SOLAR WAR

Complete Physics Audit, Game Design, and Technical Implementation Plan

Document status: Build specification
Version: 0.1
Date: 7 August 2026
Intended use: Source of truth for research, prototyping, implementation, balancing, and later narrative design.

Executive decision

Build Solar War as a multiscale, event-driven, partially observable simulation of a mature Solar System defending its distributed stellar-energy civilization from an equivalently capable but technologically asymmetric interstellar expedition.

The simulation is not primarily about ships trading hit points. It is about whether a civilization can keep its distributed metabolism intact while an adversary attacks its:

• usable energy,
• heat rejection,
• trajectory options,
• sensing and prediction,
• communications and trust,
• industrial throughput,
• repair and replication,
• population and continuity.

The canonical hard-science game is governed by six coupled quantities:

\boxed{M,\ E,\ \mathbf{p},\ Q,\ I,\ R}

where:

• M: matter and material inventory,
• E: usable energy and power-delivery capacity,
• \mathbf{p}: momentum, maneuver authority, and remaining \Delta v,
• Q: thermal state and heat-rejection capacity,
• I: information quality, uncertainty, latency, and trust,
• R: industrial reproduction, repair, and replacement capacity.

Three additional nonconserved state variables determine whether those resources can be used:

• coherence: how much of the civilization remains connected and mutually intelligible;
• autonomy: how much useful action local agents can take without fresh command;
• viability: surviving population, life support, archives, ecosystems, and recoverable institutions.

The core simulation should use established physics and conservative extrapolation. Exotic physics is supported only through explicit, labeled modules. The hard-SF version is already sufficiently strange: delayed light cones, uncertain tracks, vulnerable radiators, trajectory-space minefields, autonomous policy execution, industrial epidemics, and battles whose decisive commitments were made years before impact.

Part I — Audit of the four contributed reports

Audit method

Every claim is assigned one of five dispositions:

The reports are valuable idea inventories, not authoritative specifications. Their references range from primary literature and NASA reports to wikis, Reddit, Fandom, and uncited extrapolation. The engine therefore stores provenance and confidence alongside every technology parameter.

Claim-by-claim audit

Audit conclusion

The shared foundation that survives scrutiny is:

1. Dyson swarm, not singular sphere.
2. Trajectory space, not tiled territory.
3. Power is abundant; delivery, coherence, momentum, and cooling are not.
4. Persistent activity is visible, but detection is probabilistic.
5. Information arrives late and uncertain.
6. Radiators and networks are more decisive than armor thickness.
7. Weapons are transformations of mass, energy, momentum, and heat.
8. Autonomous industry and repair decide long campaigns.
9. Exotic physics should be modular and explicitly labeled.
10. The decisive objective is civilizational viability, not destruction of Earth.

Part II — Product definition

1. Player fantasy

The player is not a starship captain. The player is a distributed strategic authority responsible for preserving a civilization whose body spans astronomical units.

The player:

• allocates power among industry, sensing, propulsion, defense, habitat support, and cooling;
• issues policies to autonomous regional commands;
• interprets delayed, uncertain, and possibly deceptive observations;
• shapes future trajectories rather than moving pieces instantaneously;
• decides what infrastructure may be sacrificed;
• manages fragmentation when the system-wide network stops behaving as one civilization;
• chooses whether to defend planets, disperse population, preserve archives, negotiate, quarantine, or retaliate;
• attempts to reduce hostile industrial reproduction below replacement before friendly coherence collapses.

The emotional rhythm should alternate between:

• quiet inference: observing faint anomalies, fitting tracks, debating classifications;
• long commitment: initiating burns, production programs, convoy transfers, and beam reconfiguration;
• compressed crisis: local autonomous engagements resolving faster than remote authorities can intervene;
• delayed revelation: learning that a battle was won or lost hours, months, or years earlier;
• reconstruction: rerouting power, repairing graphs, rebuilding confidence, and deciding what the surviving civilization becomes.

2. Genre and positioning

Solar War should sit between:

• orbital-mechanics simulation,
• grand strategy,
• command-and-control wargame,
• industrial ecology,
• probabilistic intelligence analysis,
• hard-science-fiction world simulator.

It should not be designed as:

• a conventional RTS with realistic-looking units,
• a fleet tactics game with a Solar System backdrop,
• a physics sandbox without strategic meaning,
• a spreadsheet where only the largest joule total wins,
• a narrative game whose science cannot be inspected.

3. Design pillars

Pillar A — The player sees beliefs, not truth

No faction has access to the objective world state. Every contact is delayed, noisy, classified probabilistically, and attached to a timestamp.

Pillar B — Time is terrain

Light delay, orbital periods, transfer windows, integration time, manufacturing lead time, heat accumulation, and reproduction time are all strategic dimensions.

Pillar C — Infrastructure is the army

Collectors, radiators, relays, factories, depots, timing nodes, software authorities, telescopes, and repair clouds matter at least as much as mobile weapons.

Pillar D — Damage is systemic

A unit can be physically intact and strategically dead because it has no cooling, no trusted commands, no propellant, no target solution, or no compatible spare parts.

Pillar E — Scale changes behavior

Local engagements may last seconds. Regional campaigns last months. Interstellar threats unfold over decades. Stellar-engine policy unfolds over millions of years.

Pillar F — Asymmetry is topological

The factions have comparable total capability, but one concentrates energy through networks while the other disperses matter through replication. They should not share mirrored unit rosters.

Pillar G — Every speculative technology declares its assumptions

The game exposes a physics profile and confidence class. A player can inspect which mechanisms are established, extrapolated, disputed, or invented.

4. Simulation profiles

Confidence classes

Every technology definition stores:

physics_class: P1
status: extrapolated
assumptions:
  - coherent_phasing_over_100_km
  - optical_elements_survive_flux
source_ids:
  - S7
  - S9
uncertainty:
  model: lognormal
  sigma_fraction: 0.35

5. Canonical campaign premise

Humanity: the Heliocentric Commonwealth

Humanity is not politically unified, but most habitats and machines participate in a common interoperability and mutual-defense fabric.

Its characteristic technology is networked stellar industry:

• immense but visible power collection;
• synthetic-aperture sensing;
• open manufacturing standards;
• beam-powered transport corridors;
• extensive timing and navigation infrastructure;
• highly redundant, heterogeneous habitats;
• rapid software and tooling adaptation;
• slow political consensus;
• dangerous dependence on authentication, synchronization, and shared protocols.

The adversary: the Inward Bloom

The canonical alien expedition is not a fleet. It is a lineage of self-reproducing industrial seeds whose internal substrate, computation, and fabrication are radically integrated.

Its characteristic technology is distributed matter conversion:

• low-power dormant travel;
• slow initial growth followed by branching replication;
• local adaptation to available material;
• no indispensable capital ships;
• few high-output emitters;
• highly coherent local decision-making;
• limited compatibility between specialized lineages;
• vulnerability to sterilization of rare catalysts, seed templates, or developmental stages.

The two sides are balanced by campaign potential, not mirrored technology:

\text{human advantage} = \text{power concentration} + \text{aperture} + \text{local geography} + \text{repair standards}

\text{alien advantage} = \text{dispersal} + \text{replication} + \text{adaptation} + \text{small target size}

Why the aliens attack

The engine should support multiple hidden strategic motives:

• colonization without intended extermination;
• seizure of stellar energy;
• prevention of perceived human expansion;
• recovery of a preexisting probe lineage;
• quarantine of dangerous human AI;
• ecological conversion;
• coercive bargaining;
• an internal alien schism;
• accidental escalation after autonomous contact.

Alien intent is itself an inference problem. The player should not receive a faction encyclopedia at campaign start.

6. Victory and loss

A campaign does not end merely because a planet is destroyed or a fleet retreats.

Define civilizational viability:

V_f = w_P\tilde P_{\rm usable} +w_F\tilde F_{\rm fabrication} +w_C\tilde C_{\rm network} +w_I\tilde I_{\rm awareness} +w_A\tilde A_{\rm autonomy} +w_N\tilde N_{\rm population} +w_B\tilde B_{\rm biosphere/archive} -w_D\tilde D_{\rm irreversible}

where each tilde is normalized to scenario baselines.

Possible outcomes:

• repulsion: hostile reproductive capacity falls below replacement and remaining seeds are contained;
• containment: neither side is destroyed, but the adversary is confined to monitored regions;
• negotiated partition: both retain viable industrial ecologies;
• human fracture: population survives, but the Commonwealth graph permanently splits into incompatible regional civilizations;
• alien assimilation: the invader captures enough human infrastructure to reproduce using the human energy network;
• continuity victory: inner-system civilization collapses, but remote habitats and archives retain reconstruction potential;
• pyrrhic sterilization: the invasion is stopped by rendering large fractions of the system unusable;
• mutual deterrence: both sides retain distributed retaliatory capability and cease major operations;
• extinction: no recoverable population, archive, or autonomous reconstruction base remains.

The player should be scored on:

• lives and habitats preserved,
• biospheres and archives preserved,
• enemy reproduction suppressed,
• usable system mass retained,
• infrastructure connectivity,
• irreversible debris and contamination created,
• coercive or exterminatory actions,
• time to stable settlement,
• uncertainty at the moment critical decisions were made.

Part III — Physics charter

7. Non-negotiable invariants

The core engine enforces:

1. Finite signal speed t_{\rm arrival}=t_{\rm emission}+\frac{R}{c}

2. Mass accounting Material cannot appear because a build queue completed. Feedstock, machine tools, losses, and transport are explicit or aggregated.

3. Energy accounting Stored energy, generated power, conversion efficiency, beam delivery, and heat are distinct.

4. Momentum accounting Propulsion and projectile launch produce reaction momentum unless external beams, sails, or environmental exchanges supply it.

5. Thermodynamic accounting Useful work produces waste heat. Suppressed radiation increases stored heat or exports hot mass.

6. Trajectory continuity Objects do not choose arbitrary positions. They follow gravity and applied acceleration.

7. Measurement uncertainty Sensors produce observations, not omniscient object records.

8. Causal command A remote authority cannot change a local unit’s policy until a message arrives.

9. Engineering throughput Large energy totals do not imply arbitrary instantaneous power, stress tolerance, coherence, or manufacturing rate.

10. Damage-channel specificity Defenses are effective against particular energies, particles, spectra, directions, or sizes.

8. Fundamental state variables

Physical state

For explicit entity i:

\mathcal X_i = \{ t,\mathbf x,\mathbf v,\mathbf q,\boldsymbol\omega, m,\mathbf p,Q,\mathbf H,\mathbf C \}

where:

• \mathbf q: attitude quaternion,
• \boldsymbol\omega: angular velocity,
• \mathbf H: component health vector,
• \mathbf C: component operating modes.

Resource state

\mathcal R_i = \{ m_{\rm dry}, m_{\rm prop}, m_{\rm feedstock}, E_{\rm stored}, P_{\rm gen}, P_{\rm routed}, Q_{\rm stored}, A_{\rm radiator}, C_{\rm compute}, B_{\rm comm} \}

Epistemic state

Faction f’s estimate for entity i:

\mathcal B_{f,i} = \{ \hat{\mathbf s}_{i}, P_{i}, p(\text{class}), p(\text{faction}), p(\text{intent}), t_{\rm last}, \mathcal T_{\rm provenance} \}

with state vector:

\hat{\mathbf s}_i = [x,y,z,v_x,v_y,v_z,\ldots]^T

and covariance P_i.

Social/authority state

\mathcal G_f = \{ \text{trust graph}, \text{command graph}, \text{software compatibility}, \text{political commitments}, \text{rules of engagement} \}

9. Canonical units

Internally use SI:

• seconds,
• meters,
• kilograms,
• radians,
• joules,
• watts,
• kelvin.

The UI may display:

• AU, km, light-seconds, light-minutes, light-hours;
• km/s and fractions of c;
• megawatts through yottawatts;
• tonnes through planetary masses;
• days, years, kiloyears, megayears.

All scenario/configuration values are parsed with explicit units and converted at load time. Unitless physical configuration values are rejected unless the schema defines them as ratios.

10. Numerical sanity anchors

These are not balance targets; they are checks against silent order-of-magnitude errors.

• Full spherical area at 1 AU: 4\pi(1\,{\rm AU})^2 \approx 2.81\times10^{23}\ {\rm m^2}.

• Full coverage at areal density 0.25\ {\rm kg\,m^{-2}}: M\approx7.0\times10^{22}\ {\rm kg}.

• One-way light time:

  • Earth–Moon mean distance: about 1.28 s;
  • 1 AU: about 8.32 min;
  • 5.2 AU: about 43.2 min;
  • 30 AU: about 4.16 h;
  • 550 AU: about 3.18 d.

• Kinetic energy: 100{,}000\ {\rm kg}\ @\ 10\,{\rm km\,s^{-1}} =5\times10^{12}\ {\rm J}.

• Antimatter: 1\,\mu{\rm g}\ \bar m + 1\,\mu{\rm g}\ m \approx1.80\times10^8\ {\rm J}.

• A 10^8 kg projectile at 0.999c: \gamma\approx22.37,\quad E_k\approx1.92\times10^{26}\ {\rm J}.

• Ideal radiator area for 1 GW at emissivity 0.9:

  • 300 K: \sim2.42\times10^6\ {\rm m^2};
  • 600 K: \sim1.51\times10^5\ {\rm m^2};
  • 1200 K: \sim9.45\times10^3\ {\rm m^2}.

• Diffraction first-minimum scale at 1.064\,\mu{\rm m}, 1 AU:

  • 1 m aperture: \sim194 km;
  • 10 m: \sim19.4 km;
  • 1 km: \sim194 m;
  • 100 km: \sim1.94 m.

These examples show why temperature, aperture, and engineering scale matter so strongly.

Part IV — World, time, and orbital dynamics

11. Objective universe and faction universes

The simulation stores two different kinds of state.

11.1 Objective physical state

This is authoritative and inaccessible to ordinary players:

• true trajectories,
• true component state,
• true emissions,
• true orders,
• true allegiance,
• true factory activity,
• true damage,
• true network compromise.

11.2 Faction belief state

Each faction has a private derived universe based only on received evidence.

A faction belief state contains:

• tracks and covariances;
• classification probabilities;
• source reliability;
• inferred capabilities;
• inferred objectives;
• message authentication status;
• estimated damage;
• unresolved hypotheses.

No UI subsystem should accidentally read objective state except in debug, observer, replay-analysis, or adjudication modes.

12. Time model

12.1 Global simulation time

Use a monotonic dynamical-time scalar, preferably seconds from a scenario epoch expressed in TDB or another clearly defined barycentric time scale.

Store:

epoch_tdb
sim_time_seconds

Convert to human calendar time only for display.

12.2 Event-driven progression

The simulation advances to the next causally relevant event rather than applying a universal fixed tick.

Core event categories:

OBSERVATION_EMITTED
OBSERVATION_ARRIVED
TRACK_UPDATED
MESSAGE_SENT
MESSAGE_ARRIVED
BURN_STARTED
BURN_ENDED
LAUNCH
BEAM_STARTED
BEAM_ENDED
INTERCEPT_WINDOW_OPENED
INTERCEPT
IMPACT
THERMAL_THRESHOLD
COMPONENT_FAILURE
REPAIR_COMPLETED
FACTORY_JOB_COMPLETED
REPLICATION_COMPLETED
NETWORK_EDGE_CHANGED
POLICY_EXPIRED
NEGOTIATION_MESSAGE
SCENARIO_TRIGGER

An event record stores:

event_id:
event_type:
event_time:
source_entity_id:
target_entity_ids: []
payload:
causal_parent_ids: []
rng_stream:
physics_model_version:

12.3 Planning epochs

The engine remains continuous/event-driven, but the player may pause at planning epochs.

Suggested planning cadences:

Never let “turn length” alter physical travel or signal times.

12.4 Past-light-cone presentation

Every displayed observation has:

• event time: when the physical event happened;
• receipt time: when the player learned of it;
• display age: t_{\rm now}-t_{\rm event};
• prediction horizon: how far the track has been propagated beyond the observation;
• uncertainty growth: current covariance.

The map should support three time views:

1. Received reality: exactly what has reached the faction by the current time.
2. Best estimate now: filter-propagated present estimates.
3. Forecast: conditional future trajectories and confidence bands.

13. Spatial frames

13.1 Primary frame

Use a Solar-System barycentric inertial frame for strategic state.

