still working on this... =[ seriously is there a way to transform your 3d object thats facing the camera into 2D coordinates so the mouse can pick any point if the object is capable of rotating around?
I thought I could do it somehow if I rewrite the code to use the OCD. OCD works real good, but I can't move further from there.

it possible would, but i can't seem to be able to apply it to the 3d Cube Grid   but this looks similar enough to what i need, i will try it and come back with the results.

thanks!!!

thanks antiplastic!!! it actually works. I used boxes in intersections for now and just to keep it simple tried to change fill() of the boxes once the cursor is over them. it really is almost perfect. almost.... at some certain angles the boxes change fill on their own.  i just have two more questions. how would i make buffer boxes all in different colours without having to fill them all separately? and I've tried to run it with spheres just now. That made the program really slow. Is there a way to speed it up?
At the moment my grid mesh is quite rare and I understand that one or two spheres won't make it as slow as spheres on every intersection, but later on the program will have a denser mesh and i think that will make the program near to unusable due to all the spheres in buffer. Cab I increase somehow the dedicated buffer size?

any answers or hints are really welcome

at this stage I got this far:

import processing.opengl.*;

PGraphics bf;
float expo = 1;
float rotx, roty;
float rate = 0.01;

void mouseDragged()
{
 rotx = rotx + (pmouseY - mouseY)*rate;
 roty = roty - (pmouseX - mouseX)*rate;
}

void setup()
{
 size(700,400,OPENGL);
 bf = createGraphics(width,height,P3D);
 stroke(150);
}

void draw()
{
 background(255);
 translate(width/2,height/2,0);
 if (keyPressed && keyCode==UP)
 {
   expo = expo + 0.01;
 }
 if (keyPressed && keyCode==DOWN)
 {
   expo = expo - 0.01;
 }
 rotateX(rotx);
 rotateY(roty);
 scale(expo);
 for (int x = -height/2; x<=height/2; x+=100)
 {
   for (int y = -height/2; y<=height/2; y+=100)
   {
     for (int z = -height/2; z<=height/2; z+=100)
     {
       pushMatrix();
       translate(x,y,z);
       line(x,y,z,-x,y,z);
       line(x,y,z,x,-y,z);
       line(x,y,z,x,y,-z);
       pushMatrix();
       translate(x,y,z);
       color inbf = bf.get(mouseX,mouseY);
       fill(inbf);
       box(10);
       popMatrix();
       popMatrix();
     }
   }
 }
 bf.beginDraw();
 bf.background(0);
 bf.translate(width/2,height/2,0);
 bf.rotateX(rotx);
 bf.rotateY(roty);
 bf.scale(expo);
 for (int x = -height/2; x<=height/2; x+=100)
 {
   for (int y = -height/2; y<=height/2; y+=100)
   {
     for (int z = -height/2; z<=height/2; z+=100)
     {
       bf.pushMatrix();
       bf.translate(x,y,z);
       bf.pushMatrix();
       bf.translate(x,y,z);
       bf.noStroke();
       bf.fill(166);
       bf.box(10);
       bf.popMatrix();
       bf.popMatrix();
     }
   }
 }
}

hey, antiplastic! as I am quite new to processing (just about a month) i could not really understand all the arrays and mapping stuff. sorry could not follow your advice there.. =] though i came up with the idea to use red(), green(), blue() values of the color in the buffer and use them as translate(x,y,z). works really great. (I imagine that this is somehow the long way of arrays and hashmap). I understand that this is a very limited method as there are only 255 possible values for each axis, but for now it works for me.    thanks for all the help, it was your suggestion initially that set me on this approach. I am sure I'll need more help further down the project somewhere, but for now the code is like this:

import processing.opengl.*;

PGraphics bf;
float expo = 1;
float rotx, roty;
float rate = 0.01;

void mouseDragged()
{
 rotx = rotx + (mouseY - pmouseY) * rate;
 roty = roty + (mouseX - pmouseX) * rate;
}

void setup()
{
 size(700,400,OPENGL);
 bf = createGraphics(width,height,P3D);
 stroke(150);
}

void draw()
{
 background(255);
 if (keyPressed && keyCode==UP)
 {
   expo = expo + 0.01;
 }
 if (keyPressed && keyCode==DOWN)
 {
   expo = expo - 0.01;
 }
 translate(350,200,0);
 rotateX(rotx);
 rotateY(roty);
 scale(expo);
 translate(-350,-350,-350);
 for (int x = 150; x<=550; x+=100)
 {
   for (int y = 150; y<=550; y+=100)
   {
     for (int z = 150; z<=550; z+=100)
     {      
       line(150,y,z,x,y,z);
       line(x,150,z,x,y,z);
       line(x,y,150,x,y,z);
       pushMatrix();
       color c1 = bf.get(mouseX,mouseY);
       color c2 = color(0);
       if (c1 > c2)
       {
       translate(red(c1)*10,green(c1)*10,blue(c1)*10);
       fill(0);
       box(10);
       }
       popMatrix();
     }
   }
 }
 bf.beginDraw();
 bf.background(255);
 bf.translate(350,200,0);
 bf.rotateX(rotx);
 bf.rotateY(roty);
 bf.scale(expo);
 bf.translate(-350,-350,-350);
 for (int x = 150; x<=550; x+=100)
 {
   for (int y = 150; y<=550; y+=100)
   {
     for (int z = 150; z<=550; z+=100)
     {      
       bf.line(150,y,z,x,y,z);
       bf.line(x,150,z,x,y,z);
       bf.line(x,y,150,x,y,z);
       bf.pushMatrix();
       bf.translate(x,y,z);
       bf.noStroke();
       bf.fill(x/10,y/10,z/10);
       bf.box(10);
       bf.popMatrix();
     }
   }
 }
}
 

thanks again, antiplastic!

There's a method in processing to handle projections.

http://processing.org/reference/screenX_.html

Note: Personally, I think that this section could be improved if the method would return an array[Floats] or a PVector, otherwise you calculate the projection twice with screenX(x,y,z) and screenY(x,y,z). Here, (x,y,z) are the same.

Suggestion (note, this is not implemented):
PVector 2dPostionOnScreen = screen(x,y,z);

Where PVector is 2D.

And what about screenZ? It was a joke or a prophecy? Maybe Ben and Casey where thinking about the Avatar's 3D movie? Coming soon...  

I think screenZ should be removed from the documentation, isn't it?
http://processing.org/reference/screenZ_.html