feat(showcase): add rope-bridge

[TMHSDigital]

Adds showcase/rope-bridge. It is a rope footbridge with log posts and sills through-tenoned to the post centres. Twenty planks are seated on two laid foot ropes, the hand ropes are lashed to the posts, suspenders tie hand rope to foot rope, and the foot ropes are staked off behind each end.

Selection: it is the first unclaimed candidate in the ROADMAP pool (line 136). It is a recognizable prop with no API contract, so it goes in showcase, not examples. No open PR or feat/showcase-* branch covered it.

Binaries (all local, .scratch/):

• blender-4.5.11-windows-x64, reports Blender 4.5.11 LTS
• blender-5.1.2-windows-x64, reports Blender 5.1.2
• blender-5.2.1-windows-x64, reports Blender 5.2.1 LTS

The budget that matters: deck parabola (exit 19)

A deck laid on two straight ramps has the same ends, sag, planks and bounding box as one that hangs, so no envelope check can tell them apart. The check fits a least-squares parabola to the plank top-face centres read off the mesh. Every plank must land within 6 mm of the fit, and the fitted sag must be 0.22 ± 0.015 m.

• Default: residual 2.97 mm, sag 0.2204 m.
• --vee-deck: its fitted sag is 0.215 m, still in band, and every other budget passes. It exits 19 only on the 32.5 mm residual. live-run-proven

Falsifiers: exit codes per binary (live-run-proven)

	default	--skip-decimate	--stray-vert	--lift-z	--float-post	--short-rails	--float-planks	--loose-lashings	--float-suspenders	--lean-post	--vee-deck	--drift-planks	--slack-rails	--sharp-plank	--low-bake	--output
4.5.11	0	9	15	16	16	17	18	18	18	19	19	19	19	20	21	—
5.1.2	0	9	15	16	16	17	18	18	18	19	19	19	19	20	21	—
5.2.1	0	9	15	16	16	17	18	18	18	19	19	19	19	20	21	0

Flag	Target budget	Measured	Exit
--skip-decimate	LOD1 ratio	1.0000	9
--stray-vert	hygiene	loose_v 1	15
--lift-z	grounded zmin	0.050	16
--float-post	named supports (4 posts + 4 stakes)	worst 0.012	16
--short-rails	rope-end / tenon bite	−19.3 mm	17
--float-planks	plank seat band	−3.0 mm	18
--loose-lashings	lashing hoop band	−4.0 mm	18
--float-suspenders	one connected assembly	15 components	18
--lean-post	plumb	19.4 mm	19
--vee-deck	deck parabola	32.5 mm	19
--drift-planks	plank pitch	31.7 mm	19
--slack-rails	rail height	0.565 m	19
--sharp-plank	edge treatment	12 right-angle edges	20
--low-bake	baked texels per UV island	9.42	21

Envelope: every falsifier stays inside BBOX_TOL (20 mm). The worst cases are --float-post at +12 mm in Z and --loose-lashings at +13 mm in X and Y. live-run-proven

Checks run locally

• tests/check_falsifier_targets.py: static check passes. --run <BLENDER> --only rope-bridge also passes on all three binaries. live-run-proven

• tests/smoke/run_example.py --forbid-skip: [PASS] rope-bridge on 4.5.11, 5.1.2 and 5.2.1. live-run-proven

• Other local checks:

  • check_exit_code_readme, check_import_export_rules, check_engine_id, check_gallery_images, test_harness
  • the inline validate-manifest and validate-counts jobs, run in a clean worktree (29 showcase pieces)

  All pass. live-run-proven

• Determinism: every measured line is byte-identical across 4.5.11, 5.1.2, 5.2.1 and a second 5.2.1 run. That includes both LOD counts and the 300212 B glTF. Check-only wall-clock is about 2.0–2.2 s. live-run-proven

• Ecosystem drift check: remote action, so CI-proven only.

• 4.5 / 5.2 smoke jobs: CI-proven on this PR. needs-5.1 is applied because the pipeline bakes.

Findings the budgets forced during authoring

• Coplanar pairs. The coplanar budget counted 84 cross-shell pairs. The two lashing turns on each post were one section stacked, so their outer faces shared planes. The second turn is now rotated half a vertex step, and the count is 0.
• Right-angle edges. The edge budget found 6 right-angle edges. The foot rope's bend over the sill arris had a radius smaller than the rope, so a ring every few degrees folded the section through itself. It is now a single mitred ring, and the count is 0.
• Framing. The framing gate failed the first camera (bottom margin 0.017). The camera was re-aimed to fill 0.822 × 0.761, with a minimum margin of 0.069.

Visual gates

Contact sheet: docs/gallery/contact-sheets/rope-bridge-contact-sheet.webp. The calibration band is armature-bend / damped-track-aim / bmesh-gear.

criterion	rope-bridge	calibration	verdict
mean luminance	0.2216	0.151–0.302	in band
stage darkness	0.2233	0.145–0.236	in band
wedge warmth	+0.139	−0.43 … +0.37	warm, milder than two of the three
saturation	0.131	0.144–0.231	just under the set; in line with the timber showcase peers (archway 0.090, trough 0.136)
subject fill	0.822 wide	—	passes the framing gate
thumbnail legibility	—	—	reads as a rope bridge at tile size

The first draft's stage was 0.243, above the brightest calibration stage: a key light at 900 W flooded the back wall. It was relit at 660 W. inspection + measured

Asset sheet: docs/gallery/asset-sheets/rope-bridge.webp. Verdict: passes. Rendered alone under neutral light, it carries more construction than any reference:

• grained, chamfered planks
• laid rope, lashings hooped onto their posts, staked tie-offs

It is not the least-designed object; the rocket and the mech arm are simpler. inspection-only

Close-ups inspected (sill wrap, post lashings, stake ties, plank seats from below, deck end, elevation):

• No z-fight speckle and no chrome.
• One defect was found: the single-ring post-top dome rendered faceted. A second round-over ring was added before the hero was rendered. inspection-only

The hero was rendered on 5.2.1 through scripts/render_hero.py, and build_gallery.py was run last. The generated <img alt> and witness callout on docs/gallery/index.html and docs/gallery/rope-bridge/index.html were read and are complete. inspection-only

Not applicable, with reasons (in the piece README)

• Mirror-symmetry budget: laid rope is chiral.
• Single convex-hull collider: replaced by a 30-box compound, because a hull over a sagging deck puts walkers 0.22 m above the planks at midspan.
• Masonry, vessel, scatter, fixture, roof, forging and ring conventions: no such parts.

🤖 Generated with Claude Code