IrrPhysx 0.3 - Nvidia Physx 2.8.1 wrapper
You'll have a choice of capsule or box controllers (though I'm not really sure where box controllers are useful... probably is some situation!)
Stair climbing is something that I'll have to test out actually so thanks for that, hadn't considered it!
EDIT:
Actually stair climbing should work fine really as the capsule controller will walk up and over small boxes, I'll make sure that the necessary parameters are exposed though so you can tweak them for your own setups.
Stair climbing is something that I'll have to test out actually so thanks for that, hadn't considered it!
EDIT:
Actually stair climbing should work fine really as the capsule controller will walk up and over small boxes, I'll make sure that the necessary parameters are exposed though so you can tweak them for your own setups.
It looks awesome... i downloaded and launched it, after 1 s fps slow down from 300 to 3... it speed up to 400fps when there is no fluids... how can i fix it?
Some of errors:
..etc
Some of errors:
Code: Select all
Physx acceleytot hardware not found, simulation wikk be done in software
Code: Select all
IrrPhtysx PERFOMANCE WARNING - could not create fluid in hardware. Creating a software fluid
Code: Select all
Invalid Operaiton: Attempt to create PhysX hardware when physX hardware not accesible
JP, I just can't wait to see the 0.3 version with character controllers!! I'm using your physx wrapper in my present project and I'm really up to see these controller implementedJP wrote: bit of info for everyone waiting for character controllers, seems like i've managed to get them working pretty well, just gotta do some tidying up on the implementation then i may well release a new version with this in.
Personally I think that's the last essential thing in IrrPhysX to be implemented, fluids etc. are just an extra to me
Btw. IrrPhysX 0.2 works great for me! Great job!! Just dunno why I needed to recompile sources to force it to work with my VS2008 compiler.
Wise, wise...JP wrote:that's basically my plan from now is to release a new version as soon as there's a new feature
Well character controlls should be out pretty soon hopefully... I'm having an absolute nightmare with normal actors and setting their rotations each frame... basically if the rotation values aren't within -45 to +45 it just borks up horribly and rotates manically on all axes until it gets back into that range... bizarre and annoying... That's all that's holding me back from the next release really so I'm gonna hopefully get a bash at it tomorrow on my day off from work and if I can't get it working properly I'll probably just take out the rotation stuff for now and release again and get it fixed in the future (hopefully!), it's not an essential part of the next release so that should be fine.
Malgodur, there's no error really. You computer does not have hardware available that will accelerate Physx (i.e. a Physx PPU card or a recent GeForce GFX card which supports Physx acceleration). So that means all the calculations are done on the CPU, which is generally fine but it's not ok for fluids as they need a lot of calculation time so will kill the performance on a system without hardware acceleration (sio2 explains this in his post).
Having said that the last bit you post about the invalid operation is something to be slightly concerned about in a way, possibly an oversight by sio2, not much of a problem though as no one can use the fluid stuff as the source isn't available!
Malgodur, there's no error really. You computer does not have hardware available that will accelerate Physx (i.e. a Physx PPU card or a recent GeForce GFX card which supports Physx acceleration). So that means all the calculations are done on the CPU, which is generally fine but it's not ok for fluids as they need a lot of calculation time so will kill the performance on a system without hardware acceleration (sio2 explains this in his post).
Having said that the last bit you post about the invalid operation is something to be slightly concerned about in a way, possibly an oversight by sio2, not much of a problem though as no one can use the fluid stuff as the source isn't available!
You need to install the necessary Physx SDK and system software as described in the proper IrrPhysx thread, did you not read that? If you have done that and it's still not working then possibly you've got a different path to Sio2 so check the linker settings and make sure the paths are correct for where you installed Physx.
I just fixed it, there was wrong link in vcproj.....
Omg, old game example still crash after loading flagpole_physx.ms3d,
new gameexample compiled with removed fluids crash with irrlicht.dll bug..
Please help me...
Omg, old game example still crash after loading flagpole_physx.ms3d,
new gameexample compiled with removed fluids crash with irrlicht.dll bug..
Code: Select all
?createDevice@irr@@YAPAIrrlichtDevice@1@W4E_DRIVER_TYPE@video@1@ABV?$dimension2d@H@core@1@I_N22PAVIEventReceiver@1@@Z
-
- Posts: 168
- Joined: Sun Feb 04, 2007 3:30 pm
- Location: France
I had the same problems when i used Physx in my (non-irrlicht) game last year - I couldn't set the correct angles for physx-actors because PhysX always gave weird results.
I tried with eular rotations, matrices and also with quaternions, but the results were always the same.
Apparantly physx does some of its own internal calculations to determine orientation. Now that i think about it, maybe you could try relative (object space) rotations?
Have you asked on NVidia's forum yet?
I tried with eular rotations, matrices and also with quaternions, but the results were always the same.
Apparantly physx does some of its own internal calculations to determine orientation. Now that i think about it, maybe you could try relative (object space) rotations?
Have you asked on NVidia's forum yet?
where to download irrphysx V0.3?help meᦀ
where to download irrphysx V0.3?help me!
who can give me a copy?thanks!
my email:flyrxue@163.com
thanks again!
who can give me a copy?thanks!
my email:flyrxue@163.com
thanks again!