Re: So how do Games do it?

Stephen Chenney (schenney@CS.Berkeley.EDU)
Wed, 6 Dec 1995 17:03:46 -0800 (PST)

>Okay, If we are to make VRML an interactive world type thing than why not take
> example from commercial and very effective games? Marathon, Descent, Doom,
> Quake they all have a system of creating a world, putting things in it with
> behavior and they are, best of all, fast. If there was some way of useing a
> rendering engine like in Descent for showing VRML models don't you think that
> might be neat?... Animated textures, interactive "switches" buttons, doors,
> environments, physics, behaviors (monsters), sound ambient sound stereo
> panning, strange light sources. How do they do it?
>
> Just a wondering...
> Alex
> alex_okita@ub.com

Games do it because they are highly specialized environments and they use
very smart algorithms.

Doom and Descent both work in mazes, essentially. Therefore very little
is visible at any given time. If you have smart algorithms for figuring
out what you can see and what you can't, BEFORE rendering it, then you
end up rendering very little and hence very fast. Hint: they don't just
theow everything into a Z-buffer. Do they use one at all?
The worlds are also simple in that they are largely flat with texture
maps to give detail.

VRML places no such restrictions on worlds, and hence pays the price.
There is a big difference between a general modelling language like VRML
and a specialised one like a WAD.

Cheers,
Steve.

PS I don't intend to start a "how does doom do it" tmesad here. But the
initial question as related to VRML is relevent.

  • Next message: Stephen Chenney: "Re: So how do Games do it?"
  • Previous message: Michael Linde: "PHIL: courtesy"
  • In reply to: Alex Okita/UB Networks: "So how do Games do it?"
  • Next in thesad: Stephen Chenney: "Re: So how do Games do it?"