I tried experiment in Unity. I made object rotate and attached a child to it, so child was rotating around, because parent did. Then I tried to detach child and see if it’s gonna fly away by centrifugal force. But it just stopped. That was just example, the essence is that it is hard to do some simulations in Unity - there always problems, everything happens not the way it’s expected. Is jMonkeys physics the same or it is better for simulations ?
Sounds like you didn’t really test physics in Unity… you just assumed that all nodes are physical objects… which is not the case.
You’ll have the same issue in JME … unless you do spend the time learning how the physics engine is working…
We use Bullet physics and there is a documentation page about it. https://jmonkeyengine.github.io/wiki/jme3/beginner/hello_physics.html
Though I recommend reading about the whole aspects of the engine https://jmonkeyengine.github.io/wiki/
The child also had rigid body and restriction to vertical axis, so it’s not gonna fall down. Thanks.
scene graph
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
rigid bodies.
See how far separated they are? That’s because they have nothing to do with each other. The scene graph parent/child relationship will not affect the physics AT ALL.
If you want two things to be connected then separated later then there must be some physical joint holding them together.
1 Like
In unity don’t parent them, use joints. Same will apply in jme if you use bullet.
1 Like
Got it. Thanks people.