The best java wrapper for the best physic engine!

It CAN be achieved, we both know that's possible, it's just whether anyone has the time to go through and tweak the wrapper for every implementation to make them work similarly. If there is a stable jME-Physics wrapper around PhysX ever I'll probably take the time to make sure both ODE and PhysX implementations work the same as I want to be able to offer both (because of licensing issues this is necessary since not everyone can use PhysX).

So you want to tell us, what ever can be done with PhysX, you want to do it with ODE?   Cool! 8) . . . while you're at it, can you fix those trimesh-trimesh collisions in OPCODE (it's ok if they just work like in PhysX :P)



(hmm, I can't believe I'm diverting this thread with darkfrog, but I could not resist)

hehe…often "tweaking" in an abstraction must equate to dumbing down functionality to make it all work the same. :wink:

Ageia confirmed, that I can distribute JPhysX with whatever license I want, untill I do not redristributing their SDK. So, to use this adapter you need to agree with Ageia's license about SDK, or just about Engine (in the latter case registration is not required).



So, in short time I will establish open source on source forge or java net (any preferences?) or other.



P.S. Linux SDK already available, so jphysx will be available for this platform soon too.

Great news!



I'd prefer sourceforge as they have subversion (instead of cvs) :slight_smile:

P.S. Linux SDK already available, so jphysx will be available for this platform soon too.


Really looking forward to.
darkfrog said:

It CAN be achieved, we both know that's possible, it's just whether anyone has the time to go through and tweak the wrapper for every implementation to make them work similarly. If there is a stable jME-Physics wrapper around PhysX ever I'll probably take the time to make sure both ODE and PhysX implementations work the same as I want to be able to offer both (because of licensing issues this is necessary since not everyone can use PhysX).

IF you want two impl to act the same (same gameplay) then really whats the reason to have more than one ? Why would you tweak the PhysX impl to act similar as ODE ?
It would not even be possible because there's no way you could make PhysX's collision detection be as bad as ODE's, even if you tried hard. There's no way you can make it explode constantly etc..And if you really found a way what would the point be ?
Lets just face it's just not possible and certainly not worth the trouble.
To do it you would either downscale one impl or the gameplay itself..and I don't see why anyone would do that just to make "nice" looking code (the users of your game dont care how your code is). I totally agree that Physics should be abstracted but with the sole purpose that it is easy to replace later with another impl if something better comes up.
What licensing issues prevent everyone from using PhysX ?
Baune said:

What licensing issues prevent everyone from using PhysX ?

After Uree last message: none :D

Baune said:

There's no way you can make it explode constantly etc..And if you really found a way what would the point be ?

That's about what I meant. I think the abstraction is important for the game devs to have a common API (plus easy learning!) and they can switch between engines in the testing phase. The fine tuning must be done engine specific.

I vote for a merge into jme-physics.

it doesn't need to be switchable during runtime, but a phys-x wrapper (eg - PhysicsSpace) would be awesome… can switch implementations before game load (can make it selectable - easier to just recompile) and compare behaviours.



heh - perhaps then things wouldn't keep getting stuck inside of things (with almost no code change)



for things like joints, well, integration isn't instant… throw unimplemented exceptions (till they are implemented)



this is all from the point of view of someone who uses the physics libs and isn't interested in developing them (much), just using them… but good luck.

Have they changed the license?



It used to be free for non-commercial games but you ended up paying a license fee if you decided to start selling the game.

From what I understand it's free as long as you tell the players that you're using PhysX and tell them how to buy the Aegis physics card.

I would like to buy a card myself but don't know if it fits with my motherboard.

It is a cut from Ageia's EULA:



Free:

Commercial & non-commercial use on PC

Must keep registration information currect

Must agree to the EULA at the time of download (pops up, but is copied below)

Available for Windows & Linux (soon)

No PhysX HW support requirement

PS3 platform (through Sony pre-purchase)

All platforms through some of our middleware partnerships, such as UE3, Gamebryo 2.2, and others



$50k per platform:

Xbox 360

Fee may be waived at our discretion for multi-platform developers providing PC HW support

Fee may be waived at our discretion for some Tools & Middleware providers

Awesome!



I personally would like to see this as the primary supported implementation for jME-Physics in the future then. :slight_smile:

but but what about MacOS?



not that i really care, i don't have one  }:-@

They can use ODE. :wink:

Or make Aegis create a Mac OS X release of it. I would assume it would be extremely easy for them once they have the Linux ones.  :smiley: Much rather have OSX/Linux than Win/Linux.  :evil:



P.S. BTW, not owning a platform is not excuse enough to ignore it  ;)… I don’t own a Windows computer, and still know I might have to consider them as potential targets :’(

What about the Amiga platform ?

Baune said:

What about the Amiga platform ?

:) ... or the Game Boy?

Or the Commodore 64  :smiley:

duenez said:

Or the Commodore 64  :D

:'( I'm gonna cry now pls bring back the  64 and lets all forget about this PC nonsense. I mhz and 8 sprites is really all you need.