changed when planning is undertaken but this only affects copies of the agent’s world state, it never affects the actual agent’s world state.
So now every time the GoTo goal is selected, the atTargetNode symbol will always be false and thus the GoTo action gets picked by the planner to solve the goal. When the GoTo action is executed the effects of the action aren’t actually applied to the agent’s current state and so the agent’s atTargetNode symbol remains at false for the entirety of the game. This ploy was also used for the targetIsDead symbol but of course there are still world state symbols such as inMeleeRange, weaponLoaded etc. that are updated every frame as they are needed by the planner to discover when actions are complete or when goals are relevant.
Once the GOAP system was debugged and working with the sample actions and goals, it was imported into the Domination game and work on the agent architecture begun. However some unforeseen problems occurred during the implementation phase of the agent architecture that needed some attention.