Simulation Loop

Understanding the tick-based execution model.

The Main Loop

Each tick follows this sequence:

  1. Update NPCs: Execute NPC AI behaviors

  2. Agent Perception: Update fog of war, detect entities

  3. Agent Decision: Evaluate goals, plan actions

  4. Execute Actions: Process all queued actions

  5. Update Systems: Trading, respawns, market prices

  6. Database Logging: Save snapshots and events

  7. Advance Time: Increment tick counter

Tick Timing

  • 1 tick = 1 simulation time unit

  • 60 ticks = 1 simulation minute (by convention)

  • Real-time execution controlled by tick_rate config

Action Execution

Actions can be:

  • Instant: Complete in 1 tick (e.g., MoveAction)

  • Progressive: Take multiple ticks (e.g., GatherAction)

  • Interruptible: Can be cancelled

Goal Management

Each agent maintains a priority queue of goals:

  1. Evaluate all goals for utility

  2. Select highest-utility achievable goal

  3. Request next action from goal

  4. Queue action for execution

  5. Monitor goal completion

When goals complete or fail, agents re-evaluate.

System Updates

Trading System

  • Process pending trades

  • Expire old offers

  • Match buyers and sellers

Respawn Manager

  • Track dead entities

  • Spawn entities after delay

  • Place in safe zones

Market

  • Update prices based on trades

  • Calculate supply/demand

Database Persistence

Snapshots saved at intervals:

  • Agent state (position, health, inventory)

  • World state (resource levels, active entities)

  • Analytics metrics (economy, social networks)

All actions and events logged for analysis.