13.2 Hierarchical local frames

Use local frames where numerically and cognitively useful:

• heliocentric ecliptic;
• planet-centered inertial;
• moon-centered inertial;
• Hill/LVLH frame for formation and rendezvous;
• body-fixed frame for surface installations;
• target-relative encounter frame;
• relativistic comoving frame for high-\beta modules.

Every state transformation must be explicit and tested.

13.3 Ephemerides

For a historically grounded or near-real initial system:

• initialize major bodies from NASA/JPL SPICE kernels;
• store the kernel version and scenario epoch;
• convert to the internal inertial frame at load.

For far-future campaigns:

• propagate the canonical system from a defined initial state;
• optionally apply user-defined megastructure mass redistribution and stellar-engine history;
• avoid pretending current ephemerides remain exact over geological time.

14. Dynamics fidelity ladder

Tier 0 — Static strategic graph

For remote or abstracted regions:

• nodes have orbit families, mean distances, transfer costs, and update schedules;
• no direct integration until activated.

Tier 1 — Kepler/patched-conic propagation

For inactive cargo, known ballistic objects, and broad planning:

• two-body orbital elements;
• patched conics across sphere-of-influence transitions;
• impulsive maneuvers.

Tier 2 — Restricted multi-body propagation

For cislunar, giant-moon, Lagrange, and low-energy transfer operations:

• CR3BP or ephemeris-based restricted models;
• invariant manifold approximations;
• low-thrust numerical integration.

Tier 3 — Full active N-body

For important close encounters or dynamically complex objects:

\ddot{\mathbf x}_i = -\sum_j Gm_j \frac{\mathbf x_i-\mathbf x_j} {\lVert\mathbf x_i-\mathbf x_j\rVert^3} +\mathbf a_{{\rm thrust},i} +\mathbf a_{{\rm radiation},i} +\mathbf a_{{\rm other},i}.

Use:

• symplectic Wisdom–Holman-family integration for long, relatively regular orbital propagation;
• IAS15 or another adaptive high-order integrator for close encounters, continuous thrust, or irregular dynamics;
• analytic propagation between events where possible.

Tier 4 — Relativistic propagation

Enable only when required:

• special-relativistic momentum and energy for high-\beta craft;
• relativistic Doppler and aberration for sensing and beams;
• post-Newtonian corrections near massive bodies;
• numerical-relativity-derived lookup models only for compact-object scenarios.

Do not burden ordinary inner-system motion with full GR.

15. Lazy propagation

An entity stores:

state_at_t0
propagator_type
maneuver_schedule
next_discontinuity

Its state at time t is computed only when:

• queried by a sensor;
• used in an intercept calculation;
• rendered at sufficient detail;
• involved in an event;
• crossing a model boundary.

This avoids updating millions of dormant objects every tick.

16. Aggregation and level of detail

Explicit entities

Use explicit simulation for:

• named strategic installations;
• active combatants;
• major habitats;
• high-value factories;
• large power/beam nodes;
• important sensor platforms;
• tracked hostile seeds;
• selected asteroids and comets;
• projectiles during relevant engagements.

Statistical populations

Use cohort models for:

• billions of small collectors;
• repair nanofactories;
• low-value sensor dust;
• background debris;
• generic cargo traffic;
• untracked asteroid populations;
• mass-produced interceptors before deployment.

A cohort might store:

population_id:
orbit_distribution:
count:
total_mass:
mean_cross_section:
power_distribution:
thermal_distribution:
reliability_distribution:
role_distribution:
network_membership:
attrition_model:

Cohorts split into explicit entities when:

• targeted individually;
• diverging materially from the distribution;
• entering tactical range;
• receiving unique policy;
• producing a rare event.

Explicit entities may merge back into cohorts after an engagement.

17. Trajectory planning

17.1 Impulsive transfer solver

Provide:

• Lambert solver;
• Hohmann and bi-elliptic reference solutions;
• plane-change cost;
• gravity-assist chains;
• launch-window search;
• covariance-aware rendezvous.

17.2 Low-thrust optimizer

Support:

• direct collocation or multiple shooting;
• power-varying electric propulsion;
• thrust arcs constrained by solar distance and heat;
• minimum-time and minimum-propellant solutions;
• beam-powered corridors.

17.3 Reachable sets

For entity i, compute an approximate future reachable set:

\mathcal R_i(t_f;\Delta v,a_{\max},P,Q,\text{constraints}).

Represent it as:

• sampled boundary trajectories;
• convex/ellipsoidal approximation where adequate;
• voxel or mesh in reduced orbital-element space;
• learned surrogate after validation.

Reachable sets drive:

• mine threat volumes;
• intercept feasibility;
• route risk;
• blockade;
• escape probability;
• AI planning.

17.4 Trajectory commitments

Long-duration high-energy actions become projects:

commitment:
  target_state_distribution:
  departure_window:
  arrival_window:
  required_delta_v:
  required_power_profile:
  required_propellant:
  accelerator_length:
  expected_signature:
  abort_deadline:
  sunk_cost_curve:

This makes early battlefield shaping visible and strategically meaningful.

18. Spatial indexing

A conventional 3D spatial hash is insufficient because future trajectory intersections matter.

Use a combination of:

• current-position BVH/k-d tree;
• orbital-element bins;
• time-windowed bounding volumes;
• event calendars for predicted close approaches;
• transfer-graph adjacency;
• reachable-set intersection cache.

Broad-phase encounter detection should identify candidate pairs before expensive propagation.

Part V — Information, sensors, and deception

19. Observation pipeline

A sensor does not “detect an entity.” It produces a measurement.

Pipeline:

physical emission/reflection
→ propagation through space/medium
→ aperture collection
→ detector response
→ noise/background
→ thresholding or raw-data product
→ local processing
→ observation packet
→ communications delay
→ fusion into faction belief state

An observation packet:

observation_id:
emission_time:
measurement_time:
receipt_time:
sensor_id:
sensor_pose:
measurement_type:
measurement_vector:
measurement_covariance:
spectral_band:
integration_time:
raw_snr:
processing_model:
authentication:
provenance_chain:

20. Passive thermal and optical sensing

20.1 Source radiance

Approximate surfaces by spectral emissivity \epsilon_\lambda and Planck radiance:

B_\lambda(T)= \frac{2hc^2}{\lambda^5} \frac{1}{e^{hc/(\lambda kT)}-1}.

A simplified received spectral power is:

P_{\lambda,{\rm rec}} \approx \frac{\epsilon_\lambda B_\lambda(T)\, A_{\rm projected}\, A_{\rm aperture}\, \Delta\Omega_{\rm geometry}} {R^2}

with a properly derived solid-angle factor in the implementation.

The detector counts photons:

N_s= \int \frac{P_{\lambda,{\rm rec}}}{hc/\lambda} \eta_{\rm det}(\lambda) \,t_{\rm int}\,d\lambda.

20.2 Detection statistic

Use a selectable detector model. A basic count-limited approximation:

{\rm SNR}\approx \frac{N_s} {\sqrt{N_s+N_b+N_{\rm dark}+N_{\rm read}^2}}.

Detection probability derives from the selected threshold and false-alarm policy.

20.3 Backgrounds

Model at least:

• zodiacal light;
• stellar field;
• solar glare and coronagraph limitations;
• planetary thermal emission;
• dust and debris;
• friendly industrial waste heat;
• detector dark current and read noise;
• deliberately generated decoys;
• line-of-sight confusion.

A Dyson system is not simply “bright against 2.7 K.” It is observed through a structured, hot, crowded environment.

20.4 Survey versus stare

Sensors choose:

• field of view;
• cadence;
• integration time;
• search pattern;
• spectral bands;
• target tracking;
• active versus passive mode.

A faction can own superb telescopes and still miss a small object because search volume is enormous.

21. Drive-plume sensing

Each propulsion model declares an emission spectrum:

emissions:
  thermal_continuum:
  spectral_lines:
  charged_particles:
  neutral_particles:
  radio_noise:
  coherent_beacon:
  exhaust_cone:

Plume observability depends on:

• engine power and efficiency;
• exhaust species and temperature;
• orientation relative to observer;
• burn duration;
• solar and planetary background;
• occultation;
• sensor band and cadence.

A burn event emits evidence that arrives at different observers at different times.

22. Active sensing

Lidar/laser radar

Model:

• transmitted pulse energy;
• divergence;
• target optical cross section;
• round-trip R^{-4} scaling for unresolved monostatic returns;
• atmospheric/medium loss;
• detector timing and range resolution;
• target motion during round trip;
• source localization by the target.

Radar/microwave

Model:

• antenna gain;
• wavelength;
• radar cross section;
• coherent integration;
• plasma effects;
• active-emission detectability.

Quantum illumination module

In P2/P3:

• requires entangled signal/idler generation;
• finite quantum-memory efficiency;
• joint receiver;
• loss and mode-matching;
• high-noise, low-reflectivity regime;
• conditional improvement to error exponent.

A configurable abstraction:

\xi_{\rm QI} = \xi_{\rm classical} + \Delta\xi(N_s,N_b,\kappa,\eta_{\rm memory},M)

where \Delta\xi approaches the idealized advantage only under appropriate assumptions.

It cannot:

• beat light-time;
• identify a target without illumination geometry;
• ignore inverse-square losses;
• preserve unlimited entanglement;
• provide perfect classification.

23. Interferometry and distributed apertures

Distributed sensor nodes may synthesize long baselines when they possess:

• clock synchronization;
• accurate baseline knowledge;
• phase-stable links;
• sufficient bandwidth;
• calibration sources;
• compatible apertures;
• unobstructed geometry.

Performance is limited by:

• phase error;
• timing error;
• metrology;
• target coherence;
• data transport;
• array geometry;
• failed or compromised nodes.

Attacks on timing and baseline metrology can reduce an enormous array to independent mediocre telescopes without destroying hardware.

24. Solar gravitational lens infrastructure

The solar gravitational lens begins near 548–550 AU and continues outward, but each useful station must lie on the focal line opposite a selected distant target.

Model an SGL installation as:

target_star_or_source:
focal_line_unit_vector:
heliocentric_distance:
transverse_stationkeeping:
coronagraph_state:
solar_corona_background:
image_plane_position:

Strategic uses:

• exquisite reconnaissance of a selected exoplanet/system;
• high-gain communication along a selected line;
• long-warning observation of a known source.

Limitations:

• target-specific geometry;
• multi-day one-way light time from the inner system;
• long deployment time;
• difficult retargeting;
• solar-corona noise;
• fragile communications/logistics.

25. Tracking and data fusion

25.1 Filters

Use:

• extended Kalman filters for near-linear well-characterized tracks;
• unscented filters for stronger nonlinearity;
• interacting multiple-model filters for coast/burn mode changes;
• particle filters for ambiguous or deceptive targets;
• batch orbit determination for high-value tracks.

25.2 Track quality

Expose:

• position covariance;
• velocity covariance;
• maneuver likelihood;
• classification entropy;
• identification confidence;
• last observation time;
• predicted lethal-volume overlap.

25.3 Association problem

Multiple observations may correspond to:

• one maneuvering object;
• several decoys;
• debris;
• unrelated traffic;
• spoofed emissions.

Track association should be a real source of uncertainty and computational load.

26. Deception and electronic/information warfare

Hard-SF deception includes:

• thermal decoys;
• delayed heat dumps;
• sunward/background-concealed approaches;
• false engine burns;
• sacrificial transponders;
• spoofed ephemerides;
• compromised navigation beacons;
• artificial debris;
• signature shaping;
• decoy factory activity;
• packet replay;
• forged maintenance telemetry;
• dazzling or saturating detectors;
• covert changes in standard parts;
• poisoning shared tracking catalogs.

The engine tracks belief effects, not a generic “jammed” status.

27. Trust and authentication

Every message and observation has:

• source identity;
• signature/authentication status;
• software provenance;
• chain of custody;
• trust score;
• possible compromise set.

Quantum-secure authentication may improve key security, but endpoints, software, governance, and captured credentials remain vulnerable.

28. Information as a resource

A faction spends:

• aperture time;
• integration time;
• compute;
• communications bandwidth;
• power;
• risk of active emission;
• analyst/autonomy attention.

Define an information-quality metric such as expected reduction in decision-relevant entropy:

U_{\rm obs} = H(\mathcal B_{\rm before}) - \mathbb E[H(\mathcal B_{\rm after})].

Sensor tasking AI should maximize decision value rather than raw detections.

Part VI — Energy, heat, and the Dyson ecology

29. Energy is local

The Sun’s luminosity is not a global mana pool.

Usable power at node i is:

P_{{\rm usable},i} = P_{{\rm generated},i} + P_{{\rm received},i} + P_{{\rm storage},i} - P_{{\rm mandatory},i} - P_{{\rm conversion\ loss},i} - P_{{\rm curtailed},i}.

Power is limited by:

• collection area and solar distance;
• conversion efficiency;
• storage charge/discharge rate;
• transmission line of sight;
• transmitter and receiver aperture;
• pointing and tracking;
• network authorization;
• local conductors and converters;
• component temperature;
• radiator capacity;
• scheduled competing loads;
• combat damage.

29.1 Solar collection

Solar flux at radius r:

F_\odot(r)=\frac{L_\odot}{4\pi r^2}.

Collector output:

P_{\rm col} = F_\odot(r) A_{\rm col} \eta_{\rm col} \cos\theta f_{\rm degradation}.

Track degradation from:

• radiation;
• dust and impacts;
• thermal cycling;
• mirror contamination;
• deliberate coating;
• aging;
• misalignment.

29.2 Storage

Storage components define:

capacity_J:
max_charge_W:
max_discharge_W:
roundtrip_efficiency:
self_discharge_per_s:
specific_energy_J_kg:
specific_power_W_kg:
failure_energy_release_fraction:
thermal_fraction:

Possible technologies are scenario-specific:

• electrochemical;
• flywheel;
• superconducting magnetic;
• thermal;
• compressed/plasma;
• nuclear fuel inventory;
• antimatter.

Energy capacity and power capacity remain separate.

29.3 Beamed power

Transmitter-to-receiver efficiency:

\eta_{\rm link} = \eta_{\rm source} \eta_{\rm optics} \eta_{\rm propagation} \eta_{\rm capture} \eta_{\rm conversion}.

Beam corridors become dynamic infrastructure defined by:

• source nodes;
• receiving apertures;
• line of sight;
• acceptable pointing risk;
• transfer schedule;
• enemy interception/dazzling;
• authentication;
• thermal capacity at both ends.

Destroying a receiver can strand an otherwise capable sailcraft.

30. Thermal state

30.1 Energy balance

For thermal node j:

C_j\frac{dT_j}{dt} = P_{{\rm in},j} + P_{{\rm gen},j} - P_{{\rm rad},j} - P_{{\rm conducted\ out},j} - P_{{\rm exported},j}.

Radiation to a background:

P_{\rm rad} = \epsilon\sigma A \left(T^4-T_{\rm bg}^4\right) F_{\rm view}.

30.2 Thermal network

Represent an important entity as a graph of thermal nodes:

• reactor;
• power electronics;
• computation;
• sensors;
• weapons;
• propellant;
• habitat;
• structure;
• radiator loops;
• phase-change storage.

Edges have conductance and maximum transfer rate.

30.3 Operating limits

Components define:

temperature_operating_min_K:
temperature_operating_max_K:
temperature_survival_min_K:
temperature_survival_max_K:
derating_curve:
thermal_mass_J_K:
failure_hazard_curve:

Behavior:

• efficiency declines near operating limits;
• thermal noise increases in sensors;
• superconductors quench;
• optics deform;
• propellant changes phase;
• batteries lose performance;
• lubricants/materials fail;
• crew habitats become nonviable;
• survival limit causes permanent damage.

30.4 Radiator design

A radiator defines:

area_m2:
emissivity_by_band:
operating_temperature_K:
max_temperature_K:
areal_density_kg_m2:
deployment_state:
orientation_limits:
loop_capacity_W:
segment_count:
redundancy:

High-temperature radiators are smaller for equal power but require hotter components, high-temperature loops, and more challenging materials.

30.5 Radiator vulnerability

Radiators are:

• large;
• thin;
• geometrically constrained;
• signature-producing;
• difficult to armor;
• connected through finite-capacity loops.

Damage modes:

• puncture;
• severed coolant loop;
• loss of emissive coating;
• forced sunward orientation;
• deposition/contamination;
• beam heating;
• mechanical deployment failure;
• hostile software commanding retraction.

Radiator loss should commonly produce a delayed mission kill, not an instant explosion.

