Continuous call of a method in a FSM Action

This topic has 0 replies, 3 voices, and was last updated 11 years, 5 months ago by Gavalakis.

Viewing 6 posts - 1 through 6 (of 6 total)

Author

Posts

September 30, 2014 at 8:04 pm #19549

edamiani

Participant

Hi, maybe it’s in front of my eyes, but since I can’t find it, let me ask: whenever I set a method inside a FSM Action, it’s called once and that’s it, but I wanted to use the called method as some sort of Update() function — is there a way of making it being called continuously?

Also, is there a way of accessing the OnStateEnter(IState), OnStateUpdate(IState) and OnStateExit(IState) callbacks of each state inside the visual editor, so I can assign a different action for each one of them?

Thanks and best regards! 🙂
Edgard

October 5, 2014 at 4:36 pm #19554

Gavalakis

Keymaster

Hello Edward,

I am very sorry, I missed your post :/

Regarding you first question, I will add a “Repeat” bool check for those Tasks in the next update : )

OnStateEnter, OnStateExit and OnStateUpdate, are callbacks called on any MonoBehaviour on the agent game object that implement those methods, so that you optionaly can if you want, take complete control of the state’s actions through code.
Can you please clarify what you mean with your 2nd question?

Thanks and again sorry for this late reply.
Cheers!

December 6, 2014 at 10:33 pm #19553

fkollmann

Participant

Has this feature been implemented? I fail to find a ‘repeat’ checkbox both on the task or the state node. Regards, Felix

December 7, 2014 at 12:24 am #19552

fkollmann

Participant

Regarding the 2nd question of Edgard… What he means is that you have separate action lists of state enter, exit, and update… currently it’s on state enter, only.

That would be greatly appreciated if this would be possible to implement. (In fact that is precisely what I was looking for.) Regards, Felix

December 7, 2014 at 1:36 am #19551

fkollmann

Participant

Okay, from my POV there is an super-easy already existing approach to it: create a custom task that describes the whole state:

using UnityEngine;

using NodeCanvas.Variables;

namespace NodeCanvas.Actions

{

[Category("GnomeVille")]

[AgentType(typeof(IntervalTransporterSimulationObject))]

public sealed class IntervalTransporterStateIdle : ActionTask

{

private IntervalTransporterSimulationObject _target;

protected override string OnInit()

{

_target = (IntervalTransporterSimulationObject)this.agent;

_target.SetPaused(false); // ensure it's running

return null;

}

protected override void OnExecute()

{

//EndAction(true);

}

protected override void OnUpdate()

{

_target.MoveToClampedPosition();

StartCoroutine(_target.PlayAnimationIdle(true));

}

protected override void OnStop()

{

StartCoroutine(_target.WaitForAnimations());

}

protected override void OnPause()

{

}

December 8, 2014 at 6:12 pm #19550

Gavalakis

Keymaster

Hello Felix,

Indeed, creating a single task that IS actually the state, is a way to go if you do coding.
Alterinatively (as said earlier) you can leverage the State callbacks on any Monobehaviour on your agent game object. Very similar to how the new Unity 5 StateBehaviours work.

Now, I am a bit skeptical on having 1 action list for each OnEnter, OnStay and OnExit. Both because that the UI will become very cluttered and I fear that in the end it will create more confusion. Often, sticking to limitations create a better workflow. At least in my point of view 🙂

Final, the reapeat, has not yet been implemented as there were some issues, which have resolved in the work I am lately doing, so hopefully it will be there for the State Node.

Thanks and best regars,
Gavalakis

Author

Posts

Viewing 6 posts - 1 through 6 (of 6 total)

You must be logged in to reply to this topic.