using Blender to make collision maps

I’m finally getting my Blender made terrains/maps to load into the Ralph sample, now that I know how to edit my .egg files to include Grid {polyset keep descend} to the Grid group.

I started with just making one shape…a grid. Then I added some spheres and cubes that were half above-half below the grid. And I noticed that Ralph could walk through them. I had to add Cube {polyset keep descend} to these groups as well to get Ralph to collide into them. I guess the ray can see through objects that aren’t tagged.

Its kind of funny. The visible objects to me are “invisible” to Ralph while the invisible objects (collision meshes) to me are “visible” to him.

The collision detection and the visualization are really 2 different parts. So you have models you can see and models you can collide with. Often you would want the visible models to be nicely detailed, while collision models are less detailed. This is because it takes less cpu time for the collision detection if the collision models are simpler. Note that by using “keep” the way you did, a model is both visible and collides.