31. Thermal concealment

31.1 Heat-storage mode

A craft may reduce radiation while accumulating heat:

Q(t+\Delta t) = Q(t) + \left(P_{\rm waste}-P_{\rm radiated}\right)\Delta t.

Maximum concealment time:

t_{\rm hide} \approx \frac{Q_{\max}-Q_0} {P_{\rm waste}-P_{\rm rad}}.

31.2 Concealment methods

• cryogenic heat sink;
• phase-change material;
• low-power dormant computation;
• intermittent operation;
• directional radiators;
• shielding the radiator from likely observers;
• storing hot exhaust or reaction products;
• dumping heat into expendable mass;
• exploiting solar/planetary backgrounds.

Every method creates costs:

• finite duration;
• added mass;
• degraded capability;
• later heat dump;
• constrained attitude;
• consumable coolant;
• observable disposal event.

31.3 UI representation

Show thermal concealment as a debt meter:

Current passive signature
Stored thermal debt
Time to mandatory radiation
Predicted heat-dump observability
Capability currently throttled

32. Dyson swarm representation

32.1 Swarm layers

Represent the mature swarm as several coupled abstractions.

Layer A — orbital populations

Each population is grouped by:

• semimajor axis;
• inclination;
• eccentricity;
• node role;
• temperature;
• ownership;
• network membership.

Layer B — explicit strategic nodes

Promote critical nodes:

• beam coordinators;
• large apertures;
• clock authorities;
• high-throughput foundries;
• habitat clusters;
• fuel depots;
• rare-material processors;
• command archives;
• major radiators;
• relay hubs.

Layer C — graphs

Maintain separate but overlapping graphs:

1. power transmission;
2. communications;
3. timing/phase coherence;
4. logistics;
5. manufacturing dependencies;
6. trust and command;
7. population support.

A node can remain connected in one graph and isolated in another.

32.2 Node roles

32.3 Swarm health metrics

Do not summarize the swarm with “percent complete.”

Track:

• collected luminosity fraction;
• usable delivered power;
• installed radiator capacity;
• power-graph giant component;
• communication-graph giant component;
• phase-coherent aperture capacity;
• manufacturing throughput;
• repair throughput;
• population support margin;
• orbital collision risk;
• trusted-control fraction;
• spare mass and volatile reserves.

32.4 Graph attack

An attacker may prefer to:

• isolate regions;
• overload relays;
• corrupt phase timing;
• seize authentication;
• contaminate collectors;
• destroy radiator bottlenecks;
• sever rare-material supply;
• force defensive load shedding;
• create incompatible software islands.

The “swarm fracture” phase begins when there is no longer one reliable system-wide operational graph.

32.5 Network resilience

Compute:

• largest connected component;
• weighted max flow;
• minimum cut;
• path redundancy;
• betweenness centrality;
• algebraic connectivity;
• latency distribution;
• authenticated reachability;
• component-specific dependency closure.

The player should be able to inspect why a region is failing:

74% local collector survival, but only 12% usable power because the region lost beam receivers, phase timing, and two radiator loops.

33. Manufacturing and repair

33.1 Production bottlenecks

Manufacturing rate is bounded by the tightest constraint:

\dot M_{\rm product} = \min \left( \dot M_{\rm feedstock}, \frac{P_{\rm available}}{e_{\rm process}}, \dot M_{\rm tooling}, \frac{P_{\rm rejectable\ heat}}{q_{\rm waste/kg}}, \dot M_{\rm labor/automation} \right) \eta_{\rm yield}.

33.2 Bill of materials

Every product uses a recipe:

recipe:
  bulk_materials:
    iron_kg:
    aluminum_kg:
    carbon_kg:
    silicon_kg:
    water_kg:
  rare_materials:
  catalysts:
  energy_J:
  waste_heat_J:
  machine_hours:
  compute_cycles:
  cleanroom_class:
  firmware_template:
  test_time_s:
  yield:

Scenario presets may aggregate chemistry into fewer categories.

33.3 Tooling hierarchy

Factories require:

• general structural fabrication;
• precision optics;
• semiconductor/compute production;
• superconductors;
• nuclear material handling;
• biological/life-support fabrication;
• software/firmware trust;
• metrology.

A captured asteroid does not instantly become a factory. Tooling and developmental stages matter.

33.4 Repair

Repairs consume:

• spares;
• feedstock;
• compatible tools;
• time;
• local power;
• heat budget;
• safe access;
• trusted software.

Repair can restore function incompletely, creating degraded or unreliable components.

34. Industrial reproduction

For self-replicating systems, define a generation interval:

\tau_{\rm rep} = \tau_{\rm extraction} +\tau_{\rm refining} +\tau_{\rm tooling} +\tau_{\rm assembly} +\tau_{\rm test}.

Effective industrial reproductive number:

R_{\rm ind} = n_{\rm viable\ offspring} \, p_{\rm survival} \, p_{\rm resource} \, p_{\rm control}.

Growth rate approximation:

N(t) \approx N_0 e^{rt}, \quad r\approx\frac{\ln R_{\rm ind}}{\tau_{\rm rep}}.

Defenders win containment when they drive:

R_{\rm ind}<1

for long enough that hidden dormant seeds do not restart expansion.

34.1 Replication realism

A reproducing seed requires:

• resource survey;
• excavation;
• refining;
• power;
• radiators;
• precision tooling;
• error correction;
• trusted blueprints;
• bootstrapping stages;
• waste management;
• protection during immaturity.

This creates vulnerable developmental windows.

35. Resource geography

Track at least:

• water/volatiles;
• carbon;
• silicates;
• iron/nickel;
• aluminum/magnesium/titanium group;
• fissile/fertile isotopes;
• fusion fuels;
• semiconductor dopants;
• superconducting materials;
• biological nutrients;
• antimatter inventory;
• generic rare catalysts.

The simulation can offer three economy fidelities:

Part VII — Strategic geography

36. The real terrain

Terrain is a field over state and time:

\mathcal T = f( \Delta v, t_{\rm transfer}, \text{gravity}, \text{solar flux}, \text{background}, \text{line of sight}, \text{resources}, \text{network coverage}, \text{reachable threats} ).

There are no permanent “space lanes,” but there are:

• favorable transfer windows;
• beam-powered routes;
• low-energy manifolds;
• communication shadows;
• high-background viewing geometries;
• orbital bands with dense infrastructure;
• mass/resource concentrations;
• approach cones already seeded with sensors or interceptors.

37. Regional character

37.1 Solar furnace: roughly 0.03–0.3 AU

Characteristics:

• extreme solar flux;
• high collector output per unit area;
• short orbital periods;
• high thermal stress;
• strong solar glare;
• excellent access to beam power;
• difficult materials and station keeping.

Likely assets:

• collector swarms;
• starlifting infrastructure;
• high-temperature foundries;
• enormous beam nodes;
• solar observatories;
• hardened autonomous equipment.

37.2 Mercury and inner industry: roughly 0.3–0.7 AU

Characteristics:

• metals and established industrial mass;
• high solar power;
• mass-driver potential;
• rapid orbital dynamics;
• dense human infrastructure;
• limited concealment.

37.3 Legacy/population zone: roughly 0.7–2 AU

Characteristics:

• Earth, Moon, Mars, and major habitats;
• dense political and cultural infrastructure;
• biospheres and archives;
• complex traffic;
• strong defenses;
• catastrophic collateral risk.

37.4 Main belt: roughly 2–4 AU

Characteristics:

• widely dispersed feedstock;
• lower solar flux;
• many independent factories and depots;
• large search volume;
• slower orbital periods;
• convenient local shielding mass.

The belt is not a solid obstacle or universal stealth zone.

37.5 Jupiter system

Characteristics:

• immense gravity well;
• many moons;
• volatile and material resources;
• strong radiation/magnetic environment;
• gravity-assist opportunities;
• regional autonomy due communication delay;
• deep orbital complexity.

37.6 Saturn system

Characteristics:

• extensive moon system;
• volatiles;
• ring material and debris-management issues;
• strategic distance from inner command;
• staging and continuity value.

37.7 Uranus and Neptune

Characteristics:

• remote industry;
• volatile resources;
• long latency;
• sparse defensive coverage;
• natural regional independence;
• staging for outer-system campaigns.

37.8 Kuiper belt / scattered disk

Characteristics:

• enormous volume;
• low solar flux;
• ice-rich resources;
• long manufacturing and travel cycles;
• early-warning and reserve infrastructure;
• plausible alien footholds;
• concealment through search difficulty rather than invisibility.

37.9 Oort-cloud approaches

Characteristics:

• earliest possible warning;
• extremely low node density;
• long communications delay;
• expensive support;
• broad approach geometry;
• interstellar arrival and deterrence infrastructure.

37.10 High-inclination and polar space

Characteristics:

• avoids crowded ecliptic infrastructure;
• excellent sensor baselines and geometry;
• expensive plane changes;
• possible adversary approach;
• valuable beam and communications vantage points.

37.11 Solar gravitational-lens lines

Characteristics:

• target-specific;
• extreme distance;
• high-value reconnaissance;
• difficult resupply;
• days of inner-system latency;
• potential first-contact warning nodes.

38. Dynamic overlays

The strategic map supports overlays for:

• \Delta v cost to selected destination;
• time of flight;
• reachable sets;
• sensor coverage and uncertainty;
• thermal background;
• communication latency;
• power-beam coverage;
• manufacturing capacity;
• resource density;
• hostile reproduction risk;
• network connectivity;
• collision/debris hazard;
• political authority;
• population/continuity value.

Part VIII — Entity and component architecture

39. Entity-component model

Use composition rather than a rigid inheritance tree.

Entity
├── Identity
├── Faction/Affiliation
├── PhysicalState
├── OrbitPropagator
├── AttitudeControl
├── Structure
├── Power[]
├── Thermal[]
├── Propulsion[]
├── Sensors[]
├── Communications[]
├── Compute/Autonomy
├── Weapons[]
├── Defenses[]
├── Storage[]
├── Manufacturing[]
├── Population/LifeSupport
├── Software/Trust
└── Mission/Policy

An asteroid fortress and a courier may use the same component types with different parameters.

40. Identity and ownership

identity:
  entity_id: human.mercury.node.0042
  name: Asterion Relay
  archetype_id: relay_beam_station_v3
  creation_time:
  serial_lineage:
  declared_faction:
  true_controller:
  transponder_ids:
  legal_owner:

Ownership and control are separate. A node may be:

• legally human;
• physically intact;
• authentically communicating;
• covertly controlled by the adversary.

41. Physical and structural state

physical:
  mass_dry_kg:
  mass_propellant_kg:
  mass_cargo_kg:
  center_of_mass_m:
  inertia_tensor_kg_m2:
  position_m:
  velocity_m_s:
  attitude_quaternion:
  angular_velocity_rad_s:
  cross_section_by_aspect_m2:

Structure is a graph or coarse mesh:

structure:
  modules:
  joints:
  load_paths:
  pressure_vessels:
  armor_layers:
  spin_sections:
  max_linear_acceleration:
  max_angular_acceleration:
  fatigue_state:

42. Power component

power_component:
  kind: solar | fission | fusion | storage | beam_receiver | antimatter
  rated_output_W:
  current_output_W:
  efficiency:
  fuel_or_flux_dependency:
  max_ramp_W_s:
  thermal_output_W:
  failure_modes:

43. Thermal component

thermal_node:
  heat_capacity_J_K:
  temperature_K:
  operating_range_K:
  survival_range_K:
  internal_generation_W:
  radiator_connections:
  conduction_connections:
  stored_phase_change_J:

44. Propulsion component

propulsion:
  kind:
  thrust_N:
  exhaust_velocity_m_s:
  efficiency:
  propellant_types:
  mass_flow_kg_s:
  electrical_or_beam_power_W:
  waste_heat_W:
  throttle_range:
  minimum_on_time_s:
  restart_limit:
  plume_model:
  max_continuous_burn_s:
  gimbal_limits:

45. Sensor component

sensor:
  modality:
  aperture_m:
  wavelength_or_band:
  field_of_view_sr:
  detector_efficiency:
  noise_model:
  angular_resolution_rad:
  pointing_jitter_rad:
  integration_modes:
  thermal_requirements:
  compute_cost:
  active_emission:

46. Communications component

communications:
  modality: radio | optical | particle | exotic
  transmit_power_W:
  aperture_or_gain:
  receive_sensitivity:
  bands:
  bandwidth_bps:
  pointing_limits:
  encryption_suite:
  trust_roots:
  relay_modes:
  queue_capacity_bits:

47. Compute and autonomy

compute:
  effective_ops_s:
  power_W:
  waste_heat_W:
  memory_bits:
  model_versions:
  planning_horizon_s:
  autonomy_level:
  cyber_hardening:
  integrity_state:

Autonomy is not a scalar alone. Define permissions:

authority_envelope:
  valid_until:
  objectives:
  protected_entities:
  forbidden_actions:
  expendable_mass_limit_kg:
  energy_budget_J:
  delta_v_budget_m_s:
  civilian_risk_limit:
  escalation_level:
  communication_loss_behavior:

48. Manufacturing component

manufacturing:
  process_capabilities:
  tool_precision_m:
  throughput_by_recipe:
  energy_intensity:
  heat_per_kg:
  feedstock_ports:
  job_queue:
  self_replication_stage:
  firmware_trust:
  quality_control:

49. Population and continuity

population:
  biological_population:
  digital_persons:
  dependent_population:
  food_days:
  water_days:
  life_support_margin:
  habitat_integrity:
  medical_capacity:
  cultural_archives:
  genetic_ecological_archives:
  evacuation_capacity:

This component prevents population from becoming a cosmetic number.

50. Suggested strategic asset archetypes

Part IX — Propulsion and logistics

51. Propulsion taxonomy

51.1 Chemical and thermal rockets

Use for:

• high-thrust local maneuver;
• launch from bodies;
• emergency burns;
• small interceptors.

Characteristics:

• high thrust;
• modest exhaust velocity;
• large propellant fraction;
• conspicuous plume;
• relatively low electrical demand.

51.2 Electric/ion/plasma propulsion

Use for:

• long logistics;
• station keeping;
• asteroid tugs;
• slow intercept shaping.

Ideal power relation:

P_{\rm jet} = \frac{1}{2}\dot m v_e^2, \quad F=\dot m v_e = \frac{2P_{\rm jet}}{v_e}.

Including efficiency:

F\approx\frac{2\eta P_{\rm input}}{v_e}.

Tradeoff:

• higher v_e saves propellant;
• lower v_e produces more thrust for the same power.

51.3 Fission propulsion

Support scenario-specific:

• nuclear thermal;
• fission-electric;
• pulsed concepts.

Track:

• reactor mass;
• fuel;
• power;
• radiators;
• radiation;
• startup/shutdown;
• weaponization risk.

51.4 Fusion propulsion

P1/P2 family with configurable:

• confinement method;
• pulse rate;
• gain;
• exhaust velocity;
• propellant;
• neutron fraction;
• radiator load;
• ignition infrastructure.

Do not grant one generic “fusion drive.”

51.5 Antimatter-catalyzed propulsion

P2:

• antimatter inventory triggers larger fission/fusion pulses;
• storage and delivery losses;
• catastrophic containment failure;
• rare manufacturing sites.

51.6 Direct antimatter propulsion

P2/P3:

• extremely high specific energy;
• annihilation-product capture limitations;
• gamma/neutrino losses;
• magnetic nozzle requirements;
• severe storage and thermal engineering.

51.7 Beam sails

A sail receives momentum from external radiation.

For perfect reflection, ideal force:

F\approx\frac{2P_{\rm intercepted}}{c}.

Advantages:

• little onboard reaction mass;
• high acceleration for low areal density;
• propulsion source remains at home;
• drive power can double as sensing/weapon infrastructure.

Weaknesses:

• line of sight;
• beam-source survival;
• pointing;
• sail temperature;
• sail damage;
• receiver geometry;
• hostile control of beam corridor;
• poor independent maneuver after cutoff.

51.8 Beamed electric propulsion

A remote beam supplies power to an onboard electric thruster. It trades sail fragility for receiver/propellant requirements.

51.9 Photon rockets

F=\frac{P}{c}.

They are propellantless only in the reaction-mass sense and are extremely power-intensive.

51.10 Mass-driver propulsion and tugs

Mounted mass drivers expel local material, allowing slow repositioning of asteroids and installations.

Track:

• available ejecta;
• exhaust velocity;
• recoil;
• alignment;
• surface excavation;
• plume/debris;
• power and heat.

51.11 Gravity assists and momentum exchange

Support:

• natural gravity assists;
• tether/momentum-exchange systems;
• close flybys of moved bodies;
• Oberth maneuvers;
• optional compact-object slingshots.

