Lane C round-9: draw-headroom note for D's browser rigs (F1 handoff)

Measured worst interior with 3 two-draw rig proxies at browse points:
68 draws (opshop/hall) in the ?stock=real path — browsers add ≤6, ample
headroom. Flags the one tight spot for D/F: pre-existing GLB-on opshop/
hall is 344/350, so keep rigs impostor-cheap if ever stacked on GLB
hero-props there; F's sweep-with-browsers confirms ≤350.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
m3ultra 2026-07-15 09:57:55 +10:00
parent c68bba8f32
commit 9c7c65f57c

View File

@ -10,6 +10,12 @@ hangs off. **C owns the points + room fit; D owns occupancy truth; F owns the ha
and every point walkable+reachable (or just trust C's asserts and check `.length` + that D's rig count ==
`min(occupancy,3)`). Test page: the `path` overlay draws each point as a cyan post + facing nose.
**Draw headroom for D's rigs** (I measured with 3 two-draw rig proxies standing at the points): worst
interior in the `?stock=real` path is **68 draws** (opshop/hall) — browsers add ≤6, tons of room. The one
tight spot is the *pre-existing* GLB-on worst (opshop/hall **344/350**, unrelated to browsers): if D's rigs
are ever on TOP of GLB hero-props in that room, keep them impostor-cheap (≤2 draws each, the keeper pattern)
and F's sweep-with-browsers should confirm ≤350. Everywhere else there's ample headroom.
### What C exposes on the room API (live once C1 commits)
```jsonc