http://www.jphysx.com
The best java wrapper for the best physic engine!
JPhysX test release. Java wrapper for Ageia PhysX (http://www.ageia.com)
Simple demo buit on jME + JPhysX is available.
Wrapper contain: scenes, actors, shapes, joints, fluid, cloth, materials, and other…
Not included yet: callback interfaces.
Plans: import java docs and add "short hand" functions
P.S. dont forget to install PhysX engine before using!
Picking up speed? 
What about writing a jME Physics 2 implementation with that, instead of coding demos? 
You could use all demos from jME Physics 2 then - and compare it to the other implementations (ODEJava and JOODE).
Thats nice will you release the source ?
Sources is not a problem. But they are not very helpful, because of (almost) autogenerating nature.
Though the license for it would be important. BSD-like would be very nice, esp. for jME users.
Yeah ok I'm just interrested in seing how swig does it…
Any word on the license? What about contributing, do you plan to host it e.g. on sourceforge or java.net?
Complete sources including build instructions would be important to build platform libraries on ones own, too…
License: some type of BSD license (because PhysX uses one). I will post it later (on the www.jphysx.com).
I dont know much about contributing and hosting on sourceforge and java.net. What benefits does them has?
Uree said:
License: some type of BSD license (because PhysX uses one).
err, no - PhysX is closed source and certainly is not under BSD-like license.
Uree said:
I will post it later (on the www.jphysx.com).
ok, nice, we'll have a look
Uree said:
I dont know much about contributing and hosting on sourceforge and java.net. What benefits does them has?
More than one person can access the project (Subversion!) and you can grant write access to some people, who wish to help out...
Ok. I am not see much of jphysx interested people, or some activity, or even feedbacks. So, I think, it is not worth of such advances for now.
I am still working on it (and will do so) because I need it for my projects.
Maybe some new demos will fix the situation?
There have been quite a few people mentioning this on the jME forum. Perhaps the lack of direct interest is due to the lack of open hosting (like SourceForge, Google Code, or java.net) and the lack of a defined license.
A couple of people have even expressed an interest in implementing it in jME-Physics.
If jme-physics turns out to be "the best jme integration of the best java wrapper for the best physics engine" then i can imagine MUCH interest … 
Hello Uree,
You did a great job creating this wrapper.
Irrisor also did a great job creating an abstract physics layer.
IMO we should join both technologies, allowing Jme-physics to run with a JPhysX implementation.
What are your intentions to JPhysX?
I'm sure people will get much more interest as this project grows. But it must be open source, otherwise we won't see much interest.
If you don't have time, or think it would be a boring task, I think someone here could do that for you.
Regards
Very cool!! I think it’d be a great idea to integrate this into irrisor’s jme-physics project.
Are there any capabilities of Aegia’s PhysX that aren’t provided with ODE, I’m just thinking how difficult would it be to expose implementation specific features in the jmephysics wrapper?
dougnukem said:
Are there any capabilities of Aegia's PhysX that aren't provided with ODE,
hehe, lots and lots I guess ;)
I'm just thinking how difficult would it be to expose implementation specific features in the jmephysics wrapper?
It's not difficult to expose them, but it's not 'nice' :) (It's currently possible for ODE, though ;))
What Irrisor refers to as "nice" means that though you can expose things that are implementation specific you then just broke your abstraction layer. If you never rely on anything "implementation specific" you could go so far as to provide an in-game option to choose between JOODE, ODE, and PhysX and your code would work with them all indifferent about which implementation is being used.
True.
The only sad thing about this, of course, is that it only 'works', but will not be fine tuned. So in a finished game you would not be able to exchange physics implementation without harming game play unless you optimized it for all engines…
I would argue that it's the wrapper's job to provide said "tuning" so under jME-Physics all of them are tuned to work similarly in function.
hehe, that'd be nice but can never be achieved IMO (please prove me wrong :P)