Prototype graphics-state world-worker walking

[SimGamerJen]

Summary

Adds the first controlled locomotion experiment on top of the Alpha 4 movement telemetry probe.

What the probe established

• the real player exposes PlayerMover -> PlayerGraphicsState -> HumanGraphicsComponent
• PlayerGraphicsState is where isIdling, isWalking, and isRunning actually transition
• the raw animation getter does not reliably expose those boolean states even while movement is visibly active
• GIANTS' on-foot update applies graphics state first, then model yaw/position, then updates the human graphics component

Prototype scope

This PR intentionally does not add pathfinding or target selection.

It adds one diagnostic command:

• hpWorld walk [slot] [metres] (default 5 m)
• hpWorld walkstop [slot]

The selected placed worker:

1. keeps its current facing,
2. accelerates into a 1.35 m/s straight walk,
3. advances along terrain for the requested distance,
4. decelerates to a stop,
5. persists its new world placement.

During the test a HumanGraphicsComponentState owns absSpeed, idle/walk/run flags, movement direction, grounded state and NPC state. The Alpha 3 FACE ME/root-yaw layer is suppressed for that worker while locomotion is active, so there is one graphics owner during the experiment.

Test focus

1. Confirm [WorldLocomotion] Loaded ... prototype-1 appears.
2. With A placed outdoors on level ground, run hpWorld walk A 5.
3. Observe whether A uses a genuine walking gait rather than sliding/swivelling.
4. Confirm acceleration and stopping look smooth.
5. Confirm there is no jitter.
6. Confirm the worker remains at the new location after completion/reload.

If straight walking is visually correct, the next experiment can add heading changes/curved movement. If it is not, the log will tell us whether the graphics-state route is being accepted before we invest in pathfinding.

[SimGamerJen]

In-game locomotion validation — PASS

Tested on FS25 1.21.1.0 with Rhys (slot A), outdoors on level ground.

Observed result:

• worker visibly moves using a walking gait rather than jittering
• movement speed is comparable to the player's slow walking movement
• no jitter observed
• acceleration/start and deceleration/stop are smooth; visually comparable to native player movement
• overall presentation is close to what a real player walking on a multiplayer server would look like

Log validation:

• prototype loaded successfully: 2.2.0.0-alpha4-locomotion-prototype-1
• repeated hpWorld walk A 5 tests completed successfully
• each 5 m movement ran at the intended 1.35 m/s and completed in ~4.03–4.04 s
• each ended with WALK STOP A reason=target-reached
• no locomotion Lua errors observed

Conclusion: the HumanGraphicsComponentState / applyState() locomotion route is viable. Next experiment should add heading changes/curved movement while preserving this single graphics-state ownership model.