Hey guys,
i am trying to create this effect in a more suitable way since this is using physics and you all know how annoying that can be. Exspecially since i later want to move points along a curve depending on the position of the following cubes. - that gets really buggy
If you know any tutorials that are about stuff like that let me know 
I really have to start getting into geo nodes man, that would make so many things faster.
Cheers
Ah i see thanks a lot! I dont really get why the position multiplied with the distance only moves along the normal but i take it ^^. If i may be so bold to ask how would you go further in implementing this effect:
This is my Setup from the Side. Ultimately i want a circular array out of 8 or more balls that move along a curve towards the " goal empty".
But this is already so much help thank you.
The big thing that you’re showing is inertia and acceleration. Geometry nodes do not know the state of the file last frame, so you can’t get those with GN.
If you want to use rigid body physics to do stuff without worrying about physics later, you might think about baking the physics to keyframes.
tricky.blend (901.8 KB)
we can get acceleration using exponential math @bandages
if you want true acceleration, we can set the initial position to the derived positions in upbge as well as encode a velocity to a meshes vertex positions to recall next frame
Those aren’t accelerating. They are changing position based on the position of the “force field” in a non-linear fashion, but they’re not accelerating, they’re just jumping to that position. Their position is solely a function of the control, rather than a function of the control and their old position.
That’s not to say that it’s not what Geospinelsulfid wants. All we know so far is, it’s like physics, but without physics. Maybe they don’t care about inertia.
We can recreate limited rigid body physics in a Blender fork, or we can recreate limited rigid body physics in modern Blender with some scripts, but either way, we’re doing a lot of work to recreate functionality that already exists, and that won’t be any easier to use than regular old physics.
I think what Geospinelsulfid ought to do here is say why they find Blender physics annoying. Some annoyances may be solvable with a different workflow. Some problems they’re having might be be solvable with a different hierarchy of action. Some problems they’re having are going to show up in whatever non-physical workarounds people can figure out anyways.
Damn witchcraft dude that looks awesome ty. While it does look amazing, bringing upbge into the mix is too much for my small brain since im already out after ur math setup. But fun to know what is possible. This gives me a great base to work on thanks.
True thats not “real” acceleration but i dont care if the appearance is the same. Since the goal was anyway to create a setup to play around with these variables and see what looks best.
Altough i dont understand half of it i still would argue that creating it with geo nodes is way faster, versatile and safe than to try this purely with physics since they really can get messy fast. I am amazed all the time what complex behaviours are possible with the physics system but i was already giving up creating that circular array and copying constraints around in a changing positive and negative space which is why i came here
thanks for all the input
