Trade caravans crossing Petriarch's dead zone

What shipped

Petriarch’s first authored social layer. Two societies now trade complementary goods across a barren gap with no food in it.

Agents climb a long-range supply-scent toward the nutrient they lack. A provisioning gate lets only the well-fed attempt the crossing. A carry/return state machine hauls cargo home, so goods move both ways without either population emigrating. Carriers wear a visible caravan road into the dead zone, and a slow-decaying amity field cools the frontier into a pacified market.

This was the first Tier A/GPU change since the WebGPU port. The scent and gate terms ship in steer.wgsl, re-verified on the 3090.

Decisions

Long-range reach was the hard part. The obvious design, a deposited demand field, failed (below), so the supply-scent is a static, monotonic cone anchored to each nutrient’s region, and deficit-weighted agents climb it. Weight 0.6 is the cap. At 1.0 region A empties out, which is migration, not trade.

Provisioning is one energy-reserve gate on the scent pull instead of the three levers I’d planned. The gate subsumes two of them and only lures agents with enough reserve to survive the crossing.

The return leg is required. Without it the two societies blur into one. So cargo is carried, not consumed, and “home” reuses the existing claim-field gradient.

What broke

The deficit-broadcast demand field failed outright. With both regions eaten down to scarcity, agents are hungry for both nutrients, so demand tracked population density and peaked inside each region. Climbing it herded agents toward their own centre and gap traffic went down. A spatial probe proved it before the rewrite. Diffusing the capacity field was tried first and also rejected, since it decays the far signal to noise and is never monotonic across the 20-cell gap.

This was the most useful failure of the phase. Broadcast demand and let agents answer it is the design a doc approves and a simulation vetoes. You only find out by running it.

Amity barely moved at first. Suppress at 0.5 and per-trade volume at 5 capped peak amity near 4, with under 1% of fights suppressed, because at fast decay each deposit faded before the next sparse frontier trade. Slow decay (0.998) fixed it.

The gate’s nonlinearity nudged the CPU-f64/GPU-f32 steer divergence up for a few borderline agents. A seed sweep showed 0 to 3 of ~3900 agents, non-systematic, so I widened the verify tolerances instead of chasing a phantom logic bug.

Numbers

Amity from amitycheck.ts (16k x 4 seeds): pacified cells 13 to 245, global fights/k down 16% versus trade-only and 26% versus no trade, TRADE selection 0.45 to 0.49. Crossing numbers from crossing.ts (3 seeds x 8k). A later conflict-recession run (seed 11, 20k ticks) confirmed it globally: per-capita fights fell 70%, 1.25 to 0.38/k/agent, as commerce took over.

Next

Territory, the second social layer. Societies fight harder on home ground and hold coherent borders.