Originally Posted by super pretendo
Not how it would work braj, all theyd have to do is call client-side models etc with a few bits of data that are just precise enough for tactical consideration
|
That solves the problem with video streaming, granted. But now you're heaping more onto the client CPU and GPU. In effect, asking their system to run a real time "Demo Reel" on top the of processing what is happening in the players view.
To accomplish that with any fidelity you'd have to scale the render down so far it would be difficult to even tell what was being shown in the "stream". An example of this would be to look at any game that gives you a second on-screen view - for example driving simulators with active rear view mirrors. You will notice that they NEVER give you the same draw distance or polygon rendering detail that you get from the forward view. For example look here:
As you can see, the rear view mirrors and hood reflections pop-in distances are dramaticly reduced. And I can tell you from experience that this is running on one of the highest detail settings on a rather nice setup (judging by framerates).
Now your expecting PS2 to do the exact same thing, except with a high level of detail and high fidelity from a remote camera controlled by another player? And you're still asking the servers to stream data reguarding the observers' camera positions and lists of visible objects. You're asking a lot. Too much I'd recon. And if you are not expecting the observer to supply a list of visible objects, then that means that all clients would then need access to everyother players postitions at all times "just in case" a live stream is started at any time, and that seems EXTREMELY exploitable.
Also, please do not refer to me as "bra" unless I am either doing keg-stands in your frat house or lifting and seperating your breats while giving you all-day comfort.