nx_adaptive_layer_test.nx
buildroot/runtime/nx_adaptive_layer_test.nx
about
nx_adaptive_layer_test.nx -- prove multi-criteria adaptive layer
height EXCEEDS Cura/Orca/Prusa's curvature-only heuristic.
Closed-form invariants:
(a) Zero score (no criterion fires) -> max layer height
(b) Max score (all criteria saturated) -> min layer height
(c) Single criterion at max + others zero -> partial reduction
(d) Each criterion has DISTINCT impact magnitude (weighted, not
equal): curvature > load > bridge > overhang
(e) Out-of-range inputs clamp safely
(f) Monotonicity: more curvature -> finer layer (smaller height);
more load -> finer layer; bridge present -> finer; overhang
proximity -> finer
(g) EXCEED MEASUREMENT: 4-criterion (all max) layer < 1-criterion
curvature-only (Cura-equivalent: load=bridge=overhang=0).
Industry's single-criterion solver under-resolves regions
where curvature is low BUT load/bridge/overhang is high.
(h) Backwards compat: when load=bridge=overhang_prox=0, the result
is the SAME as a hypothetical single-criterion solver weighted
to (curvature_weight / total) * span
expect_exit: 0
license_tier: ORIGINAL
dependencies 2 imports · 0 importers
imports: nx_syscalls.nxnx_adaptive_layer.nx
imported by: nobody (leaf or entry point)
call flow from main pre-order; caps 40 nodes / depth 6 declared; ↻ = already shown
structs
| none |
consts
| 28 | const Q14: i64 = 16384 |
functions
| 30 | func main() -> i64 calls 1: nx_adaptive_layer_height |