Clock geometry for the remaining six stretches
This saved-only screen checks whether the two-sided policy can obtain the correct adjacent packet intervals. No original recordings, native predictions or observed fits. Supported geometry does not establish convergence or change any disposition.
| Stretch | Failed primary checks with supported geometry | Failed repeat checks with supported geometry | Passing smooth repeats farther than 1 µs from proposed knot |
|---|---|---|---|
0e73879f98c3b701fd7c42df |
2/2 | 1/1 | 1 |
370cec66878c4c1cf84b4ceb |
2/2 | 2/2 | 0 |
50f24131843071ae9312c082 |
1/1 | 1/1 | 0 |
b4e44ddc0f130ecd45674c78 |
4/4 | 2/2 | 2 |
bb8061a43b4f14b6262ceda1 |
1/1 | 1/1 | 0 |
fd7b81e48a4122b149c4925b |
1/1 | 1/1 | 0 |
All endpoint geometry · Preserved clock calculations · Two-sided software qualification.
Each proposed boundary relocation is recorded. Close transitions are retained separately; a side sample that cannot be placed reliably must abstain. Native integration, continuity and nuisance-correction checks still need separate finite validation. Passing repeats at different smooth endpoints require timing-profile investigation; they are not replaced by a newly proposed boundary. The current tally remains 47 globe / 0 flat / 6 unresolved.