[MUD-Dev] Cg - Lowering the cost of 3D games?
frankc at maddog.com
Tue Jun 18 22:59:16 New Zealand Standard Time 2002
From: <Daniel.Harman at barclayscapital.com>
> Its not just about the APIs intrinsic value, its a shift in
> thinking to a higher level approach and trying to make 3D easier
> rather than just prettier. Resource management and collision
> shouldn't have to be something that gets re-implemented for every
> project. Of course whether a gfx chipset vendor like nvidia will
> drive it that way is another matter. As others have pointed out,
> they need this API so that developers will start using their
> card's advanced features and encourage the consumer to keep
> upgrading. Whilst nvidia like to push out new generations of
> cards every 6 months, I don't know anyone who feels compelled to
> buy each iteration.
> Right now writing 3D just seems harder than necessary
The place to start is just to build a core 3D client-server. This
means to pick a 3D engine, use a simple network protocol, and use or
build a simple network library. You also include some sort of
object management system -- anything from MySql to an object system
like in Cold, or something else. Doesn't matter-- build one version
with bought, borrowed, stolen, revamped components and get it
working. Then improve and throw away the old stuff.
There has to be enough in the core where a person could built a
mud-type game even if the client can't handle 3D and audio. The
server at least will act like the client can. Isn't that the
philosophy behind Worldforge?
The hardest part on the server side (beside scene management) is
navigation. Mob creation, placement, modification is not much
harder than in a text mud. Combat system is similar to a text base
except now you are dealing with 3D space and movement.
Before long, if the core is decent enough, there will be at least a
half dozen competing projects all trying to improve the core.
MUD-Dev mailing list
MUD-Dev at kanga.nu
More information about the MUD-Dev