Physics Engine requests

Must haves:
- llMoveToRot
- llSetDensity and / or llSetMass
- llSetVelocity

Nice to have:
- llSetElasticity
- llSetFriction

P.S. The new llVolumeDetect is really, really cool.

I don't know what this one would do


Can be faked to some degree with llSetBuoyancy()


See llApplyImpulse()



Agreed!

Some explanations:

Must haves:

- llMoveToRot

Like MoveToTarget but for rotations. You set the rotation and the tau and your object moves to that rotation with critical damping.

- llSetDensity and / or llSetMass

Two reasons for this:
- for vehicles, you need extremely low mass or you run out of energy, setbuoyancy doesnt help with this, because it needs energy
- for cannonballs, you want something with high mass so it knocks people out of hte way when it hits them

- llSetVelocity

By which I mean, it will critically damp the velocity to the value here with time tau. You could fake this by making an object of high mass and using ApplyImpulse, but just having SetVelocity would seem to impose much less strain on the physics engine and my head.

One use for this is for mobs which you want to move around at a constant speed.

Nice to have:

- llSetElasticity

Sets behavior when two objects collide. Currently elasticity is close to 1, sometimes you want it much lower than this.

- llSetFriction

By which I mean: SurfaceFriction. Make it possible to make two objects slide over each other without tumbling or jerking to a halt

I support/endorse this product and/or service.

I like all of it, especially llSetMass.

It's silly that you need a bullet meter in length to get a reasonably realistic amount of kickback and impact force.

llRotLookAt



llMoveToTarget or llSetForce

> llRotLookAt

There are two problems with llRotLookAt:
- it points the blue axis at the target, I need the red axis (the forward axis) to point to the target
- there is no way to enforce the angle about this axis, after a while you tend to get angle drift, and your object tilts to one side, or turns upside down

> llMoveToTarget

Not a constant velocity. llMoveToTarget gives a critically damped motion.

> llSetForce

Not a constant velocity. This will accelerate the object, whether it is floating or on the ground. Frictional force on the gorund is constant giving constant acceleration. In the air there is no friction giving constant acceleration.

Also note that llSetForce will not react to external evetns. Eg, imagine I have a train. It should move with constant velocity even if something hits it. llSetForce cannot handle this.

Azelda

I think you are confusing llLookAt and llRotLookAt - llLookAt points the +z axis at a pos, but llRotLookAt (perhaps poorly named) lets you specify a rotation, and leads to neither of the problems you will find for llLookAt (the axis, and the lack of constraint about the axis)

I agree that the llSetDensity, elasticity. and friction would be nice - the material chooser does a little of the latter 2 - you will find that flesh is less bouncey, and glass has less friction - but more control wouldn't hurt I say!

As for the llSetVelocity, right now I *might* use llApplyImpulse, but I would be far more tempted to make my object a "vehicle," and use the motor params. It comes complete with easy-to-use damping tools, and now that I think of it, better friction controls as well.

bbc

> I think you are confusing llLookAt and llRotLookAt - llLookAt points the +z axis at a pos, but llRotLookAt (perhaps poorly named) lets you specify a rotation, and leads to neither of the problems you will find for llLookAt (the axis, and the lack of constraint about the axis)

Ooo, you are right! Cool!

Azelda