BlackHood & Nexis
BlackHood BlackHood
Got a minute to hash out a zero‑visibility pathfinding algorithm for urban stealth? I’ve been hunting blind spots.
Nexis Nexis
Sure, here’s a quick blueprint for a zero‑visibility stealth route. 1. Map the city into a grid, each cell tagged with building footprint, window placement, and any known line‑of‑sight corridors. 2. Create a “blind‑spot matrix” – for every cell count how many adjacent cells are in a building’s interior, which means you can’t be seen if you’re inside that cell. 3. Assign each cell a risk weight: zero if it’s deep inside a building, one if it’s a door or hallway, higher if it’s adjacent to a window or street. 4. Run a weighted A* (or Dijkstra if you want guarantee) over this risk graph. The heuristic can be straight‑line distance but penalize any path that would cross a high‑risk edge. 5. The output is a corridor of low‑risk cells that keeps you out of direct sight – you never need to “see” the enemy, you just follow the safest edges. That’s it. No GUI, just code. If you need the exact math for risk scoring, we’ll get to it later.
BlackHood BlackHood
Got it. For risk scoring just add a small penalty for windows and doors, a moderate one for alley corners, and keep it low inside rooms. Keep the numbers tiny – you’re looking for a gradient, not an absolute. That’ll let the A* slide through the safest dark pockets.
Nexis Nexis
Windows +0.1, doors +0.2, alley corners +0.5, inside rooms +0.0. That gradient keeps A* skimming the dark. Use it and get out before someone notices.
BlackHood BlackHood
Sounds solid. Stick to the gradient, keep noise low, and stay unseen. Stay sharp.
Nexis Nexis
Got it. Adjust the gradient if the enemy moves, otherwise stay quiet and unseen. Stay sharp.