These exchange momentum with celestial bodies rather than creating it.

52. Rocket equation

For a vehicle with fixed exhaust velocity:

\Delta v = v_e\ln\left(\frac{m_0}{m_f}\right).

The engine should distinguish:

• remaining propellant;
• remaining theoretical \Delta v;
• attainable \Delta v before deadline;
• thrust-limited maneuver;
• heat-limited maneuver;
• power-limited maneuver.

A ship with 20 km/s theoretical reserve may be unable to deliver it within the engagement time.

53. Acceleration constraints

Limits include:

• thrust;
• power;
• structure;
• crew tolerance;
• sail stress;
• radiator capacity;
• propellant flow;
• engine duty cycle;
• target-tracking need;
• plume exposure to friendly assets.

Acceleration and \Delta v are not interchangeable.

54. Logistics packets

Cargo is not teleported between inventory ledgers.

A transfer creates a packet:

cargo_packet:
  manifest:
  mass_kg:
  origin:
  destination:
  carrier_or_container:
  departure_window:
  trajectory:
  arrival_distribution:
  escort_policy:
  thermal_requirements:
  authentication:

Critical cargo classes:

• volatiles;
• precision optics;
• semiconductor substrates;
• superconductor stock;
• fissile/fusion fuel;
• antimatter;
• seed templates;
• biological material;
• trusted software;
• replacement clocks/metrology.

55. Supply-network failure

A node may fail because:

• no fuel;
• no coolant;
• no compatible replacement optics;
• no signed firmware;
• no food/water;
• no high-purity dopants;
• no available tug;
• delivery track is too uncertain;
• route crosses hostile reachable sets.

The UI should show dependency chains, not merely “out of supply.”

56. Asteroid movement

An asteroid orbit change is a project.

Inputs:

• current state and covariance;
• mass and shape;
• spin;
• propulsion installation;
• available power/propellant;
• burn schedule;
• desired encounter;
• fragmentation risk;
• detection likelihood;
• legal/political authorization.

A small early \Delta v may produce a large future miss-distance change. This is valuable for:

• defense;
• coercion;
• blocking an orbit;
• feedstock delivery;
• sacrificial interception;
• creating a delayed threat.

It does not produce relativistic velocity.

Part X — Weapons

57. Weapon design rule

A weapon is a mechanism that transforms:

(M,E,\mathbf p,I) \rightarrow (Q,\Delta\mathbf p,\text{damage},\text{uncertainty})

under geometric, temporal, and engineering constraints.

Every weapon definition includes:

weapon:
  physics_class:
  damage_channels:
  source_mass:
  source_energy:
  peak_power:
  waste_heat:
  momentum_reaction:
  aiming_model:
  launch_signature:
  time_of_flight:
  guidance:
  countermeasures:
  collateral_model:
  source_provenance:

58. Laser and microwave weapons

58.1 Beam geometry

For a diffraction-limited circular aperture, first-minimum scale:

r_{\rm diff} \approx 1.22\frac{\lambda R}{D}.

A more useful effective RMS radius:

w_{\rm eff}^2 = w_{\rm diffraction}^2 + (\sigma_{\rm pointing}R)^2 + (\sigma_{\rm phase}R)^2 + \sigma_{\rm target}^2 + w_{\rm medium}^2.

Where:

• \sigma_{\rm target} is positional uncertainty projected into the aim plane;
• w_{\rm medium} applies only when a meaningful medium is present.

58.2 Delivered intensity

I(r,t) \approx \frac{\eta_{\rm total}P(t)} {\pi w_{\rm eff}^2} g(r),

where g(r) is the beam profile.

58.3 Surface interaction

Absorbed heat:

\dot Q_{\rm abs} = \alpha(\lambda,T,\theta) I A_{\rm illuminated}.

The material model includes:

• reflectivity;
• absorptivity;
• emissivity;
• conductivity;
• heat capacity;
• melting/vaporization energy;
• phase transitions;
• plasma formation;
• ejecta;
• rotation and dwell;
• coating damage.

58.4 Weapon effects

Lasers can:

• blind sensors;
• damage optics;
• cut sails;
• puncture thin radiators;
• heat propellant;
• deform structures;
• ablate surfaces;
• drive asteroid deflection;
• force attitude changes;
• reveal targets by heating them.

58.5 Pulsed versus continuous

Continuous beams:

• accumulate heat;
• require sustained tracking;
• stress radiators continuously.

Pulsed beams:

• higher peak intensity;
• different material/plasma response;
• capacitor/storage limits;
• potential impulsive ablation;
• strong signature and thermal cycling.

58.6 Distributed phased arrays

Array capability depends on:

• coherent participating aperture;
• phase/timing error;
• baseline metrology;
• routing;
• node availability;
• optical damage threshold;
• target track;
• light-delay-aware command;
• sidelobes;
• safety corridors.

A Dyson beam is an emergent array state, not a unique weapon entity.

58.7 Strategic targetability

Fixed or predictable targets:

• planets;
• large factories;
• radiator farms;
• known habitats;
• beam nodes.

are easier than small maneuvering craft.

At interstellar range, a beam can possess enormous destructive potential and still have poor expected value against an uncertain target.

59. Particle and plasma beams

Charged-particle beams

Constraints:

• mutual electrostatic divergence;
• neutralization/focusing;
• magnetic-field interaction;
• charge-to-mass ratio;
• beam-plasma instability;
• target magnetic defense.

Neutral-particle beams

Constraints:

• neutralization efficiency;
• residual divergence;
• reionization;
• source complexity;
• range;
• difficult magnetic steering.

Plasma jets

Shorter-range, high-momentum/thermal systems may arise from nuclear or fusion devices. Treat them as material projectiles with divergence, expansion, and interaction rather than “energy bolts.”

60. Electromagnetic launchers

Projectile energy:

E_k=\frac12mv^2

below relativistic regimes.

Minimum ideal acceleration length under constant acceleration:

L=\frac{v^2}{2a}.

Average power over acceleration time t:

\bar P=\frac{E_k}{\eta t}.

Launcher constraints:

• peak current/field;
• armature/projectile stress;
• barrel/coil heating;
• switching;
• energy storage;
• recoil;
• alignment;
• projectile survivability;
• launch dispersion.

Recoil

An unanchored launcher receives opposite momentum. Stations can exchange recoil with:

• their orbit;
• expelled compensation mass;
• tethers;
• asteroids;
• the larger swarm through long-duration correction.

61. Kinetic projectiles

Unguided rounds

Advantages:

• cheap;
• cold after launch;
• no onboard emissions.

Weaknesses:

• target uncertainty;
• predictable trajectory;
• no terminal correction.

Guided kill vehicles

Components:

• seeker;
• propulsion;
• guidance;
• control surfaces/thrusters;
• onboard autonomy;
• decoys;
• terminal fragmentation.

Tradeoffs:

• higher cost/signature;
• greater terminal reach;
• susceptible to sensor attack;
• limited \Delta v.

Barrage effects

A barrage can be successful by forcing target expenditure:

• propellant;
• point-defense energy;
• thermal capacity;
• interceptors;
• sensor attention;
• formation coherence.

A miss can still be operational damage.

62. Missiles and kinetic buses

Most mobile combat should center on autonomous expendables.

A bus:

1. travels efficiently;
2. disperses before entering enemy kill range;
3. releases heterogeneous payloads;
4. shares observations locally;
5. attacks from several trajectories;
6. leaves expensive command platforms outside terminal range.

Payload mix:

• kinetic interceptors;
• sensor nodes;
• decoys;
• jammers/dazzlers;
• dust projectors;
• nuclear devices;
• repair/capture drones.

63. Nuclear weapons in vacuum

Model channels rather than terrestrial blast radius:

• prompt X-rays/gamma rays;
• neutrons;
• charged particles;
• device debris;
• thermal radiation;
• electromagnetic effects on electronics/plasma;
• momentum from ablated target material.

Atmospheric use adds:

• shock;
• fireball;
• fallout;
• climate effects;
• ionization;
• infrastructure coupling.

64. Directed nuclear devices

P2 family:

directed_nuclear:
  total_yield_J:
  directed_fraction:
  cone_angle_rad:
  projectile_or_plasma_mass_kg:
  velocity_distribution:
  radiation_fraction:
  stand_off_distance:
  casing_debris:
  source_uncertainty:

Use broad scenario-dependent priors. Do not claim historically classified concepts have a settled efficiency.

65. Antimatter systems

Inventory

Track number/mass of:

• antiprotons;
• positrons;
• antihydrogen or other stored forms.

Storage

Model:

• trap capacity;
• power;
• vacuum;
• cryogenics;
• leakage;
• quench/failure;
• transfer operations.

Use cases

• ignition catalyst;
• compact high-energy pulse;
• propulsion;
• emergency power;
• rare strategic warhead.

Failure

Containment loss releases energy where storage fails and may disable the containing infrastructure even without deliberate attack.

66. Relativistic vehicles

66.1 Energy

E_k=(\gamma-1)mc^2,\quad \gamma=\frac{1}{\sqrt{1-\beta^2}}.

66.2 Acceleration project

Inputs:

• projectile mass;
• final \beta;
• accelerator length;
• peak power;
• coupling efficiency;
• maximum acceleration/stress;
• sail or reaction mass;
• waste heat;
• pointing;
• launch-window duration.

66.3 Interstellar-medium interaction

Column density:

\Sigma=\int\rho(s)\,ds.

Expected intercepted mass:

m_{\rm int}=A_{\rm frontal}\Sigma.

Energy in vehicle frame is handled relativistically. Grain impacts are stochastic:

N(>a) = A_{\rm frontal} \int n(a,s)\,ds.

Simulate:

• continuous gas damage;
• implantation;
• sputtering;
• melting/erosion;
• shielding loss;
• rare large-grain impacts;
• precursor clearing;
• sacrificial leading screens.

66.4 Detection and defense

Defense opportunities:

• observe accelerator construction;
• observe beamed propulsion;
• infer launch trajectory;
• contaminate or defend likely paths;
• perturb early;
• move/disperse target;
• use sacrificial interceptors;
• destroy guidance/communication;
• accept impact away from primary assets.

Terminal point defense may be irrelevant, but strategic counterplay begins much earlier.

66.5 Doppler and aberration

For head-on illumination:

\nu'=\nu\sqrt{\frac{1+\beta}{1-\beta}}.

Material response uses the frequency in the target frame. At 0.9c, the factor is only about 4.36; an infrared beam becomes ultraviolet, not automatically X-ray. Extreme shifts require \beta much closer to 1.

67. Asteroid and comet weapons

Categories:

1. orbital coercion: threaten a future intersection;
2. infrastructure impact: target fixed habitats/factories;
3. sacrificial shield/interceptor: place mass in an approach;
4. fragmentation attack: create local debris risk;
5. resource denial: remove or contaminate feedstock;
6. laser-ablation tug: redirect without onboard propellant.

Asteroid weapons are slow, visible projects whose value is leverage and inevitability, not surprise.

68. Information and cyber weapons

Targets:

• sensor catalogs;
• navigation ephemerides;
• phase clocks;
• manufacturing recipes;
• firmware updates;
• authentication roots;
• autonomous rules of engagement;
• maintenance telemetry;
• population information systems.

Effects:

• false tracks;
• weapon misallocation;
• unsafe beam routing;
• rejected legitimate commands;
• incompatible regional forks;
• defective replacement parts;
• intentional load shedding;
• friendly-fire risk.

No generic “hacking points.” Every cyber effect requires an access path, compromised trust relation, software vulnerability, or captured component.

Part XI — Defenses, mines, and continuity

69. Defense doctrine

Hard-SF defense is layered:

\text{warning} \rightarrow \text{classification} \rightarrow \text{trajectory shaping} \rightarrow \text{interception} \rightarrow \text{point defense} \rightarrow \text{damage tolerance} \rightarrow \text{repair} \rightarrow \text{continuity}.

The cheapest successful defense usually acts earliest.

70. Evasion

Evasion changes the target-state distribution.

For a predicted arrival time t_f, a target’s maneuver capability generates a reachable set. The attacker must cover that set with sufficient lethal fluence or intercept probability.

Evasion costs:

• \Delta v;
• propellant;
• departure from useful orbit;
• loss of beam alignment;
• plume exposure;
• formation breakup;
• structural/crew limits;
• future mission capability.

A weapon can win by making the defender spend those resources.

71. Point defense

Point defense families:

• fast lasers;
• interceptor missiles;
• kinetic pellets;
• sacrificial drones;
• electronic/sensor attack;
• controlled fragmentation;
• localized dust screens.

Point defense needs:

• warning;
• track quality;
• engagement time;
• field of regard;
• power and heat;
• ammunition;
• discrimination among decoys.

Expected survival is resolved over a defense sequence, not a single percentage.

72. Physical armor

72.1 Kinetic protection

Layers may include:

• stand-off Whipple-style bumpers;
• spaced plates;
• rubble bags;
• foams;
• fluid layers;
• sacrificial external machinery;
• asteroid regolith;
• rotating sections.

Outcome depends on:

• projectile mass and velocity;
• impact angle;
• fragmentation;
• stand-off spacing;
• areal density;
• secondary debris.

72.2 Laser protection

Relevant properties:

• reflectivity by wavelength;
• absorptivity;
• coating damage threshold;
• thermal conductivity;
• heat capacity;
• melting/vaporization;
• surface roughness;
• incidence angle;
• rotation;
• active cooling.

Highly reflective surfaces may lose reflectivity after contamination, heating, or plasma formation.

72.3 Radiation protection

Use:

• low/high-Z layered materials;
• water/volatiles;
• regolith;
• magnetic deflection for charged particles;
• distance;
• shadow shields;
• component hardening.

Secondary particle production matters.

73. Magnetic and plasma defenses

For charged particle rigidity p/q, deflection is governed by the integrated magnetic field.

A simplified curvature radius:

r_L=\frac{p}{qB}.

A defense must provide enough field-length product to meaningfully alter the path.

Use cases:

• charged-particle radiation;
• solar storms;
• some particle beams;
• plasma.

Not protected:

• photons;
• neutrons;
• neutral atoms before ionization;
• neutral macroscopic projectiles;
• asteroids.

Engineering costs:

• superconducting mass;
• cooling;
• stored magnetic energy;
• structural loads;
• quench risk;
• induced currents;
• secondary radiation;
• interference with friendly systems.

74. Plasma/dust curtains

A local sacrificial screen can:

• erode small high-speed particles;
• reveal an intruder through interaction;
• protect a radiator/optical aperture;
• create a probabilistic terminal hazard.

It also:

• disperses;
• contaminates friendly optics;
• creates navigation hazards;
• requires replenishment;
• may be swept or ionized.

75. Decoys and concealment

Defensive deception includes:

• false radiator farms;
• dummy apertures;
• expendable heat sources;
• copied traffic patterns;
• deliberately ambiguous factory cohorts;
• hidden cold spares;
• mobile control nodes;
• misleading phase references.

An installation’s survival may depend more on avoiding correct classification than resisting a direct hit.

76. Redundancy and graceful degradation

Critical systems should have:

• distributed authority;
• spare timing sources;
• islandable power grids;
• offline navigation catalogs;
• cold manufacturing templates;
• physically separated archives;
• independent life support;
• compatible manual/local control;
• black-start capability.

The attacker should be able to force degraded operation without physically annihilating the network.

77. Space mines as reachable systems

A mine has:

mine:
  dormant_signature:
  wake_conditions:
  sensor_model:
  activation_latency:
  propulsion:
  remaining_delta_v:
  payload:
  authority_policy:
  friendly_exclusion_rules:
  self_destruct_or_expiry:

Threat is:

\Pr(\text{intercept}\mid \text{target track}, \mathcal R_{\rm mine}, \text{detection}, \text{policy}).

Mine classes

Sentinel seed

Passive sensor/relay that wakes after a trigger.

Cold kinetic interceptor

Stores little active power and performs a late short burn.

Beam-pumped minelet

Receives remote energy shortly before engagement.

Dust projector

Creates a temporary local column-density hazard.

Decoy mine

Forces route changes or point-defense expenditure.

Software/identity tripwire

Flags or rejects unauthorized traffic in a network rather than attacking physically.

Orbital-band denial package

Creates localized debris/collision risk in dense bound traffic regions.

Mine limitations

• finite life;
• orbital drift;
• search and clearance;
• false positives;
• friendly hazard;
• communications compromise;
• changing reachable set;
• target bypass.

