Page 3 of 11

Re: new Physx

PostPosted: Saturday, 28.June 2014, 08:04
by ikam
it should be a problem with the pause/restart trigger, simul is not runned., i'll check that
for kinematics it's normal it is related to animation script, not to physx, kinematics only collide but notdriven by physics

Re: new Physx

PostPosted: Saturday, 28.June 2014, 08:35
by Skinnytorus
for kinematics it's normal it is related to animation script, not to physx, kinematics only collide but notdriven by physics

Ok for that.

Re: new Physx

PostPosted: Saturday, 28.June 2014, 10:17
by Skinnytorus
Is there a way to make the explosion start from the place of collision?

Re: new Physx

PostPosted: Saturday, 28.June 2014, 15:02
by ikam
I changed the pause detection. Now I'm sure when the time is running or not, so when the physx simulation should run or not.
If you experience again no simulation, it means that the problem is other... let see.

For explosion, I don't know... when it start to collide it chain wakeup all other collider that are in contact too, it will need a kind of delay or trigger but don't know how, I'll study the physx doc maybe there is some kind of feature.

Re: new Physx

PostPosted: Saturday, 28.June 2014, 16:03
by Skinnytorus
If you experience again no simulation, it means that the problem is other... let see.

I'm testing it now. So far so good. Will get back to you.

Take a look at a couple of other issues and another proposal on showing handles directly from physx operator.

Re: new Physx

PostPosted: Saturday, 28.June 2014, 16:58
by ikam
plane, rotation yes, a bug...
for handle it's not possible, because this is not the same render process than when showing from colliders or rigidbody operators.

Re: new Physx

PostPosted: Saturday, 28.June 2014, 17:20
by Skinnytorus
for handle it's not possible, because this is not the same render process than when showing from colliders or rigidbody operators.

Then maybe add this render process to DebugChannel yellow operator then? Without seeing all colliders overall scene setup is a bit of a pain...

Re: new Physx

PostPosted: Saturday, 28.June 2014, 17:29
by ikam
you can put an add (actor) operator to group all acctor ?
It's not so simple to switch to the debugchannel yellow and it's will bring many problems.

Re: new Physx

PostPosted: Saturday, 28.June 2014, 17:39
by Skinnytorus
Add operator? Ah! I see. Thanks for the tip! That's much better. :)
If you experience again no simulation, it means that the problem is other... let see.

Sadly, the problem remains. However the 'no sim' issue happens less often. I don't know what's going on. After playing with collider params in my bugs1.wz4 file one stack works fine, the other just won't start sim after pressing F6. After some time both stacks start working ok... I'm stuck :-k

Re: new Physx

PostPosted: Saturday, 28.June 2014, 18:35
by Skinnytorus
Another interesting problem: How to align a kinematic RB to a spline? Say... I would like to animate a rocket with complex trajectory hitting a wall...
Or maybe this will be implemented in Physx particles?

Re: new Physx

PostPosted: Saturday, 28.June 2014, 19:25
by Skinnytorus
I wonder how to make a hole in the wall without causing the entire wall to explode?
Maybe try to create a ChunkSelect mesh operator + add Selected/Unselected Chunks option in the Debris to apply dynamic RigidBody to selected chunks only?
The rest of the chunks to be static RB by using another RB operator (Unselected option).

Re: new Physx

PostPosted: Monday, 30.June 2014, 16:04
by ikam
Not stable at all yet, but a little physx particles attempt (to try with the physx_particle branch)

Re: new Physx

PostPosted: Monday, 30.June 2014, 18:12
by Skinnytorus
Yey! :) Another step forward! Still buggy, but the particles actually work! I had to show PxCloud first for it to work (otherwise crash is imminent).
Congrats, man! =D>

Re: new Physx

PostPosted: Monday, 30.June 2014, 21:20
by ikam
yes concluding test. it is even faster when enabling the particles gpu mode ;)

Re: new Physx

PostPosted: Tuesday, 01.July 2014, 06:10
by CybeREX
I little test my rebuilded old one scene with static and dynamic geometry, and now I must point, that results more impressive than before new physX.
Before all objects fall when ball hit the building, but now, I find a way, to increase friction power in mesh and overall looks better.
Here is the test scene.