Gravity Troubles

[lucas]

Hi,
I'm experimenting with gravity, where instead of a direction, objects fall to a point, which lies in the middle of a static circle.
In practice, it is like a planet.
However, I'm having a little trouble moving a character around. In order to move, I force the character (a circle) in a direction perpendicular to the angle towards the planet.
The problem arises when the character floats off the planet. I assume it is because it still has the velocity of the last few seconds, when the angle toward the planet was different.
Its kinda hard to explain, so here is a video.

Any solutions? How could 'translate' the velocity to the new perspective?

Lucas

[slembcke]

To get a correct clockwise/counterclockwis vector, normalize then get the perpendicular vector of the delta from the position of the object to the gravity point.

Something like this:

Code: Select all

cpVect direction = cpvperp(cpvnormalize(cpvsub(body.p, gravity_point)));

Then you can multiply this against your forces:

Code: Select all

body.f = cpvadd(body.f, cpvmult(direction, force));

[lucas]

That's what I'm doing, and it would work fine if the direction to the gravity point never changes. It does, however, and velocity from previous forces are pointing in the wrong direction. What I'm asking is how I can make the velocity point in the new direction. Maye cpvrotate (except I have no idea how complex rotation works)?

Watch the video to help explain.

[slembcke]

Increase gravity and/or drag? I don't know if you want to fudge the velocity numbers. That seems like an odd thing to do to me.

If you wanted to try the rotation idea, cpvrotate() the velocity by the normalized delta, then cpvunrotate() by the previous normalized delta. That's the best thing that I canthink of.