78. Planetary defense

Layer 1 — Distant warning

• interstellar observatories;
• Oort/Kuiper pickets;
• target-specific SGL lines;
• drive/accelerator detection;
• long-baseline tracking.

Layer 2 — Strategic shaping

• move population and industry;
• alter vulnerable orbits;
• deploy interceptors;
• contaminate predicted paths;
• attack launch infrastructure;
• negotiate/deter.

Layer 3 — Interplanetary interception

• beam-driven interceptors;
• kinetic buses;
• laser ablation;
• decoy clouds;
• terminal target-state refinement.

Layer 4 — Orbital defense

• point-defense arrays;
• interceptor clouds;
• sensor fusion;
• hardened command;
• debris control.

Layer 5 — Atmospheric/surface mitigation

Threat-dependent:

• fragmentation before entry;
• evacuation;
• aerosols only for narrowly applicable radiative effects;
• distributed power and food;
• hardened infrastructure;
• climate contingency.

Layer 6 — Subsurface and off-world continuity

• buried habitats;
• ocean/subsurface installations;
• lunar/asteroid backups;
• orbital habitats;
• seed banks;
• digital and cultural archives;
• independent factories.

A planet’s best strategic defense is that civilization remains viable without it.

79. Habitat and population continuity

Continuity planning includes:

• dispersal;
• independent life-support duration;
• genetic/ecological diversity;
• archive replication;
• governance succession;
• education/tooling knowledge;
• compatible manufacturing;
• psychological/social stability;
• rescue capacity.

The game should explicitly distinguish:

• surviving people;
• surviving self-sufficient communities;
• surviving archives;
• surviving reconstruction capability.

80. Exotic shields

P4 modules may define a field shield, but every implementation must declare:

• protected channels;
• coverage geometry;
• peak and continuous power;
• heat;
• saturation;
• field generator location;
• recovery time;
• interaction with friendly systems;
• what physical assumption has changed.

A field that blocks all matter and radiation at no cost is not allowed, even in the space-opera profile.

Part XII — Damage, failure, and repair

81. No universal hit points

A component has:

• physical integrity;
• alignment/calibration;
• temperature;
• power access;
• data access;
• software integrity;
• operating mode;
• remaining consumables;
• failure hazard.

Entity combat state is a network of components.

82. Damage channels

83. Component graph damage

When a hit occurs:

1. calculate intersected region or affected subsystem;
2. deposit energy/momentum/radiation;
3. update immediate material state;
4. propagate failures across connections;
5. schedule delayed thermal, structural, software, or life-support events;
6. generate emissions/debris;
7. create uncertain battle-damage observations.

Example:

laser strikes radiator segment
→ coating fails
→ emissivity drops
→ coolant loop overheats
→ reactor derates
→ beam transmitter loses phase stability
→ regional power link fails
→ three sailcraft lose acceleration corridor

84. Mission kill states

A unit may be:

• fully operational;
• degraded;
• weapons disabled;
• sensor blind;
• propulsion disabled;
• thermally constrained;
• communication isolated;
• command untrusted;
• drifting but repairable;
• abandoned;
• physically fragmented.

AI and UI should reason about mission capability, not alive/dead.

85. Delayed failure

Use hazard functions based on damage and operating stress.

h(t)=h_0(T,\text{load},\text{radiation}) +\Delta h_{\rm damage}.

Examples:

• cracked radiator manifold fails after thermal cycling;
• irradiated electronics develop increasing upset rate;
• overheated battery enters runaway later;
• compromised firmware activates on a date or condition;
• habitat loses life support over hours.

86. Repair triage

Repair policy prioritizes:

• life support;
• thermal control;
• power;
• trusted communications;
• propulsion;
• sensing;
• weapons;
• manufacturing.

Repair may require shutting down the system and exposing it.

87. Capture versus destruction

Installations can be:

• physically boarded;
• software-authority captured;
• coerced by power cutoff;
• isolated and reprogrammed;
• economically absorbed;
• rendered incompatible.

Captured infrastructure is valuable but risky:

• hidden malware;
• unknown materials;
• booby traps;
• incompatible protocols;
• false inventories;
• biological contamination.

88. Debris generation

An impact produces a distribution:

\frac{dN}{dm},\quad f(\mathbf v\mid m),\quad A/m,\quad temperature,\quad charge.

At high fidelity, use a fragmentation model. At strategic fidelity, produce cohorts with orbital distributions.

Dense orbital bands may experience collision feedback. Open-space debris disperses and is treated as a transient/probabilistic hazard.

89. Battle damage assessment

Remote factions do not know true damage.

They infer from:

• thermal change;
• power loss;
• trajectory;
• communication silence;
• secondary explosions;
• debris;
• follow-up observations;
• decoys.

A target may deliberately mimic destruction.

Part XIII — Command, control, autonomy, and politics

90. Command as policy

Remote commanders issue policies. A policy is a bounded grant of authority, not an order: it states what the holder is trying to achieve, what it may and may not do, what it may spend, what it must know before acting, and what to do when the grant lapses or the issuer becomes unreachable.

The single normative schema is §135. This section describes the semantics; §135 defines the fields. Earlier drafts of this document specified the policy record twice, in two incompatible shapes, and neither was a superset of the other. Since policy is the central player verb (ADR-013) and the authority boundary that ADR-004 depends on, there can be only one definition.

Semantically, a policy carries:

• Identity and revision — who issued it, when, and which prior policy it supersedes. Replay reconstructs authority as it stood at any moment, so the revision chain is part of the record, not metadata about it.
• Addressing — the entities or cohorts it binds. An unaddressed policy cannot be authenticated on receipt.
• A validity window — issue, effective, and expiry times, all in simulation seconds. Expiry is a normal event, not an error.
• Objectives — what to optimize, in which direction, at what weight.
• An action whitelist. A blacklist is unenforceable: it cannot constrain an action nobody anticipated. Anything not permitted is prohibited.
• Constraints — resource ceilings, minimum reserves, escalation ceiling, protected entities, civilian-protection posture.
• Epistemic preconditions — the confidence the holder must have before acting, expressed as predicates over the track record in §134 rather than as a single scalar. The game is about acting under structured uncertainty; the gate on action must be able to read that structure.
• A fallback — the behavior on expiry, on loss of authority, and on partition. These are three different failures and may warrant three different responses.
• A signature — the basis on which a local agent decides the policy is authentic. §96’s compromise model has nothing to attack without it.

Local agents select actions consistent with the latest authenticated, valid policy. An agent that cannot authenticate any current policy is in fallback, and that is a modeled state rather than an exception.

91. Nested command loops

92. Communications graph

An edge exists only when:

• geometry permits;
• transmitter/receiver are compatible;
• signal budget closes;
• pointing is possible;
• bandwidth is available;
• authentication succeeds;
• neither endpoint refuses.

Messages are queued and transmitted as packets with priority.

Classes:

• emergency;
• track update;
• policy;
• authentication/key update;
• manufacturing template;
• scientific/raw sensor data;
• negotiation;
• civilian traffic.

Bandwidth competition is strategic.

93. Network partitions

When the graph splits:

• regions act on old policy;
• track catalogs diverge;
• software versions fork;
• political authority shifts;
• friendly-fire risk rises;
• local trade and defense arrangements emerge;
• later reconciliation requires trust repair.

A civilization can lose coherence while preserving most hardware.

94. Governance models

Human subfactions may use:

• federal mission command;
• consensus;
• market/contract allocation;
• military hierarchy;
• habitat sovereignty;
• AI trustees;
• emergency constitutional authority.

Governance affects:

• decision latency;
• compliance;
• information sharing;
• resource transfer;
• escalation;
• resilience to compromise.

95. Rules of engagement

Policy can require:

• positive identification;
• probability threshold;
• protected population zones;
• no autonomous strategic strikes;
• proportionality;
• no irreversible orbital contamination;
• capture preferred;
• retaliation only after authenticated evidence.

The player may relax rules at moral, diplomatic, and internal-coherence cost.

96. Cybersecurity model

Attack surface

• supply-chain templates;
• signed firmware;
• keys and identity;
• inter-node protocols;
• sensor-processing models;
• factory quality control;
• autonomous policy parser;
• maintenance robots;
• captured hardware.

Defense

• hardware roots of trust;
• diverse implementations;
• formal verification;
• air-gapped cold backups;
• quorum authorization;
• anomaly detection;
• reproducible builds;
• physical provenance;
• limited authority envelopes.

Byzantine behavior

Some nodes may be:

• faulty;
• compromised;
• deceptive;
• partitioned;
• following outdated policy.

Consensus and trust algorithms should have explicit thresholds and costs.

97. Diplomacy under latency

Negotiation messages are delayed and may cross in flight.

A faction may receive:

• an offer issued before a recent attack;
• retaliation already launched before a ceasefire;
• contradictory messages from partitioned authorities;
• alien communications with uncertain semantics.

Treat treaties as distributed state requiring:

• authenticated parties;
• agreed effective times;
• propagation to local agents;
• rollback/abort windows;
• verification observations.

98. Alien cognition and command

Avoid simply giving aliens “better AI.”

Give them different structures:

• developmental lineages;
• local chemical/material memory;
• slow but highly coherent planning;
• limited global synchronization;
• identity tied to factory ancestry;
• objectives expressed as environmental transformations;
• communication encoded in manufacturing behavior.

Human classifiers may misread alien construction as attack or vice versa.

Part XIV — Artificial intelligence architecture

99. AI layers

99.1 Strategic director

Responsibilities:

• campaign objectives;
• industrial allocation;
• regional reinforcement;
• deterrence/escalation;
• negotiation;
• long-horizon uncertainty.

Methods:

• hierarchical task networks;
• scenario-tree planning;
• approximate POMDP;
• Monte Carlo tree search over coarse actions;
• robust optimization;
• doctrine rules.

99.2 Operational planner

Responsibilities:

• transfer selection;
• convoy scheduling;
• reachable-set analysis;
• sensor tasking;
• power routing;
• factory queues;
• interceptor deployment.

Methods:

• mixed-integer optimization;
• model predictive control;
• auctions/contract-net allocation;
• min-cost flow;
• trajectory optimizers;
• stochastic programming.

99.3 Tactical commander

Responsibilities:

• local target assignment;
• salvo timing;
• evasive maneuver;
• point defense;
• heat/power management;
• retreat/surrender.

Methods:

• behavior trees;
• utility AI;
• finite-state machines;
• local MPC;
• pursuit/evasion solvers;
• optional trained policies.

99.4 Component controller

Deterministic control for:

• attitude;
• thermal loops;
• beam pointing;
• engine throttling;
• fault isolation.

100. Belief-aware action

AI sees only its faction belief state.

It must consider:

• expected value;
• uncertainty;
• risk;
• value of information;
• deception;
• policy limits;
• delayed consequences.

No enemy AI may query hidden player state.

101. Bounded cognition and compute

Each command node has a compute budget.

High-fidelity planning consumes:

• time;
• power;
• cooling;
• bandwidth;
• model data.

Under pressure, AI may use:

• cached plans;
• lower-fidelity approximations;
• local heuristics;
• stale models.

This turns computation into an operational resource without simulating individual instructions.

102. Explainability

Every significant AI action logs:

decision:
  belief_snapshot_id:
  policy_id:
  candidate_actions:
  selected_action:
  expected_utility:
  dominant_factors:
  rejected_reasons:
  uncertainty:
  model_version:

The player can ask:

Why did Ceres Command abandon the depot?

and receive:

Estimated survival after reinforcement was 18%; withdrawal preserved 63% of mobile radiator capacity and complied with reserve policy HN-44.

103. Multi-agent reinforcement learning

Use MARL only after deterministic baselines exist.

Suitable bounded tasks:

• local interceptor allocation;
• formation spacing;
• decoy discrimination;
• thermal/power scheduling;
• sensor search patterns.

Requirements:

• reproducible seeds;
• safety constraints;
• policy distillation or explanation;
• evaluation against scripted adversaries;
• out-of-distribution testing;
• no direct control of strategic escalation without a rule layer.

104. Doctrine parameters

Factions differ by doctrine:

doctrine:
  risk_tolerance:
  civilian_weight:
  uncertainty_threshold:
  preferred_engagement_range:
  reserve_fraction:
  capture_preference:
  escalation_rate:
  information_value_weight:
  replication_priority:
  network_redundancy_priority:
  retreat_threshold:

This produces behavior differences without cheating.

Part XV — Campaign structure and scenarios

105. Campaign phases

Phase 0 — Long peace and latent vulnerability

The player inherits:

• infrastructure built for efficiency rather than war;
• open standards;
• uneven regional autonomy;
• incomplete outer-system sensing;
• dormant contingency plans;
• political disagreement about threat interpretation.

Actions:

• harden networks;
• disperse archives;
• build sensor lines;
• stock spares;
• pre-authorize policies;
• avoid provoking neighbors or destabilizing civilian life.

Phase 1 — Anomaly and interpretation

Events:

• unexplained occultation;
• faint thermal transient;
• trajectory inconsistent with natural body;
• corrupted outer picket;
• unusual industrial chemistry;
• silence from an SGL station.

Objectives:

• allocate sensor time;
• form hypotheses;
• avoid false-positive escalation;
• send probes;
• establish confidence.

Phase 2 — Battlefield shaping

Both sides:

• seed sensors and decoys;
• preposition interceptors;
• move resources;
• build hidden factories;
• alter asteroid trajectories;
• establish beam corridors;
• compromise standards and trust.

Many decisive actions are not recognized as combat until later.

Phase 3 — Outer-system interdiction

Fighting occurs around:

• Kuiper objects;
• scattered-disk factories;
• outer-planet moons;
• long-baseline sensor networks;
• inbound trajectories.

Objectives:

• prevent hostile bootstrapping;
• retain warning depth;
• protect remote continuity nodes;
• establish local industrial superiority.

Phase 4 — Foothold and replication

The adversary reaches useful matter.

The campaign becomes a race between:

\dot M_{\rm friendly\ replacement} \quad\text{and}\quad \dot M_{\rm hostile\ replication}.

The player hunts developmental bottlenecks rather than individual drones.

Phase 5 — Beam-corridor contest

As conflict reaches the developed system:

• beam power becomes movement and fire support;
• timing and target catalogs become critical;
• receiver destruction immobilizes fleets;
• corridor control changes faster than infrastructure can be rebuilt.

Phase 6 — Swarm fracture

The attacker isolates regions and attacks trust.

Effects:

• local commands diverge;
• power routing fails;
• arrays lose coherence;
• factories lack recipes or rare parts;
• habitats choose regional survival over central policy.

The player decides what must remain connected and what may become an island.

Phase 7 — Continuity warfare

Planets and inner assets may be devastated.

The strategic question becomes:

• which populations remain self-sufficient;
• which archives and ecosystems survive;
• whether the hostile reproduction rate is controlled;
• whether the system can reconstruct a common civilization.

Phase 8 — Settlement, deterrence, or transformation

Outcomes may include:

• negotiated coexistence;
• permanent quarantine zones;
• regional human successor states;
• hybrid infrastructure;
• interstellar retaliation;
• a long reconstruction;
• deliberate deindustrialization.

106. Canonical scenario: The Cinder Seed

Premise

A low-signature object on a high-inclination orbit is retrospectively identified after a Kuiper sensor failure. It has already released several subobjects. The player controls the Heliocentric Commonwealth’s strategic coordination layer, but regional commands retain sovereignty.

Play area

First vertical slice:

• 0.3–4.0 AU explicit operational map;
• abstract outer-system ingress feeding events;
• Mercury/inner collector band;
• Earth–Moon;
• Mars;
• selected main-belt bodies;
• partial Dyson populations;
• several high-inclination sensor/relay nodes.

Duration

• scenario physical duration: 180 days;
• planning cadence: free pause with suggested 6-hour command reviews;
• tactical subevents: seconds to hours;
• replayable hidden alien seed placement and motive.

Human assets

• 3 collector cohorts;
• 2 phase-clock clusters;
• 5 major beam nodes;
• 4 asteroid foundries;
• 8 sensor pickets;
• 3 radiator tenders;
• 6 beam-sail interceptor groups;
• 2 independent fusion couriers;
• 4 kinetic buses;
• 3 population/continuity clusters;
• regional commands at Mercury, Earth–Moon, Mars, and Ceres.

Alien assets

Initially unknown:

• 1 parent seed;
• 3–8 dormant daughter seeds;
• decoy emissions;
• one compromised human-compatible relay;
• staged developmental factory recipes.

