Has ioquake3 reached its full potential?

[coffeenet]

Hi,
If ioquake3 had more CPU space, by lowering its computation requirements.
Are there CPU hungry additions that can be done that would benefit from this additional CPU space?
I asked this in somewhere else, and I got moving the graphics away from vertex array to more advanced graphical rendering methods.
But, I was told that it was a VERY difficult task to implements.

Does anyone have any suggestions? (preferably other than graphics)

For example(just an example),
The simplest example would be with more CPU power, you can have more players; thus more possibilities for more advanced and exciting Game-modes.

Another example,
With less CPU requirements, you can port the Game to be a mobile phone Game.

Big or small, any idea is welcomed! I am looking for people to brainstorm this with me!

Any help is appreciated!
Yasir

p.s. Since my understanding is that ioquake3 is virtually the same as Quake3, I took the liberty of asking here. ioquake3 forum doesn't allow me to loggin as well : (

[MKJ]

you might want to ask this question in either Programming or Q3Disc.

[Eraser]

Uh, id Tech 3 is technology from more than 12 years ago. I don't think CPU or GPU power is a limiting factor when running id Tech 3 games anymore. The whole way the engine works is just severely outdated.

[obsidian]

Most of the limits for Quake 3 are imposed by the renderer, which is GPU based. Q3 simply does not take advantage of new GPU features. You could rewrite the entire renderer and network code and everything else as some 3rd party mods have done, but yes that would require a lot of work.

The source code is out there for both Quake3 and IOQuake3 so anyone could branch it and build it into a modern engine if they really wanted to.

What exactly is your goal here? You seem to be asking questions without any real direction.

[xer0s]

Oh look, it's a Quake thread...

[Eraser]

The source code is out there for both Quake3 and IOQuake3 so anyone could branch it and build it into a modern engine if they really wanted to..

You could waste time doing that or do the sensible thing and pick up something like UDK or CryEngine 3 instead.

[plained]

Most of the limits for Quake 3 are imposed by the renderer, which is GPU based. Q3 simply does not take advantage of new GPU features. You could rewrite the entire renderer and network code and everything else as some 3rd party mods have done, but yes that would require a lot of work.

The source code is out there for both Quake3 and IOQuake3 so anyone could branch it and build it into a modern engine if they really wanted to.

What exactly is your goal here? You seem to be asking questions without any real direction.

i got a tingle there 4 a min ey

[bitWISE]

The biggest limit on the Q3 engine for today's world is that it wasn't designed around hardware shaders. After that, I would say the animation system isn't robust enough. I've got a version of it that I wrote from scratch in C# that does deferred lighting and shading using HLSL but without high quality assets it doesn't really provide much of an improvement. The stock assets are missing important channels like spectral maps, normal maps, and height maps. And the levels aren't designed with real-time dynamic lighting in mind.

[coffeenet]

Thanks for the insightful replies! I really appreciate.
I think I understand better now. You guys are stressing that Quake3's graphics rendering is the main hurdle.

Well my goal is the following.
We were experimenting a certain programming technique for enhancing applications, but our focus is Video Games, and we used ioquake3 as our target application.
We were successful in achieving that enhanced performance. And, now we would like implement something feasible that might take advantage of that enhanced performance.
Its for educational purposes mainly. But, if we can achieve something that others may benefit clearly from, we would love to present it that way as well.
Any recommendations aside from graphics, especially since we are using a headless server.

[Whiskey 7]

.............our focus is Video Games.... .

Can't go wrong there for sure.
Best of luck and keep us advised.
There are willing beta testers ripe here for the pickings

[plained]

The biggest limit on the Q3 engine for today's world is that it wasn't designed around hardware shaders. After that, I would say the animation system isn't robust enough. I've got a version of it that I wrote from scratch in C# that does deferred lighting and shading using HLSL but without high quality assets it doesn't really provide much of an improvement. The stock assets are missing important channels like spectral maps, normal maps, and height maps. And the levels aren't designed with real-time dynamic lighting in mind.

boner achieved

[obsidian]

You sure it's not just the painted rock in your pocket?