Some interesting work at Opensim -- they call it "Ninja physics" -- basically, supporting jointed assemblies of prims. This is an important first step towards more algorithmically controlled robots and NPC's. Here's some video:
http://video.google.com/videoplay?docid=-1419364493373489348 Just a swing, but the implications are pretty serious. Add motors and servos to the joints, and you have something approaching a Gazebo-style robotics simulator. What this means is, you can create vehicles and legged robots that operate using realistic physical principles, rather than typical video game quasi-animated fakery. I can enumerate all the reasons this matters if anyone wants, but the short answer is it's important because a shared virtual world can't be faked like a videogame; it needs internal consistency, because of the open-ended nature of things. It needs to be more like a simulation, and less like animation. As to how this relates to a viewer: let's say I want to simulate a dog this way. Right now, I can use prims and textures, but I have no way to add anything like flesh, fur, clothing or hair to these objects. The joints can be done server-side, since the client faithfully represents the position and rotation of each prim. Simulating stuff that requires deformable meshes, or other complex graphics, will require client-side cooperation, and additions to the protocol. These issues need to be thought out in some detail, even if we're not going to implement everything immediately. The most important thing to get right initially is to not bind ourselves to the limitations of the SL protocol. Obviously it's the protocol we will support first, but the architecture must be flexible enough to make it feasible to support something much more sophisticated down the line. In any case, it's kinda cool. -danx0r --~--~---------~--~----~------------~-------~--~----~ this list: http://groups.google.com/group/realxtend realXtend home page: http://www.realxtend.org/ -~----------~----~----~----~------~----~------~--~---
