In article <483511e3@news.povray.org>, war### [at] tagpovrayorg says...
>   http://en.wikipedia.org/wiki/Collision_detection
> 
Hmm. There is no "collision" between the mouse pointer and the object, 
save in the most basic sense that something like the POV-Ray camera can 
be said to "collide" with an object via its vector. The issue here is 
detecting one such collision, based on another. I.e., you can't "test" 
the bounding box *unless* you already know that the pointer made 
apparent contact with the object, and then you have to determine *which* 
face of the bounding box that "had to be" for it to have struck it. Mind 
you, I can see where its possible they are already using a bounding 
system to limit how many tests they have to make, and therefor they may 
be testing if something "entered" the boundary, then if it actually 
struck the object. That would make things a whole heck of a lot simpler, 
but since this is all done server end, for security reasons, I don't 
have access to source to check...

I suppose the wiki is applicable only if they are doing something "like 
that" already. If not, then I suppose a second test against the bounding 
box created would work. Frankly, I am not real certain that they "are" 
using something to limit how many tests they need. Their code often runs 
slow, there asset services, which track who owns things, goes down way 
to often, etc. Even *basic* stuff in the script system seems to be 
missing, since it never occurred to them that it made sense to have it.

Example - Every person/group has a unique ID. The "test" function to 
tell who touched an object only tests "one" ID at a time, and you can 
only have one such script in each object, which means you have to use 
multiple objects, each with a seperate test, to determine if one of of a 
list of people/groups touched it. Some people get around this using HTTP 
requests and external servers to handle the tests, but imho, its damn 
stupid that they waste time and resources to run multiple scripts, in 
multiple objects, which "still" have to be sent to the client, along 
with textures and other data, to do what a simple "Is the ID if the 
person that touched it in a *list* of IDs?" I mean, this is a complete 
"duh!", but its missing.

So, I am not betting that they necessarily are using bounding, or other 
tests effectively to determine what got touched, client side or server 
side. :(

Still, I think I can probably post the idea anyway, since its possible 
they do have something in place that could be adapted to allow it.

-- 
void main () {

    if version = "Vista" {
      call slow_by_half();
      call DRM_everything();
    }
    call functional_code();
  }
  else
    call crash_windows();
}

<A HREF='http://www.daz3d.com/index.php?refid=16130551'>Get 3D Models,
 
3D Content, and 3D Software at DAZ3D!</A>

Post a reply to this message