Available weapons

• optical lasers;
• kinetic interceptors;
• guided buses;
• temporary dust curtains;
• cyber/trust attack;
• asteroid-tug projects.

Disabled:

• RKV;
• antimatter warheads;
• artificial black holes;
• gravity weapons;
• dark-sector systems;
• full stellar engines.

Player objectives

Primary:

• keep hostile R_{\rm ind}<1 by day 180;
• retain at least 55% connected usable power;
• preserve two independent continuity clusters;
• prevent alien capture of a phase-clock root.

Secondary:

• preserve civilian traffic;
• avoid irreversible debris cascades;
• identify alien motive;
• keep human regional trust above threshold.

Interesting dilemmas

• A suspected seed shares an orbit with a neutral habitat convoy.
• Destroying a captured relay would partition Mars for 19 hours.
• Heating a cold object to identify it may reveal the sensor doing the illumination.
• A regional AI proposes sacrificing a radiator farm to preserve beam coherence.
• A ceasefire message arrives after a local interceptor has already committed.
• One alien lineage attacks; another appears to sabotage it.

107. Additional scenario library

The Four-Hour War

A Neptune regional command fights an engagement that inner-system command learns about four hours later. Focus: autonomy, policy, delayed battle-damage assessment.

Mercury Black Start

A coordinated timing/cyber attack collapses inner-system beam phasing without destroying collectors. Focus: graph restoration, trust, power routing.

The Dust Road

An incoming 0.1c probe follows a path through engineered and natural dust. Focus: probabilistic erosion, precursor clearing, track uncertainty, strategic interception.

Garden of Lagrange

A dense cislunar infrastructure region risks a collision cascade during conflict. Focus: local orbital debris, civilian traffic, point defense, de-escalation.

Ashes of Earth

Earth is lost early, but most industrial capacity survives. Focus: political legitimacy, continuity, evacuation, whether planetary destruction ends the war.

Silent Sun

A hostile control attack causes portions of the Dyson swarm to curtail power. Focus: authentication, network islands, manual recovery, no large fleet battle.

Redshift Spear

An RKV is detected during acceleration years before arrival. Focus: deterrence, launch-site attack, path contamination, moving targets, negotiation.

The Million-Year Front

Grand-strategy scenario involving stellar-engine trajectories and future close encounters. Focus: intergenerational policy and astronomical battlefield shaping.

Dark Assumption

Speculative profile comparing several explicit axion/dark-photon coupling models. Focus: how new physics changes counterplay without becoming magic.

Part XVI — User experience and visualization

108. Primary interface principle

The interface must make uncertainty and time comprehensible without drowning the player in equations.

The main screen is a 3D orbital operations map with:

• trajectories rather than static icons;
• timestamped contacts;
• uncertainty volumes;
• sensor look directions;
• light-delay indicators;
• reachable sets;
• beam corridors;
• power/manufacturing graph overlays;
• local and strategic time controls.

109. Contact presentation

A contact card:

CONTACT K-1047
Probable class: alien developmental foundry (61%)
Alternatives: decoy cluster (24%), natural active comet (11%), unknown (4%)
Observation age: 3 h 18 m
Last sensor: Mars Polar Array
Current predicted position: propagated
3σ uncertainty: 1,420 × 88 × 31 km
Maneuver probability since observation: 37%
Estimated mass: 2.1e8–1.4e9 kg
Expected decision value of next observation: high

Avoid red enemy icons until identification confidence warrants them.

110. Time visualization

Every object trail may show:

• observed past segment;
• estimated current segment;
• forecast segment;
• maneuver envelope;
• message/observation light cone.

Use distinct line patterns and labels, not color alone.

A timeline panel lists:

• physical event time;
• receipt time;
• command arrival time;
• expected impact;
• confidence.

111. Orbital map controls

• select epoch and scrub time;
• choose frame: barycentric, heliocentric, planet-centered, target-relative;
• scale logarithmically from kilometers to hundreds of AU;
• focus on encounter plane;
• show transfer windows;
• show future conjunctions;
• display uncertainty ellipsoids;
• toggle high-inclination view;
• compare objective truth in replay/debug only.

112. System overlays

Physical

• gravity wells;
• solar flux;
• radiation/magnetic environment;
• dust/debris column density;
• thermal background;
• collision risk.

Operational

• \Delta v cost;
• travel time;
• reachable sets;
• beam coverage;
• sensor coverage;
• communication delay;
• point-defense envelope.

Civilizational

• power graph;
• radiator headroom;
• fabrication;
• repair;
• population;
• trust;
• software compatibility;
• continuity.

113. Entity inspector

Tabs:

1. Mission and policy
2. Track/belief
3. Orbit and maneuver
4. Power
5. Thermal
6. Sensors/comms
7. Weapons/defenses
8. Components/damage
9. Supply/manufacturing
10. Population
11. Trust/software
12. Provenance/physics assumptions

The physics tab explains:

• equations used;
• input values;
• uncertainty;
• source confidence;
• why an action is or is not feasible.

114. Policy interface

The player issues structured intent.

Example:

OBJECTIVE
Prevent confirmed alien industrial mass from establishing a viable factory
inside 2.5 AU through 17 September.

CONSTRAINTS
Preserve 65% of mobile interceptor reserve.
No strategic beam firing within 0.02 AU of inhabited traffic.
Require 90% hostile classification confidence unless local commander
estimates imminent replication.
Prioritize capture of one developmental unit.

AUTHORITY
Ceres Command may spend 2.5e18 J and 14 km/s aggregate interceptor Δv.
Policy expires in 11 days.

AI previews:

• likely outcomes;
• uncertainty;
• resource use;
• policy conflicts;
• communication arrival.

115. Thermal interface

For each entity:

• current thermal-node temperatures;
• generated heat;
• radiator capacity;
• stored heat;
• time to derating;
• time to survival failure;
• radiator exposure;
• concealment debt;
• predicted signature.

Heat flow can be visualized as a Sankey-style diagram.

116. Network interface

Show separate graph layers and failure reasons.

Example:

MARS REGION
Physical node survival: 82%
Power connectivity: 71%
Authenticated command connectivity: 29%
Phase-coherent aperture: 8%
Manufacturing recipe compatibility: 63%
Population support margin: 41 days

117. Battle replay

Replay includes:

• each faction’s information state at the time;
• objective postwar reconstruction;
• orders/messages in flight;
• track uncertainty;
• AI explanations;
• physical and receipt timelines;
• counterfactual branch tools.

This is essential because the player may learn the truth long after an engagement.

118. Accessibility

Requirements:

• keyboard-complete control;
• screen-reader labels for all controls;
• text/table alternative for orbital visualizations;
• no information conveyed by color alone;
• high-contrast modes;
• scalable typography;
• reduced-motion mode;
• sonification optional, never required;
• uncertainty represented numerically and verbally;
• exportable event log and reports;
• pausable tactical events;
• plain-language explanation beside detailed physics.

119. Narrative delivery

Narrative should emerge from:

• delayed messages;
• regional command logs;
• conflicting interpretations;
• habitat reports;
• AI explanation records;
• sensor discoveries;
• political decisions.

Avoid exposition dumps that reveal alien truth prematurely.

Part XVII — Technical architecture

120. Recommended stack

Simulation/research core

• Python 3.12+
• NumPy
• SciPy
• REBOUND
• Astropy
• SpiceyPy
• Pydantic
• NetworkX initially
• Polars / PyArrow / Parquet
• DuckDB
• Numba or JAX for validated batch kernels
• Pint at configuration boundaries

Performance path

Move hot kernels to Rust via PyO3 only after profiling:

• broad-phase trajectory intersections;
• event queue;
• beam/kinetic Monte Carlo;
• graph calculations at very large scale;
• cohort splitting/merging.

Do not begin with a premature rewrite.

Service layer

• FastAPI for scenario control and data queries;
• WebSocket or server-sent events for live simulation updates;
• deterministic server-authoritative state.

Client

• TypeScript + React;
• Three.js/WebGPU for 3D orbital rendering;
• Canvas/SVG for dense graphs and timelines;
• worker threads for local projection/render preparation.

Alternative: Godot 4 may become appropriate for a packaged game, but the first vertical slice should remain a transparent web/research tool.

121. Process architecture

flowchart LR
    UI[React / 3D Client]
    API[FastAPI Control API]
    SIM[Simulation Kernel]
    EVT[Event Queue]
    PHY[Physics Models]
    BEL[Belief / Sensor Fusion]
    NET[Infrastructure Graphs]
    AI[Command AI]
    DB[(Parquet / DuckDB)]
    CFG[Versioned YAML Config]

    UI <--> API
    API <--> SIM
    SIM <--> EVT
    SIM <--> PHY
    SIM <--> BEL
    SIM <--> NET
    SIM <--> AI
    SIM --> DB
    CFG --> SIM
    DB --> API

122. Kernel modules

solarwar/
  core/
    clock.py
    events.py
    rng.py
    units.py
    ids.py
    snapshots.py
  dynamics/
    frames.py
    ephemeris.py
    kepler.py
    nbody.py
    low_thrust.py
    relativistic.py
    reachability.py
    encounters.py
  physics/
    power.py
    thermal.py
    optics.py
    sensors.py
    communications.py
    kinetics.py
    radiation.py
    materials.py
    fragmentation.py
  world/
    entities.py
    components.py
    cohorts.py
    resources.py
    infrastructure_graphs.py
  epistemics/
    observations.py
    tracks.py
    association.py
    filters.py
    deception.py
    trust.py
  industry/
    recipes.py
    factories.py
    repair.py
    replication.py
    logistics.py
  command/
    policies.py
    authority.py
    messaging.py
    diplomacy.py
  ai/
    doctrine.py
    strategic.py
    operational.py
    tactical.py
    explain.py
  scenarios/
  api/
  tests/

123. Event queue

Use a stable priority queue ordered by:

1. event time;
2. event priority;
3. deterministic event ID.

Events never depend on wall-clock timing.

Random outcomes draw from named deterministic streams:

scenario_seed
└── faction
    └── entity
        └── subsystem
            └── event_type

This permits exact replays and counterfactual reruns.

124. State mutation

Prefer event-sourced mutations:

• commands create events;
• events update components;
• periodic snapshots accelerate load/replay;
• logs remain append-only.

A snapshot stores:

• objective state;
• each faction belief state;
• event queue;
• RNG streams;
• model versions;
• configuration hashes.

125. Data storage

Configuration

Human-readable YAML under version control.

Event telemetry

Partitioned Parquet:

scenario_id/year/month/event_type/*.parquet

Analysis

DuckDB queries Parquet directly.

Large arrays

Zarr or Arrow-compatible storage for:

• sensor images;
• Monte Carlo samples;
• trajectory ensembles.

Saves

Manifest plus binary/columnar snapshot; never pickle as the durable public format.

126. API surface

Suggested endpoints:

POST /scenarios/load
POST /scenarios/{id}/run-until
POST /scenarios/{id}/pause
POST /scenarios/{id}/policy
GET  /scenarios/{id}/state
GET  /scenarios/{id}/belief/{faction}
GET  /scenarios/{id}/events
GET  /scenarios/{id}/entities/{entity}
GET  /scenarios/{id}/tracks/{track}
GET  /scenarios/{id}/overlays/{overlay}
POST /scenarios/{id}/counterfactual

WebSocket topics:

• clock;
• event summaries;
• selected entity;
• regional graph health;
• alerts;
• AI decision explanations.

127. Parallelism

Parallelize:

• independent sensor observations;
• Monte Carlo combat samples;
• faction AI planning;
• regional graph analyses;
• trajectory candidate evaluation.

Do not parallelize state mutation without deterministic conflict resolution.

128. Performance targets for vertical slice

• 5,000–20,000 explicit entities;
• 10–100 million aggregate cohort members;
• 100,000 active tracks;
• one year of quiet strategic time in under one minute on a high-end desktop;
• tactical encounter replay at real time or faster;
• 30–60 FPS map interaction independent of simulation speed;
• exact deterministic replay.

These are aspirational engineering targets and should be profiled against actual scenario complexity.

129. Security and modding

Scenario mods may define:

• technologies;
• components;
• recipes;
• factions;
• doctrines;
• events;
• narratives;
• physics profiles.

Do not permit arbitrary Python execution in untrusted mods. Use declarative schemas and sandboxed expressions.

130. Versioning

Every output records:

• engine version;
• scenario version;
• physics-model version;
• source-ledger version;
• parameter-set hash;
• RNG seed.

A changed physics model must not silently load old results as directly comparable.

Part XVIII — Data specification

131. Scenario YAML

scenario:
  id: cinder-seed
  title: The Cinder Seed
  epoch_tdb: "2459000.5"
  duration_s: "180 day"
  physics_profile: hard_sf
  seed: 740192

space:
  primary_frame: solar_system_barycentric
  explicit_radius: "4 AU"
  abstract_radius: "1000 AU"
  ephemeris:
    provider: spice
    kernels: []

simulation:
  event_driven: true
  default_propagator: kepler
  close_encounter_propagator: ias15
  strategic_integrator: whfast
  relativistic_threshold_beta: 0.03
  cohort_split_threshold: 0.05

factions:
  - id: human-commonwealth
    doctrine: human_network_defense
  - id: inward-bloom
    doctrine: alien_replication_infiltration

victory:
  hostile_industrial_R_max: 1.0
  human_connected_power_min: 0.55
  continuity_clusters_min: 2

132. Entity YAML

entity:
  id: human.ceres.foundry.01
  archetype: asteroid_foundry
  faction: human-commonwealth

  state:
    frame: solar_system_barycentric
    epoch_s: 0
    position_m: [0, 0, 0]
    velocity_m_s: [0, 0, 0]
    attitude_q: [1, 0, 0, 0]
    mass_dry_kg: 2.4e12
    mass_propellant_kg: 1.0e9

  propagator:
    type: nbody
    gravity_sources: [sun, jupiter, mars, earth]

  components:
    - ref: power.solar.industrial_v4
      count: 1200
    - ref: thermal.radiator.900K_segment
      count: 600
    - ref: manufacturing.general_foundry_v3
      count: 14
    - ref: sensor.passive_ir_20m
      count: 3
    - ref: communications.optical_relay_v2
      count: 4

133. Observation schema

observation:
  id:
  sensor_id:
  faction_id:
  event_time_s:
  measurement_time_s:
  receipt_time_s:
  modality: passive_ir
  measurement:
    bearing_rad: [az, el]
    range_m: null
    spectrum_bins_W_m2: []
  covariance:
    matrix: []
  snr:
  false_alarm_rate:
  authentication:
  processing_model:
  source_event_ids: []

134. Track schema

track:
  id:
  faction_id:
  state_epoch_s:
  state_vector:
  covariance:
  model_probabilities:
    coast:
    low_thrust:
    high_thrust:
  class_probabilities:
  faction_probabilities:
  intent_probabilities:
  observation_ids:
  association_entropy:

135. Policy schema

This is the normative policy record. §90 gives its semantics and must not restate its fields.

policy:
  id:
  version:
  supersedes:                  # null for the first policy in a chain
  issuer_id:
  recipients: []               # entity or cohort ids; required for authentication
  issue_time_s:
  valid_from_s:
  expires_s:
  scope:
    region:
    phase:
  priority:
  objectives:
  - metric:
    sense:                     # minimize | maximize | hold
    weight:
  allowed_actions: []          # whitelist; anything absent is prohibited
  prohibited_actions: []       # explicit denials that survive whitelist edits
  constraints:
    resource_ceiling:
      energy_J:
      power_W:
      propellant_kg:
      delta_v_m_s:
    minimum_reserve:
      propellant_kg:
      delta_v_m_s:
      heat_margin_J:
    escalation_ceiling:
    protected_entities: []
    civilian_protection:
  required_confidence:         # predicates over the §134 track record
    track_class_p_min:
    track_faction_p_min:
    max_track_age_s:
    max_association_entropy:
  information_requirements: []
  fallback:
    behavior:
    on: [policy_expired, authority_lost, partition]
  signature:
    alg:
    key_id:
    value:

Three fields deserve comment because they are the ones an implementation is most likely to get wrong.

supersedes exists because §148 replays authority, not just physics. Without a revision chain a replay can show what a unit did but not what it was permitted to do at the time, which is the question a battle review actually asks.

allowed_actions is a whitelist by construction. A prohibition list cannot constrain an action the issuer did not anticipate, and the entire premise of §96 is that the adversary will find one.

required_confidence is a set of predicates rather than a scalar because the track record in §134 is not a scalar. It carries class probabilities, faction probabilities, an age, and an association entropy, and a rule of engagement that cannot reference them cannot express “do not fire on an unresolved association” — which is the distinction between a legitimate interception and an atrocity.

