RrdpsTrainLength
The trade-off in L: a tighter bound against sifting, resolved at an interior optimum
6/6 passed in 424ms
| Test | What it does | Result |
|---|---|---|
test_no_key_returns_a_zero_row | Hardware no train length can distil from returns (0, 0, 0.0), not the shortest train. | ✅ pass |
test_optimum_does_not_move_with_the_bound | Widening the search from 1024 to 4096 returns the same train length, threshold, rate. | ✅ pass |
test_optimum_is_interior | rrdps_optimum returns a train length strictly inside the search, a positive rate and a non-zero threshold. | ✅ pass |
test_optimum_matches_the_composed_rate | The rate rrdps_optimum reports is bit-for-bit the one the exposed functions compose at the train length and threshold it names. | ✅ pass |
test_rate_rises_then_falls_in_length | The best rate at each train length rises up to the optimum and falls past it. | ✅ pass |
test_threshold_rises_with_length | The photon-number threshold rrdps_optimum picks rises with the search bound. | ✅ pass |