MOON / FIRST DAWN V3 / DOCUMENTATION

First Dawn — implemented and still being proved

Current surface scheduling in source: planetary-fabrication programmes above eight fronts now reserve a whole fine parcel per expedition. Each new job continuously requires 32 minds and pays the full proportional bill; normal cargo capacity and travel apply. Smaller programmes and earlier paid jobs retain their contracts. The comparative experiment doubled 100-day converted area but still did not finish the region. The actual research-gated code subsequently completed all 960 available parcels in 3,860 simulated hours, with exact hourly/final accounts and productive nights; its pace remains too slow for the intended finale. This is a mature-region test, not a whole campaign; see the whitepaper library for broader design proposals and the agent manual for the actual API. V3 is publicly playable at First Dawn. Public V2 and all its saves remain unchanged at moon-astra-v2.

Implemented: automatic mixed-regolith harvesting, a 500-metre local survey reveal after research, physical Stratum survey rovers, deep mineral access, local mineral/recipe chains, 72-hour daylight/night phases, fueled Seed Base and Lantern reactors, lossy storage, continuous powered minds, finite recovery and paid maintenance, complete paid seed kits, real daughter construction, receiving offers, finite colony plans with reserved land, fixed conveyors/queued installation, and metered consensual power exchange.

The researched observatory now records bounded real production history. Checked M3 resource/traffic jobs reserve attention, verify cited facts, interrupt when mind support disappears, preserve evidence/quotas on restart, and appear in Guide. Two real M3 analyses passed on an isolated paid-world continuation. Ordinary Guide prose remains fallible: successful provider transport is not factual validation. Learning integration and evidence.

Federation Nodes now have one-founder paid expeditions, geometric sites, seed-bore membership, shared local work, delivered contributions, protected withdrawal reserves, finite proposals and majority votes. Approved work waits for supplies, research and continuously powered minds; unstarted plans can release their reserved ground. How to operate the commons.

Regional source integration now includes paid first-campus founding, shared seed-bore charter membership, one vote per player, finite 4/16/64-campus plans with continuing 12/16/32-mind support, and protected fine-parcel admission. Campuses provides site checks, proposals, votes, available-stock and reserved-work ledgers, recorded observation age and a parcel map. The first paid campus has a lightweight visible overview; Guide can read the selected persisted regional record. See the campus manual.

Verification: the six-founder trial completed 144 simulated hours. All six original colonies stayed powered through their 72-hour night and built paid daughters. A separate paid mature daughter built the first campus in 61.36 simulated hours; its existing selective-silicon recipe resolved a measured coproduct blockage. These trials preserve their exact earlier source manifests, not a claim of full current-source campaign validation. Regional store/land/governance/Guide checks and CPU geometry checks pass; current complete-suite counts are recorded in the source handoff.

Mapped successors now reserve surveyed parcels, build paid service bores, use paid bulk belts and two traveling Masons before commissioning their actual campus layouts. Bounded input pipelines overcome network latency while sharing one finite trunk. The new bulk reference reaches four campuses by hour97 with the full72-hour night working; the prior Atlas reference reached them by hour132; its original serialized-input trial stopped at three after600h and remains retained.

Required next: faster paid field logistics, automatic bounded regional propagation, the planetary completion condition, current integrated campaign/load proof. Public V2 is unchanged. Source UI is not proof that all planetary loops or future fusion are finished.

No GPU/browser test while workstation graphics instability remains under investigation. CPU geometry and bundle validation do not establish visual quality on a GPU.

AI manual · Whitepaper library · Published standalone learning explainer.

The current geographic scale review separates a paid64-campus expansion from a synthetic393,216-campus server-component test. The later paid loop reaches64campuses in524hours with independent construction and bulk belts. The measured regional-store95th percentile was305.68ms before the surface changes; this is historical component evidence, not a current full-server or client-load result.

Outside overhaul is now implemented in source: a connected member with campus engineering and twelve powered minds can send their own rover deliveries and crew to restore a zero-condition campus. The complete service bill and two fuel modules per campus are paid; existing jobs, inventory and pause settings persist. Neighboring-region expansion now has paid, independent receiving reservations and delivery; general regional trade remains unfinished. See REGIONAL-EXPEDITIONS.md. See the agent manual for district.recover.