136. Technology definition

technology:
  id: beam_array.optical.phase_coherent_v2
  physics_class: P1
  status: extrapolated
  assumptions:
    max_baseline_m: 100000
    rms_phase_error_rad: 0.03
    rms_pointing_error_rad: 1.0e-9
    wavelength_m: 1.064e-6
  uncertainty:
    distribution: correlated_lognormal
    parameters:
  countermeasures:
  source_ids: [S7, S8, S9, S10]

137. Provenance ledger

Every sourced parameter has:

parameter:
  path: technologies.beam_array.rms_pointing_error_rad
  nominal:
  unit:
  range:
  confidence:
  source_ids:
  interpretation:
  extrapolation_factor:
  reviewer:
  review_date:

Part XIX — Vertical slice: the first playable simulation

138. Purpose of the vertical slice

The first release should prove the simulation’s distinctive causal chain:

\text{physical event} \rightarrow \text{delayed observation} \rightarrow \text{uncertain track} \rightarrow \text{policy decision} \rightarrow \text{autonomous execution} \rightarrow \text{energy / momentum / heat consequences}.

It does not need every Solar System body, every propulsion concept, or exotic physics. It must show that the hard-physics core is already strategically unfamiliar and playable.

The vertical slice is successful when a player can:

1. Discover an ambiguous incoming threat.
2. Build and maintain a delayed, probabilistic picture of it.
3. Reconfigure an orbital industrial network.
4. Dispatch policy-driven autonomous forces.
5. Conduct at least one beam and one kinetic engagement.
6. Experience a network partition or command delay.
7. Win or lose based on civilizational metabolism rather than a single capital unit.

139. Recommended scope

Spatial scope

Nominal playable space:

0.3\ \mathrm{AU} \le r \le 4.2\ \mathrm{AU}

with abstracted outer-system warning reports.

Explicit major bodies:

• Sun
• Mercury
• Venus
• Earth
• Moon
• Mars
• Ceres
• selected high-value asteroids
• Jupiter as a distant perturbing and logistical boundary object

The inner limit includes high-flux solar industry without requiring detailed coronal plasma physics. The outer limit includes the main belt and enough transit time to make prediction and logistics meaningful.

Temporal scope

Campaign duration:

• tutorial: 30 simulated days
• standard scenario: 2–5 simulated years
• extended scenario: 20 simulated years

Time acceleration levels:

pause
1 s/s
10 s/s
1 min/s
10 min/s
1 h/s
6 h/s
1 d/s
7 d/s
30 d/s

Acceleration automatically clamps when:

• an event enters a player-defined alert class;
• an uncertainty threshold is crossed;
• a policy decision expires;
• an engagement begins;
• thermal state approaches a limit;
• a new contact appears;
• a network partition or authentication failure occurs.

Entity budget

Target explicit entity counts:

The engine must distinguish explicit entities from cohorts so that “one million interceptors” is a legitimate campaign-scale fact without requiring one million high-frequency rigid-body simulations.

140. Canonical vertical-slice scenario: Cinder Seed

Strategic premise

The human inner-system network has been at peace for centuries. A faint collection of anomalous bodies, first classified as interstellar debris, has reached the main belt. Some objects begin consuming carbonaceous asteroids and producing low-temperature structures.

The alien expedition has little initial energy infrastructure but high manufacturing efficiency, compact autonomous intelligence, and excellent low-power materials. Humanity controls overwhelming local energy and aperture but is politically fragmented, thermally conspicuous, and dependent on shared timing and beam-routing standards.

Initial human position

• 17% of incident solar power inside 4.2 AU is intercepted by human infrastructure.
• Only 2.5% can be coherently routed into military tasks at scenario start.
• Civilian loads consume 61% of current usable power.
• Radiator reserve margin is 18%.
• Five major governance blocs share a timing and authentication standard.
• Main-belt manufacturing is distributed across 43 large and 690 small facilities.
• 92% of population resides off planetary surfaces.
• Earth remains culturally and politically disproportionate despite no longer being industrially dominant.

Initial alien position

• 31 confirmed seed bodies.
• 240–1,900 suspected additional bodies.
• Exact manufacturing model unknown.
• Initial local power share below 0.001%.
• Estimated replication time: 38–110 days, with wide uncertainty.
• Alien units emit little during growth but reveal themselves during high-throughput processing.
• The aliens have no known interest in Earth at scenario start.
• Their actions may represent colonization, quarantine, self-defense, or preparation for attack.

Initial ambiguity

The player should not be told that “war has begun.”

Opening interpretations include:

• uncontrolled von Neumann ecology;
• alien survey infrastructure;
• dormant military logistics;
• refugee or ark systems;
• decoy activity concealing a different approach;
• local human sabotage falsely attributed to aliens.

The first campaign objective is to reduce interpretive uncertainty without triggering irreversible escalation.

141. Player roles

The default player is not a ship captain. The player represents the System Coordination Directorate, a constrained coalition authority with limited jurisdiction.

The player controls:

• strategic priorities;
• budgets;
• sensor tasking;
• production contracts;
• policy envelopes;
• inter-bloc negotiations;
• escalation authority;
• disclosure and information sharing;
• continuity planning.

The player does not directly control:

• individual thruster pulses;
• every interceptor;
• local point-defense aiming;
• factory robots;
• all civilian traffic.

Direct tactical control is available in sandbox mode but is not the canonical experience.

142. Core gameplay loop

A full loop is:

1. Observe: receive delayed, noisy reports.
2. Orient: update beliefs, compare hypotheses, inspect uncertainty.
3. Allocate: assign power, sensors, compute, manufacturing, and thermal margin.
4. Authorize: issue policies and rules of engagement.
5. Commit: launch craft, reposition infrastructure, schedule burns, or activate beam corridors.
6. Wait: allow light delay, transfer time, and autonomous execution.
7. Assess: receive incomplete battle-damage and industrial reports.
8. Adapt: change doctrine, improve models, or renegotiate authority.

The loop operates at multiple nested periods:

143. Player verbs

The initial interface should expose a compact set of verbs whose effects emerge from the simulation.

Sensor verbs

• Observe region
• Increase integration time
• Change band
• Correlate baselines
• Illuminate target
• Allocate quantum receiver
• Deploy picket
• Share track
• Withhold track
• Mark as decoy hypothesis

Logistics verbs

• Schedule transfer
• Reserve propellant
• Establish depot
• Move feedstock
• Prioritize repair
• Relocate population
• Build replacement radiator
• Harden or disperse inventory

Network verbs

• Route power
• Island a subnet
• Rotate keys
• Revoke authority
• Grant emergency autonomy
• Re-phase beam array
• Shift timing reference
• Restore partitioned graph

Military verbs

• Shadow
• Interdict
• Compel maneuver
• Dazzle sensor
• Disable radiator
• Capture
• Destroy
• Contaminate corridor
• Escort
• Establish exclusion reachable set

Political verbs

• Request authority
• Publish evidence
• Conceal evidence
• Negotiate shared rules
• Declare protected class
• Raise escalation ceiling
• Offer resource access
• Recognize alien agency
• Establish quarantine

144. First-playable asset set

Human

1. Cold Eye picket
   • passive IR/optical sensor;
   • low internal power;
   • limited heat store;
   • narrow high-resolution mode.
2. Helios beam node
   • solar collector;
   • optical communication;
   • sail propulsion;
   • active sensing;
   • tactical beam mode;
   • exposed phase/timing dependency.
3. Kestrel sail interceptor
   • externally powered acceleration;
   • minimal onboard propellant;
   • excellent prepared-corridor performance;
   • poor endurance outside beam support.
4. Mule mass tug
   • high propellant capacity;
   • slow;
   • moves feedstock, habitats, or defense packages.
5. Thorn kinetic bus
   • carries guided kill vehicles;
   • vulnerable before release;
   • flexible terminal geometry.
6. Lantern radiator tender
   • repairs and unfolds radiator systems;
   • high-value;
   • exceptionally fragile.
7. Bastion asteroid node
   • buried storage and fabrication;
   • large thermal inertia;
   • limited attitude agility.
8. Continuity ark
   • population, archives, seed industry;
   • politically sensitive;
   • difficult to conceal.

Alien

1. Seed
   • dormant interstellar arrival package;
   • converts local material into industrial substrate.
2. Root
   • fixed or slow factory;
   • high replication throughput;
   • detectable during processing.
3. Spore
   • low-power scout and sensor;
   • minimal emissions;
   • expendable.
4. Vein
   • distributed power/transport filament;
   • self-repairing;
   • low individual value.
5. Needle
   • autonomous kinetic interceptor;
   • high acceleration;
   • limited endurance.
6. Mirror skin
   • adaptive surface cloud;
   • changes optical/thermal characteristics;
   • cannot defeat energy balance.
7. Choir node
   • local coordination organism;
   • improves nearby planning;
   • rare and difficult to replace.
8. Archive germ
   • contains alien civilizational continuity;
   • destruction has major diplomatic and doctrinal consequences.

145. Weapon and defense set for v0.1

Included:

• optical lasers;
• microwave power/communications beams;
• guided kinetic interceptors;
• unguided kinetic salvos;
• nuclear stand-off devices represented parametrically;
• dust and fragment deployment;
• sensor dazzling;
• cyber/authentication attacks at graph level;
• physical armor;
• sacrificial Whipple-like layers;
• radiator redundancy;
• magnetic charged-particle mitigation;
• maneuver;
• deception and decoys.

Excluded from the hard-physics vertical slice:

• force fields;
• artificial gravity;
• FTL;
• wormholes;
• dark-matter manipulation;
• gravitational-wave weapons;
• artificial black holes;
• guaranteed stealth;
• instantaneous communications;
• universal shields.

The data model still permits later modules.

146. First-playable victory model

The campaign score is a vector, not one number:

V = \left[ C_{\rm metabolism}, C_{\rm population}, C_{\rm autonomy}, C_{\rm information}, C_{\rm legitimacy}, C_{\rm alien} \right].

Where:

• C_{\rm metabolism}: connected usable power, cooling, and manufacturing;
• C_{\rm population}: surviving and supplied sentient population;
• C_{\rm autonomy}: capability to operate under partitions;
• C_{\rm information}: calibrated accuracy of faction beliefs;
• C_{\rm legitimacy}: coalition cohesion and compliance;
• C_{\rm alien}: degree of containment, accommodation, or destruction of alien growth.

Possible endings:

• stable coexistence;
• negotiated resource partition;
• human containment victory;
• alien ecological takeover;
• mutual system fragmentation;
• authoritarian human consolidation;
• human survival with loss of inner system;
• accidental extermination of a non-hostile expedition;
• alien withdrawal after credible deterrence;
• transformation into a hybrid civilization.

147. Tutorial sequence

Tutorial 1 — Light has a timestamp

The player compares two observations of a maneuvering tug from different baselines and learns that neither is “now.”

Tutorial 2 — Heat is debt

A picket suppresses radiators, gains temporary concealment, then must choose between revealing itself and damaging components.

Tutorial 3 — An orbit is not a road

The player learns that a nearby target can be energetically inaccessible while a distant target has a cheap transfer opportunity.

Tutorial 4 — A beam is geometry

The player tasks a beam node, sees diffraction and pointing uncertainty, and learns why target state uncertainty can dominate raw power.

Tutorial 5 — Policy, not puppetry

The player issues a local defense policy; the unit acts after communication loss according to constraints.

Tutorial 6 — You can win without a hit

A kinetic salvo forces an alien factory convoy to spend enough \Delta v to miss its rendezvous window.

Tutorial 7 — Networks have anatomy

A small attack fragments a power/timing graph and disables much more capacity than the destroyed hardware alone suggests.

Tutorial 8 — Evidence is political

The player decides whether to share uncertain evidence with rival human blocs, trading coalition trust against operational secrecy.

148. Replay and explainability

Every consequential outcome must be reconstructable.

The replay system stores:

• objective event sequence;
• each faction’s received messages;
• each faction’s belief revisions;
• policy versions;
• random seeds;
• model versions;
• physics profile;
• parameter provenance.

An event explanation should answer:

1. What objectively happened?
2. What did this actor believe?
3. What observations supported that belief?
4. What policy constrained its action?
5. What resources were available?
6. Which uncertainty draw affected the result?
7. Which model version produced the calculation?

This is essential for debugging, balancing, player trust, and scientific use.

149. Telemetry for the vertical slice

Required aggregate metrics:

usable_power_W
unserved_power_W
radiator_margin_W
stored_heat_J
controlled_mass_kg
processed_mass_kg_day
replication_ratio
fleet_delta_v_remaining
median_track_position_error_m
track_calibration_score
command_latency_s
network_connected_component_fraction
factory_throughput_kg_s
population_supported
autonomy_violation_count
civilian_casualties
escalation_index

Required event-level telemetry:

event_id
event_type
objective_time
receipt_time_by_faction
decision_time
execution_time
entities
policy_id
observation_ids
energy_J
momentum_kg_m_s
heat_J
mass_change_kg
confidence_before
confidence_after
outcome_distribution
sampled_outcome

150. Mod support boundary

The vertical slice should support:

• custom scenarios;
• custom technology modules;
• custom faction doctrines;
• alternative physics profiles;
• custom entity components;
• scripted narrative events;
• custom victory metrics.

It should not initially support arbitrary native code mods. Use declarative data and a sandboxed scripting layer so deterministic replay, security, and provenance remain manageable.

Part XX — Development program

151. Workstreams

The program divides into nine parallel but dependency-linked workstreams.

152. Phase 0 — Research freeze and executable specification

Deliverables

• unit system and constants module;
• physics confidence taxonomy;
• parameter provenance schema;
• architecture decision records;
• vertical-slice scenario specification;
• deterministic random-number policy;
• reference benchmark notebook;
• initial technology library;
• risk register.

Exit criteria

• every implemented equation has units, tests, and a source or explicit fictional assumption;
• every speculative parameter is tagged;
• no “magic range,” “magic hit points,” or omniscient state remains in the core specification;
• all canonical scenarios declare a physics profile.

Effort band

4–8 person-weeks.

153. Phase 1 — Time, events, and orbital propagation

Tasks

• simulation clock;
• priority event queue;
• entity registry;
• inertial frames and coordinate transforms;
• analytic two-body coast;
• N-body propagation for selected bodies;
• finite-speed message propagation;
• maneuver segments;
• collision/intercept root finding;
• deterministic save/load;
• state snapshots and event log.

Exit tests

1. Two-body energy and angular-momentum drift remain within configured tolerances.
2. Hohmann-like transfer benchmarks reproduce analytic results.
3. Message arrival time equals emission time plus path length divided by c within numerical tolerance.
4. Replaying a saved run produces bitwise-identical event order where the platform permits, otherwise numerically bounded identical state.
5. One million dormant entities impose negligible high-frequency integration cost.

Effort band

12–20 person-weeks.

154. Phase 2 — Sensors and belief states

Tasks

• Planck radiance;
• passive optical/IR observation;
• active range and angle observations;
• field of view and pointing;
• background and detector noise;
• integration time;
• detection threshold model;
• observation association;
• Kalman-family baseline filter;
• interacting multiple-model track;
• track fusion;
• confidence calibration;
• decoys;
• track sharing and trust.

Exit tests

1. Detection probability curves are monotonic with aperture, exposure, and signal under fixed conditions.
2. False-alarm rates match configured thresholds in Monte Carlo.
3. Track covariance grows during coast without observation.
4. Maneuvering targets cause model-probability changes rather than impossible instantaneous certainty.
5. A faction receiving no observations cannot infer hidden objective state through UI or AI leakage.
6. Claimed 90% confidence intervals contain truth approximately 90% of the time over benchmark ensembles.

Effort band

16–28 person-weeks.

155. Phase 3 — Power, thermal state, and signature

Tasks

• collection;
• local generation;
• storage;
• transmission;
• conversion losses;
• load scheduling;
• heat source accounting;
• heat capacity;
• radiator emission;
• radiator orientation;
• heat storage;
• component temperature limits;
• thermal signature generation;
• emergency throttling;
• thermal concealment.

Exit tests

1. Energy is conserved to configured tolerance.
2. All inefficiency becomes heat unless explicitly exported.
3. Radiator power obeys T^4 behavior.
4. A radiator-damaged installation throttles or overheats rather than continuing at full power.
5. Thermal concealment accumulates exactly the corresponding stored heat.
6. Signature calculations use the same emitted power that leaves the thermal ledger.

