This is an earlier experiment. Read the amendments below before treating its findings as current. This source entry is marked unverified.
The question
Two test cases in the port-launch scenarios delivered nothing. The launch check compared wind strength with 0.8 of cruise speed whatever its direction, so a crosswind grounded those ports.
What was tested
We tested a direction-aware check: launch if the leg can still be flown at a fifth of cruise speed over the ground. We also tested booking arrival slots by real speed over the ground.
What was found then
The full test set stayed at 0 collisions while trips rose from 60,884 to 61,115. The long-haul scenario lost trips until the booking used real ground speed. A full rerun and a sweep of the threshold were queued.
What we learned later
Never adopted: testing paused on 2026-09-08, and the simulator still holds a craft on the ground when forecast wind exceeds 0.8 of cruise speed, by strength alone.
equilibrium/viz/sim_core.js:2416, :2567
What changed
Nothing at the time (still in test).
Numbers and source references
- 0.8
- share of cruise the old gate compares wind againstequilibrium/viz/sim_core.js:2416, :2567
- a fifth
- ground-speed threshold of the new gate (Flight Plan only)research/flight_plan.html
- 60,884 → 61,115
- matrix trips (Flight Plan only)research/flight_plan.html