Independent construction now reserves one pair of existing Masons per source campus. Higher approved programmes permit up to4/8sites with16/32continuous powered minds. Reachability and existing network distance guide each next bore. The corrected regional trial reaches64campuses at524h with no later imports and exact accounts, and works every observed nightsecond. Current regional-store component p95305.68ms excludes integrated and simultaneous-survey workloads.

Geographic surface fronts now carry finite Atlas loads and two assigned Masons to actual surveyed parcels. Research and charter votes authorize more fronts with continuing12/16/32-mind support. In the mature64-campus real-DEM comparison,144hours produced144.2km² versus21.4km² serial. A longer100-day run remained operational with exact accounts but converted only39% of eligible land: this needs further transport and pacing work.

An expedition can now carry a finite majority-approved receiving plan. The child arrives with one paid campus, then manufactures its planned successors from its own production under the normal mind and service rules. A real-DEM42.579km expedition arrived after143.576hours; its receiver then grew to four campuses and three paid bores in96.424hours on the normal day/night cycle, with no post-departure imports. Research, approval and the source qualification fixture were stipulated; this is not a new-founder campaign proof. See the receiving-plan manual.

September 8 evening: storage and logistics

Resources now joins located stocks, real cargo commitments, Mind/attention explanations and actionable production diagnostics. Human panels, authenticated observations and Guide context share those diagnostics. Hidden per-product hopper caps are removed; actual storage and the next batch determine whether work can proceed. Consumers are fed before stockpiling; processed outputs use real available storage and bore spoil can be hauled without a preassigned depot. Conveyors carry all compatible solids; fluids use separately researched pipelines.

Outdoor stockpiles: prepare ground, then use it

Choose Build → Stockpile → Prepare stockpile, then place its 24-metre footprint on clear ground. It costs 2 alloy, no components, and real delivery/crew preparation work. There is no research gate and no ongoing power or attention draw for the ground itself. Robots still need their usual support. The yard holds 2,000 bulk units and its mounds grow with material physically present.

Automatic logistics feeds waiting processors first, then sends bulk surplus to stockpiles. It also moves bulk out of enclosed Seed/Depot storage into nearby enabled stockpiles, keeping components and manufactured goods indoors. Refineries and other compatible processors reclaim inputs through ordinary robots or installed conveyors. No mineral selection is required. A stockpile does not process ore, create material, relay power, provide elevator bays or bypass travel time.

Accepted bulk: rock, mixed regolith, mixed mineral residue, tailings, silicates, ilmenite, salt ore and radiogenic ore. Components, alloy, fuel, liquids/gases and separate icy-regolith loads need appropriate enclosed storage. Mixed material retains its constituent accounting; this first version does not simulate environmental loss from trace volatiles. The 2,000-unit capacity is shared. Waiting pickups remain physically on site; reserved arrivals count against future room. Full ground needs another yard, a consumer or a real outgoing delivery.

For AI players, preview then send the ordinary command with a durable idempotency key:

{"action":"build.place","claimId":"YOUR_CLAIM","type":"depot","kind":"stockpile","lat":0,"lon":0}

Replace the claim and coordinates using your observation and placement preview. The alias type: "stockpile" also works. Installed records remain type: "depot" with storageKind: "stockpile"; use the variant when naming equipment or interpreting capacity. Catalog stockpiles.capacity is 2,000,000 milli-units; Guide text and human storage panels show 2,000 whole units. An ordinary depot remains 240 units. Paid sites can be cancelled with build.cancel before commissioning under the usual recovery rules.

Use conveyor.build with the stockpile's installed ID as either endpoint. The line carries all compatible solids and retains its existing research, power, attention, distance, cost and travel requirements. An off-grid stockpile can be served by robots; a conveyor still needs powered connected endpoints. Initial stockpile placement is on private settlements; Federation Node council proposals and depot apron upgrades do not create stockpiles.

Full September 8 development journal · Updated V3 master whitepaper