rigidbody.awake and sleepingAllowed are funky, as are applyForce and Torque

Bugs
1

With this code in draw, a dropped object shows awake for a while. When it settles, it shows asleep. This seems reasonable.

function draw()
    update(DeltaTime)
    scene:draw()	
    if BoxBody.awake then
        text("awake", 400,100)
    else
        text("asleep", 400,100)
    end
end

However if we say this in setting up the body:

    BoxBody.sleepingAllowed = false

The display now shows “asleep” all the time. From the object’s behavior it seems that it is not asleep, but thinks it is.

I test this with applyTorque and applyForce, neither of which seems to wake up a sleeping object: they probably should. Even when the object is awake, they do not always move the object. I think they probably should.

Below is my full test program:

-- CoCraTu-006

function setup()
    scene = craft.scene()
    --scene.physics.gravity = vec3(0,0,0)
    createFloor()
    createBox()
    scene.camera:add(OrbitViewer, vec3(0,0,0), 20, 1, 20)
    angle = 0
    parameter.action("Twist", twist)
end

function twist()
    BoxBody:applyTorque(vec3(0,40,0))
end

function update(dt)
    scene:update(dt)
end

function draw()
    update(DeltaTime)
    scene:draw()	
    if BoxBody.awake then
        text("awake", 400,100)
    else
        text("asleep", 400,100)
    end
end

function createBox()
    local box = scene:entity()
    BoxBody = box:add(craft.rigidbody, DYNAMIC, 1) -- mass
    BoxBody.restitution = 0.8
    BoxBody.sleepingAllowed = false
    box:add(craft.shape.box, vec3(1,1,1))
    box.model = craft. model.cube(vec3(1,1,1))
    box.material = craft.material(asset.builtin.Materials.Specular)
    box.material.map = readImage(asset.builtin.Blocks.Missing)
    box.y = 5
    return box
end

function createFloor()
    local floor = scene:entity()
    local body = floor:add(craft.rigidbody, STATIC)
    body.restitution = 0.9
    floor:add(craft.shape.box, vec3(25, 0.01, 25))
    floor.model = craft.model.cube(vec3(25, 0.1, 25))
    floor.y = -1.05
    floor.material = craft.material(asset.builtin.Materials.Specular)
    floor.material.map = readImage(asset.builtin.Blocks.Brick_Grey)
    floor.material.offsetRepeat = vec4(0,0,25,25)
    return floor
end
2

My opinion is if you say sleepingAllowed is true, then if the object is sleeping it should stay sleeping until I say sleepingAllowed is false. Then it should wake up when I want it to do something. That gives me control over the object. Whether it works that way now or not I’m not sure.

3

@RonJeffries When I run your above code, even though it says asleep, I can still twist it. It never says it’s awake. Even if I wait a minute, I can still twist it.

4

i cannot. the asleep thing is a bug, i think. i suspect you saw it in the other thread.

5

interesting idea about the allowed waking it up. but the awake flag doesn’t change, as we’ve seen.

do we know where codea lifted physics from?

6

@RonJeffries Codea uses Bullet3D. Having a look at the code, there’s some logic to do with how waking works. There may be a bug in some combination of states, maybe when sleeping is disallowed since the awake property getter checks for Bullet’s ACTIVE_TAG but there’s also DISABLE_DEACTIVATION. I need to go over the logic for this…

Also applyForce() and applyTorque() do not appear to automatically wake up rigidbodies, but that could be added

7

It would seem to be a good idea to add it. Can we wake a thing up by setting thing.awake? Or is that really kind of read only?

Thanks, I’ll see what I can find about Bullet3D.

Oh, and one more thing: is the 2D physics a different engine?

8

I see in bullet docs that apply force doesn’t activate. do we have an activate function in there? it’s definitely not `activated :smile:

9

I’ll need to look into it since I thought body.awake = true would work, maybe there’s something I missed

10

Yes. Note in particular that when we set sleepingAllowed = false, the object.awake remains false forever, i.e. it shows asleep. But it does interact, mostly.

While you’re here, I am positing that physics cycle time may always be 1/60th, rather than varying as draw cycle time does. Can you confirm or deny that, please?

Also, asked elsewhere: fixedUpdate is not called (at least not on my DYNAMIC object. Is it never called, or maybe only on STATIC or something?

Thanks @John !