Xfennec wrote:
- a real skeletal animation support, and not a morph system like the current MD2 code
Which is the difference?
Xfennec wrote:
- a system with vertex weight support
Ok
Xfennec wrote:
- our own file format, but based on the current TRI format (build another file format now, and only for animations is the wrong path !). It will automatically allows multitexturing, env.maps, shaders, and so on ...
Ok, but i'm afraid that the current tri format file is not compatible with vertices animation... If each line is 1 triangle (i'm not sure about this) then we can not perform vertex movement, vertex weighting, and so.
Xfennec wrote:
- something compatible with Blender
Ok, thenk i should start with a simple blender animation exporter.
Xfennec wrote:
- a RayODE compatible system (skeletals must be able to control RayODE elements, for example, fall back to a regular ragdoll if needed, ...)
Hmmm, here i will need a lot of help, i'm not so comfortable with ODE as you
![Wink ;)](./images/smilies/icon_wink.gif)
Xfennec wrote:
- a plugin for BVH motion capture files ! (thru Blender, why not)
BVH... i have to learn about it. I never heard about it.
Xfennec wrote:
- inverse kinematic ?
Hard... but it's a pretty feature
![Smile :)](./images/smilies/icon_smile.gif)
Xfennec wrote:
This subject is described by many developpers (free, commercial, games, engines) as a real nightmare, so being helped by libs like Cal3D shouldn't be excluded, I think.
Yep. But a few years ago i succes making an directx ms3d importer. That was a real pain, but now i have some experience, and the real hard part is the matrices management (inverses, transposed and so kind of...). Anyway using Cal3d would quit us a loooot of work...Even we could embed cal3d in the code of Raydium, so it won't be a dependency. However, cause cal3d is c++ and is object oriented, i preffer that raydium would not based in cal3d.