    var block:Polygon = new Polygon(Polygon.box(pEntity.width,pEntity.height));
 
    _body.shapes.add(block);
    _body.align();
 
    _body.space = world;
 
    _entity = pEntity;
}

If you need more information on how Nape is working, go check the documentation.
So for now, we have a basic entity initialized for working into Nape.
(more…)

What is the QuadTree ?

If we look at the wikipedia page, we can find a small explanation (it’s the best one I have found so far) :

A quadtree is a tree data structure in which each internal node has exactly four children. Quadtrees are most often used to partition a two dimensional space by recursively subdividing it into four quadrants or regions.

Why would I add it ?

The first approach to detect collisions between entities was to check for each entity if it collided with one of the others entities in the world. So you can use this method for a really small game, because the number of tests is not that big. But if you have, for example, a game with 100 entities moving around the screen, which can collide with each other, the number of tests between entities will be 100×100 !! And at least half (probably more) of them are not necessary.
The Quadtree allows you to eliminate those unnecessary checks.

How does it work ?

Each time an entity is added to the stage, we just have to look into the tree and find where to put it. A node is simply a rectangular zone with entities and other nodes in it.

So basically, any node will be divided in 4 nodes if more than n entities are completely contained within the node. For AGE, I have decided to use n = 2.

Here is a more concrete example :

Now, our tree has been set. Before checking for collisions, we have to query the tree to determine which entities are worth testing for a rectangular zone (the rectangular zone can correspond to the theoretical position of an entity, to know if we can move it or not). So, the quadtree is going to return each entity contained in a node that have an intersection with our request zone (we have to return entities that are not entirely in a sub node too).
(more…)