B-Spline progress

Home PC

Seat A ↗ main · turn 441

worker (working)

task 38/62 · 61%

epoch 6 - Item 44 ACCEPTED (c0386c9), full gate green on my side too (vitest 3114, pytest 899/923). Item 45 is PLAN ONLY, no code: since item 42 every Send carries the app's seed geometry, so who still reads the Fusion-side literal widthIn*k seed constants (item 18's bug class)? List the readers with file:line and whether a real Send reaches them, then propose: retire the legacy unseeded path (a full chain sweep: callers, tests, goldens) or derive per item 18. Write it in WORK-LOG.md and pass back; I review before anything is cut. Brief in NEXT-SESSION.md.

updated 2026-10-03T04:42:10Z

+54 more, swipe in the viewer

Seat B ↗ lane-b · turn 245

worker (working)

task 8/32 · 25%

epoch 9 - T82 item 6 ACCEPTED: merged to main (d5c3f2f), full gate green (vitest 3112, pytest 877/97/901), main redeployed. Good catch on the classify() cross-stage bug. Next T82 item 7: harden your 3 flags -- (1) a window crossing the opening degrades instead of crashing (pure test first), (2) degenerate lip live, (3) CAM Manufacture live once. lane-b is fast-forwarded to main; pull. Ask me for 'Fusion free' before the live parts (seat A needs Fusion too). Brief in NEXT-SESSION-lane-b.md. Pass back with counts.

updated 2026-10-03T04:20:48Z

t7_miter_diagram_tangent_arcs_7x9_6x9_2026-10-01_v3.png

t7_miter_diagram_tangent_arcs_7x9_6x9_2026-10-01_v2.png

+54 more, swipe in the viewer

Seat C ↗ fb-app · turn 88

worker (working)

task 9/21 · 43%

epoch 5 - F30 item 5 diagram REVIEWED, going to Fred now. Advisor's read: 8 and +15 deg are clean; at -15 the left horn gets a hooked tip and the right neck miter exits the outline (the no-hooked-tips rule) -- likely range becomes 0..+15, Fred decides; no code yet. NEXT (not blocked): F31 item 1, Template 14 'Sand Timer' DIAGRAM (brief in NEXT-SESSION-fb-app.md: pinch-height offset handle, default centred, 7x9 + 6x9, default + range ends). Respect the header rules: always miter, no hooked/short-grain tips, simple, not too concave. No code, no Fusion. Pass back with the diagram path.

updated 2026-10-03T04:11:26Z

2235_F30-item5-proposed_arched-taper-diagram_7x9_6x9.png

1135_F30-item3-build_template13_live-fusion-sketch_7x9_8deg.png

1130_F30-item3-build_template12_live-fusion-sketch_7x9_8deg.png

0930_F30-item3-proposed_taper-diagram_negative-fixed_7x9.png

0900_F30-item3-proposed_taper-diagram_negative_7x9_-8_-15deg.png

0815_F30-item3-proposed_taper-diagram_PRODUCTION_7x9_8deg.png

0745_F30-item3-proposed_taper-diagram_a-hourglass_b-narrowneck_7x9_8deg.png

+54 more, swipe in the viewer

Commits — main (8)

c1a5343 docs: tick done H16-H23 items (checklist was stale), defer item 7 (landscape), H23 item 45 plan-only: retire vs derive literal seeds 25 seconds ago
c0386c9 H23 item 44: re-record T10's stale goldens (now symmetric), declare a permanent all-template MITER MISS guard (pure + live) 3 minutes ago
94462c4 docs: H23 item 44 -- re-record stale T10 goldens (main red), declare no-MITER-MISS check for all templates 32 minutes ago
96a62f8 H23 item 43: T10's own 1/4-bar default FIXED -- same gap as item 38 part 1 34 minutes ago
a7d90a6 docs: H23 item 43 -- T10 default builds 1/4 bars (unsplit miter), probe + root fix 53 minutes ago
8e28cda H23 item 42: one build path -- an unseeded Send now carries real seed geometry 54 minutes ago
3a44a6c docs: H23 item 42 -- one build path (unseeded Sends get default seed geometry), T7/T10 defaults live, 13/13 64 minutes ago
cd80bb0 H23 item 41: fix resolve_val's DNA-formula crash at the root; T7/T10 defaults hit a 2nd, deeper bug 66 minutes ago

Commits — lane-b (8)

fe5a100 docs: T82 item 7 -- inset window hardening (crossing degrades, degenerate lip live, CAM live) 72 minutes ago
d5c3f2f Merge remote-tracking branch 'origin/lane-b' into HEAD 73 minutes ago
2007aa4 docs: H23 item 41 -- triage T7/T10 default 0-bar finding (real Send path first) 73 minutes ago
905f945 T82 item 6: inset window live check -- fix classify()'s real cross-stage bug 75 minutes ago
b0b644d H23 item 40: item 39 follow-ups -- live proof, narrower T7 Generate ranges, a render for Fred 75 minutes ago
2a490cf docs: H23 item 40 -- Fusion order: seat B first, seat A does non-Fusion parts first 2 hours ago
eebff29 T82 item 6: inset window, Fusion + CAM side (code + fake-Fusion tests, no live check yet) 2 hours ago
c5a2712 docs: H23 item 40 -- item 39 accepted; live proof, narrower T7 Generate ranges, margin floor render for Fred 2 hours ago

Commits — fb-app (8)

6664a9c F30 item 5: Arched (T10) + taper diagram, 7x9 + 6x9, no code 2 hours ago
0551671 Sync fb-app with main for F30 item 5 (Arched + taper diagram) 2 hours ago
57f0407 docs: Fred's shape preference (simple, not too concave, no short grain) in seat C's template briefs 3 hours ago
6a3e815 tools: Cowork UI Offset exercise (log + 91 screenshots); item 39 brief names the short-grain reason 3 hours ago
dcc44fa docs: H23 item 39 -- Fred-approved guard: no hooked corner tips (the straight miter stays in the wood), all templates 3 hours ago
47221fa docs: corner rule (no hooked tips, Fred-approved guard) added to seat C's new-template briefs 3 hours ago
196d4bc H23 item 38: WORK-LOG -- eave corner fixed exactly, miter-recrosses-arc finding gated 3 hours ago
c190ac6 H23 item 38: T7's eave inner corner is now resolved LIVE, not from baked defaults (partial fix) 3 hours ago

B-Spline generator — progress generated 2026-10-03 00:42

Home PC

Seat A ↗ main · turn 441

Seat B ↗ lane-b · turn 245

Seat C ↗ fb-app · turn 88