i want to code a Collision Detection, the Theory are not my Problem and i allready write a function, but i think its worst as a "bad solution" so i want to ask you how this thinks solved professional.
My way. I use a 3Ds/Obj importer from a demo. in GLModel are a function which renders the Tris.
So i use a modif. Version of this. I filter the Vertexpoint with the biggest distance to PointZero out (for each Side)
Than i filter again from this 24 Vertexpoints the biggest x,y,z and -x,-y,-z
than i say x side is x+(-1)*(-x) (for y,z the same)
So now, i know every side of my Boundary Box.
It works most well (in complex Models (4000++ poly) it don't want to work correct don't know...)
ok, i hope you can help me. Thanks.
P.S.: Sorry about my English, it's not as well as it should...