Mouse Entered and Exit aren't working how they should be (Solved)

Holo74

I'm using 2 3d area nodes. I hooked up their signals from the mouse entered and mouse exit to change color when you hover over them. However, it seems to be very glitchy. I turned off the second camera that works with the guns to see if that helps but I just encountered more bugs. It seems that the rotation of the camera changes how the mouse enter works. I found out that when I look at the floor it never works ever remotely close to the objects. When looking parallel to the ground it works most of the time with some odd occurrences. When I look straight up the hit boxes are much larger but they flicker as you get to the actual box. What seems to be going wrong here? What other info might help deduce what is going on as well? Edit: Forgot to mention that purple on the left and blue on the right are deactivated while red on the left and green on the right are highlighted. Pictures for reference:

TwistedTwigleg

To be honest, I've always just used ray-casting to handle any 3D buttons. I find it gives me more control of when I detect buttons and most importantly, it works on all angles and cameras (if using code similar to the documentation)

Maybe try using the input_event signal if you are not already and see if its more consistent? I don't know if a signal is sent for mouse enter/exit though.

Megalomaniak

I think the mouse_entered/_exited signals are for control nodes only?

Holo74

No, I found that the collision nodes also have them. The problem is that the collisions can vary so wildly with the mouse enter and exit when the camera rotates. Also, the enter and exit are mainly to know when to highlight an object. Unless there is another way to know to do that I'm kind of stuck with those signals.

Megalomaniak

Well, as @TwistedTwigleg said, you can try using raycast.

Holo74

I'm not sure how to work the raycasts. I tried to set the one in the documents but I think because it is under the player which rotates the raycast gets messed up. I set the final to variable to the cast to in the raycast and that doesn't return what I click on. In fact, I'm not even sure what I'm clicking on because I have no idea where the ray is pointing to. It seems like it is pointing in the completely opposite direction cause I can't see what it is doing in game. I tried setting it to the immediate geometry but now I can't see that either. I was able to set it up in a test project but if the camera rotates at all I need to set the rotation of the ray cast to nothing or just have it be not under the rotating camera. I thought giving a new transform with the same global origin would be the same but that doesn't help at all. Not to mention that I want to have it on a completely separate camera because now the menu clips into the ground and surrounding walls. This seems like so much of a headache with the only solution that works being a 50/50 chance of not working depending on which direction I'm looking in. Maybe I could separate the second camera completely from the players body and just have it be out in the middle of no where with all of the tools and this book menu. Though, that means the book menu can't react to the natural lighting around the player.

TwistedTwigleg

Can you share the code you are using for the raycast? In my experience, using the code provided in the documentation has worked without any issue.

Holo74

    private Camera camera;
    private const float rayLength = 1;

    public override void _Input(InputEvent @event)
    {
        if (camera == null)
            camera = PlayerController.Instance.camera;
        if (@event is InputEventMouseButton eventMouseButton && eventMouseButton.Pressed && eventMouseButton.ButtonIndex == 1)
        {
            Transform holder = new Transform();
            holder.origin = GlobalTransform.origin;
            GlobalTransform = holder;
            Vector3 from = camera.ProjectRayOrigin(eventMouseButton.Position);
            Vector3 to = from + camera.ProjectRayNormal(eventMouseButton.Position);
            CastTo = to;
        }
    }

    public override void _PhysicsProcess(float delta)
    {
        if (IsColliding())
        {
            GD.Print(GetCollider());
            if (GetCollider() is Tabs tab)
            {
                tab.OnTab(true);
            }
        }
    }

This is on the raycast node with the C# class deriving from the Raycast class.

The raycast node itself is under the camera that I use for the players head.

TwistedTwigleg

You need to send the raycast out into the physics world, that is why its not working. You will need to get the space-state and then send the raycast. Since you are getting the input in the _input function, you'll want to cache it so you can use it in _physics_process.

I believe something like this should work like you are expecting:

    private Camera camera;
    private const float rayLength = 1;
    private bool do_raycast = false;
    private Vector3 ray_origin = Vector3.Zero;
    private Vector3 ray_end = Vector3.Zero;

    public override void _Input(InputEvent @event)
    {
        if (camera == null)
            camera = PlayerController.Instance.camera;
        if (@event is InputEventMouseButton eventMouseButton && eventMouseButton.Pressed && eventMouseButton.ButtonIndex == 1)
        {
            Transform holder = new Transform();
            holder.origin = GlobalTransform.origin;
            GlobalTransform = holder;
            Vector3 from = camera.ProjectRayOrigin(eventMouseButton.Position);
            Vector3 to = from + camera.ProjectRayNormal(eventMouseButton.Position);
            
            ray_origin = from;
            ray_end = to;
            do_raycast = true;
        }
    }

    public override void _PhysicsProcess(float delta)
    {
        if (do_raycast == true)
        {
            DirectSpaceState space_state = GetWorld().DirectSpaceState;
           Godot.collections.dictionary raycast_result = space_state.IntersectRay(
               ray_origin, ray_end);
            if (raycast_result.Count > 0) {
                GD.print(raycast_result)
                if (result["collider"] is Tabs) {
                    result["collider"].OnTab(true);
                }
             }
            do_raycast = false;
        }
    }

Holo74

It nearly works... The ray detects the an object but it doesn't show up under "collider". I also had to ray length to the from. This almost solves the problem but it doesn't actually solve it fully. Is there any way to figure out the key that it might be falling under?

Holo74

Nope... your code worked and I just forgot that the tab class was on the parent object. Once I got the parent I was able to send the code to the parent and it worked... Thank you so much for the help and I'll mark this post as solved now.