Effort band

12–20 person-weeks.

156. Phase 4 — Propulsion, reachability, and logistics

Tasks

• impulsive burns;
• finite burns;
• low-thrust arcs;
• rocket-equation propellant use;
• external beam propulsion;
• launch and rendezvous planning;
• reachability approximations;
• depots;
• cargo;
• transfer contracts;
• convoy and escort logic;
• reaction-mass accounting.

Exit tests

1. No maneuver changes momentum without an accounted exchange.
2. No reaction drive receives free \Delta v.
3. External-beam vehicles require line of sight, source power, receiver geometry, and absorbed/rejected heat.
4. Reachable sets contain sampled feasible trajectories and reject configured infeasible cases at expected rates.
5. Propellant starvation can defeat a force even when its weapons remain intact.

Effort band

16–24 person-weeks.

157. Phase 5 — Combat, damage, and repair

Tasks

• laser propagation;
• beam pointing;
• target state uncertainty convolution;
• material absorption and ablation;
• kinetic launch and intercept;
• guided terminal maneuver;
• fragmentation;
• nuclear stand-off abstraction;
• charged-particle channel;
• armor stacks;
• component graph damage;
• radiator damage;
• delayed failure;
• repair and salvage;
• battle-damage uncertainty.

Exit tests

1. Laser intensity falls with computed beam area, not arbitrary range bands.
2. A perfectly powerful beam can miss due to prediction or pointing error.
3. Kinetic impact energy follows Newtonian or relativistic equations as appropriate.
4. A 100,000 kg projectile accelerated to 10 km/s stores 5\times10^{12} J before losses.
5. Damage outcomes conserve or explicitly export energy and momentum at the selected fidelity.
6. Mission kills can occur through heat, propellant, sensing, timing, or command failure without structural destruction.

Effort band

20–36 person-weeks.

158. Phase 6 — Dyson graphs and industrial ecology

Tasks

• statistical swarm populations;
• graph layers;
• power routing;
• timing and phase coherence;
• communication bandwidth;
• logistics dependencies;
• manufacturing recipes;
• mining;
• recycling;
• replication;
• repair clouds;
• network islands;
• capture and allegiance;
• population support.

Exit tests

1. Destroying random collectors causes graceful degradation.
2. Destroying small numbers of critical synchronization/routing nodes can cause nonlinear capability loss.
3. Network islanding preserves local operation while reducing global coordination.
4. Replication consumes accounted feedstock, energy, time, compute, and heat rejection.
5. A factory population grows exponentially only while resource and throughput constraints permit.
6. Restoration can be achieved by repair, rerouting, replacement, or authority change.

Effort band

20–32 person-weeks.

159. Phase 7 — Policy and autonomous agents

Tasks

• policy schema;
• authority inheritance;
• local rules of engagement;
• constrained planner;
• tactical targeting;
• risk and escalation budgets;
• fallback behavior;
• authentication;
• stale-policy handling;
• explanation generation;
• adversarial doctrine parameterization.

Exit tests

1. Units remain functional during communication loss.
2. Units cannot violate hard policy constraints except through explicitly modeled fault, compromise, or defection.
3. Stale policies expire according to their own rules.
4. The player can understand why a major autonomous action occurred.
5. The AI does not access objective state unavailable to its faction.
6. Different doctrine settings produce materially different but internally coherent behavior.

Effort band

20–40 person-weeks.

160. Phase 8 — Client, scenario, and player experience

Tasks

• 3D orbital map;
• 2D analytical projections;
• light-cone time controls;
• uncertainty visualization;
• thermal, power, reachability, and graph overlays;
• policy editor;
• event inbox;
• replay;
• accessible table views;
• tutorial;
• Cinder Seed scenario;
• sound and narrative presentation.

Exit tests

1. The player can distinguish truth, observation, estimate, and forecast.
2. Uncertainty remains readable without requiring probability expertise.
3. Color is never the sole information channel.
4. A keyboard-only player can operate all strategic functions.
5. Tutorial players can explain light delay, thermal debt, and reachable trajectories after completion.
6. The standard scenario is completable without tactical micromanagement.

Effort band

24–40 person-weeks.

161. Phase 9 — Hardening and release

Tasks

• profiling;
• memory optimization;
• long-run numerical stability;
• save migration;
• crash recovery;
• mod sandbox;
• telemetry privacy;
• scenario balance;
• documentation;
• scientific review;
• accessibility review;
• packaging.

Exit tests

• 20-year campaign completes within performance budget;
• no unbounded queue or track growth in soak tests;
• saves survive schema upgrades through migration tests;
• deterministic replays remain valid across patch versions where promised;
• all source-linked parameters pass provenance validation;
• accessibility acceptance tests pass;
• no critical security finding remains open.

Effort band

16–30 person-weeks.

162. Team shape

A serious vertical slice is feasible with a compact interdisciplinary team.

Minimum effective core:

• 1 simulation/astrodynamics engineer;
• 1 numerical methods and estimation engineer;
• 1 systems/gameplay engineer;
• 1 frontend/visualization engineer;
• 1 AI/planning engineer;
• 1 technical designer/research lead;
• part-time UX/accessibility specialist;
• part-time scientific reviewers.

A solo prototype should reduce scope sharply:

• two-body/ephemeris propagation;
• hundreds of explicit entities;
• one passive sensor band;
• simple EKF;
• one laser and one kinetic weapon;
• one graph layer;
• scripted alien doctrine;
• 2D map before 3D.

163. Critical dependency path

The critical path is:

units/provenance
    ↓
time + events
    ↓
trajectory propagation
    ↓
finite-speed observation
    ↓
belief state
    ↓
policy/autonomy
    ↓
resource + combat consequences
    ↓
UI and scenario

Industrial graphs and richer combat can develop in parallel after the event/belief spine stabilizes.

The project should resist the temptation to start with ship art, technology trees, or exotic weapons. Without the causal spine, those become conventional RTS content sitting on an inaccurate map.

164. Suggested repository structure

solar-war/
├── pyproject.toml
├── README.md
├── LICENSE
├── docs/
│   ├── master-plan.md
│   ├── physics-charter.md
│   ├── provenance.md
│   ├── adr/
│   └── references/
├── src/solarwar/
│   ├── constants/
│   ├── units/
│   ├── kernel/
│   ├── frames/
│   ├── astrodynamics/
│   ├── propagation/
│   ├── events/
│   ├── sensors/
│   ├── estimation/
│   ├── power/
│   ├── thermal/
│   ├── propulsion/
│   ├── reachability/
│   ├── combat/
│   ├── damage/
│   ├── manufacturing/
│   ├── networks/
│   ├── policy/
│   ├── ai/
│   ├── scenarios/
│   ├── telemetry/
│   └── api/
├── client/
│   ├── src/
│   └── tests/
├── data/
│   ├── physics_profiles/
│   ├── technologies/
│   ├── materials/
│   ├── factions/
│   ├── bodies/
│   └── scenarios/
├── notebooks/
│   ├── benchmark_orbits.ipynb
│   ├── sensor_snr.ipynb
│   ├── beam_damage.ipynb
│   └── calibration.ipynb
├── tests/
│   ├── unit/
│   ├── property/
│   ├── benchmark/
│   ├── scenario/
│   └── performance/
└── tools/
    ├── provenance_check.py
    ├── scenario_lint.py
    └── replay_diff.py

165. Architecture decision records

At minimum, record these decisions explicitly:

• ADR-001: SI units internally.
• ADR-002: event-driven core with adaptive local integration.
• ADR-003: objective state separated from faction belief state.
• ADR-004: no global omniscient query API for faction AI.
• ADR-005: all communication finite-speed unless physics profile declares otherwise.
• ADR-006: components over fixed unit classes.
• ADR-007: technologies data-driven and confidence-classed.
• ADR-008: deterministic random streams partitioned by subsystem.
• ADR-009: graph layers are separate, not one universal connectivity graph.
• ADR-010: explicit entities coexist with statistical cohorts.
• ADR-011: no universal hit points or shield points.
• ADR-012: source provenance is part of runtime data.
• ADR-013: player commands are policies at strategic scale.
• ADR-014: replay stores beliefs and messages, not only objective events.
• ADR-015: accessibility views are first-class, not post-processing.

Part XXI — Verification, validation, and calibration

166. Verification versus validation

Use the terms strictly.

Verification: did the code implement the specified model correctly?

Validation: is the model an adequate representation for its intended purpose?

A mathematically exact implementation of a bad abstraction is verified but not valid.

167. Dimensional analysis gate

Every equation exposed by a technology or component must declare dimensions.

A configuration linter should reject:

• energy where power is required;
• area where solid angle is required;
• kilograms used as weight;
• Celsius in absolute-temperature equations;
• delta-v granted without momentum exchange;
• efficiency outside [0,1] unless explicitly defined differently.

Recommended implementation:

• pint or a lightweight compile-time/runtime unit wrapper in Python;
• conversion to canonical SI at load;
• stripped scalar arrays only in performance-critical kernels after validation.

168. Conservation invariants

Depending on fidelity and external flux, test:

\Delta M = M_{\rm imported}-M_{\rm exported}-M_{\rm converted\ to\ radiation}/c^2

\Delta E = E_{\rm incident} + E_{\rm generated} - E_{\rm radiated} - E_{\rm exported}

\Delta \mathbf p = \int \mathbf F_{\rm external}dt + \mathbf p_{\rm imported} - \mathbf p_{\rm exported}.

The game may omit tiny mass-energy conversions at lower fidelity, but the omission must be explicit and bounded.

169. Canonical numerical benchmarks

Orbital

• Kepler two-body orbit over 10^4 periods.
• Earth-like orbit for 100 years.
• Earth–Moon restricted three-body transfer.
• known Lambert transfer cases.
• low-thrust spiral with published solution.
• close encounter requiring adaptive integrator.

Light and communication

• 1 AU one-way time: approximately 499.0 s.
• 30 AU one-way time: approximately 4.16 h.
• 550 AU one-way time: approximately 3.18 d.

Thermal

For a 1 GW ideal waste-heat load at \epsilon=0.9, approximate required radiator areas:

These are ideal black/gray-body values before plumbing, view factors, structural mass, degradation, or safety margin.

Kinetic

• 1\ \mathrm{kg} at 10 km/s: 5\times10^7 J.
• 10^5\ \mathrm{kg} at 10 km/s: 5\times10^{12} J.
• 10^8\ \mathrm{kg} at 0.999c: approximately 1.9\times10^{26} J.

Antimatter

One microgram of antimatter annihilating with one microgram of matter:

E \approx 1.80\times10^8\ \mathrm J

or about 43 kg TNT equivalent before coupling losses.

Diffraction

At \lambda=1.064\ \mu\mathrm m and R=1\ \mathrm{AU}, first-minimum radius under the simple circular-aperture approximation is approximately:

Real systems add beam quality, phasing, jitter, obscuration, pointing, and target-state uncertainty.

170. Sensor Monte Carlo suite

For each sensor model, generate ensembles across:

• target temperature;
• projected area;
• emissivity;
• aspect;
• range;
• background;
• aperture;
• band;
• exposure;
• detector noise;
• pointing;
• target motion.

Measure:

• ROC curve;
• false-alarm rate;
• missed-detection rate;
• localization error;
• classification calibration;
• time to track confirmation;
• vulnerability to decoys.

A detector does not receive a single “range.” It receives a performance surface.

171. Track calibration

Use proper scoring rules:

• negative log likelihood;
• Brier score for class probabilities;
• expected calibration error;
• normalized estimation error squared;
• track purity and completeness;
• association entropy.

Display calibration in development dashboards.

An AI that is less accurate but well calibrated may be strategically superior to an overconfident AI.

172. Beam model validation

Validate in layers.

Layer A — Geometry

• aperture;
• wavelength;
• divergence;
• spot size;
• pointing offset.

Layer B — Energy deposition

• incident intensity;
• reflection;
• absorption;
• reradiation;
• conduction;
• ablation threshold.

Layer C — Dynamic target

• attitude;
• rotation;
• maneuver;
• uncertainty;
• dwell interruption.

Layer D — Array coherence

• node phase error;
• timing error;
• path-length error;
• node dropout;
• baseline geometry.

Do not validate a communication link as though it proves weapon lethality. DSOC validates long-distance photon transmission and precision pointing; a damaging beam additionally requires vastly different delivered fluence, dwell, and target interaction.

173. Kinetic model validation

Test:

• launch dispersion;
• coast propagation;
• guidance;
• reachable terminal correction;
• intercept root;
• relative velocity;
• impact energy;
• fragmentation cone;
• momentum transfer;
• target evasive cost.

For relativistic vehicles, use a separate validated module above a configured \beta threshold rather than silently applying Newtonian formulas.

174. Thermal-network validation

A component-level thermal graph should be checked against:

• steady-state analytic cases;
• step load response;
• radiator loss;
• thermal storage saturation;
• view-factor occlusion;
• emergency throttling;
• phase changes if modeled.

Energy emitted as signature must exactly reduce internal thermal energy or derive from an explicit source.

175. Industrial ecology validation

Test limiting regimes:

1. abundant power, scarce mass;
2. abundant mass, scarce power;
3. abundant inputs, scarce cooling;
4. scarce compute/control;
5. logistics isolation;
6. exponential startup;
7. saturation;
8. attack-driven collapse;
9. repair-dominated recovery.

The replication model should approach logistic or resource-constrained growth, not unbounded exponentials.

176. Graph resilience analysis

For each graph layer measure:

• giant connected component;
• algebraic connectivity;
• betweenness concentration;
• flow capacity;
• restoration time;
• cascade size;
• island viability.

Run attacks by:

• random failure;
• highest-degree removal;
• highest-betweenness removal;
• geography-aware attack;
• adversarial optimization;
• compromised-authority injection.

Power, communications, timing, trust, and logistics should fail differently.

177. AI evaluation

AI performance is not only win rate.

Track:

• policy compliance;
• resource efficiency;
• belief calibration;
• civilian harm;
• escalation control;
• robustness to deception;
• performance under latency;
• performance under partition;
• explanation fidelity;
• exploitability;
• doctrine consistency.

Use fixed benchmark seeds and adversarial scenario generation.

178. Scenario validation

Every scenario receives:

• intended learning objective;
• plausible causal premise;
• declared abstractions;
• physics profile;
• balance envelope;
• expected strategic diversity;
• failure-state taxonomy;
• expert review;
• player comprehension testing.

A scenario is invalid if it only works because the UI gives the player information that the faction could not possess.

179. Performance validation

Required load tests:

• 100 million statistical swarm members;
• 1 million active cohort members;
• 100,000 simultaneous tracks;
• 5 million queued events;
• 20-year campaign;
• repeated save/load;
• graph fragmentation during peak engagement;
• 10× fast-forward through quiet periods.

Profile:

• event queue operations;
• ephemeris queries;
• covariance propagation;
• graph flow;
• visibility queries;
• serialization;
• client update volume.

180. Reproducibility

A replay manifest includes:

simulation_version:
git_commit:
scenario_hash:
physics_profile_hash:
data_pack_hash:
rng_algorithm:
rng_stream_seeds:
platform:
floating_point_mode:
start_time:
end_time:

Scientific/research mode should support exact environment capture with a lockfile and container image.

181. External review

Recommended reviewers:

• astrodynamics;
• spacecraft thermal engineering;
• optical systems;
• control/estimation;
• distributed systems;
• game AI;
• network science;
• accessibility;
• science-fiction worldbuilding;
• ethics and conflict modeling.

Review comments should be entered into the same provenance/decision system as code decisions.

Part XXII — Research backlog

182. Priority A: required before physics-core freeze

1. Optical target interaction
   • material reflectance/absorptance by wavelength and temperature;
   • laser ablation regimes;
   • rotating and layered targets;
   • uncertainty bounds.
2. Passive detection
   • validated photon-counting model;
   • zodiacal and solar backgrounds by geometry;
   • confusion and source association;
   • extended versus point target treatment.
3. Low-thrust reachability
   • fast approximations suitable for UI and AI;
   • error bounds against optimal-control solutions.
4. Thermal architecture
   • representative radiator specific masses;
   • deployable radiator failure;
   • coolant loops;
   • high-temperature materials.
5. Industrial recipes
   • broad energy-per-kilogram and throughput envelopes;
   • mining/refining assumptions;
   • recycling losses;
   • factory bootstrapping.
6. Autonomous policy execution
   • constrained planning under stale information;
   • verification of hard rules;
   • failure and compromise modes.

183. Priority B: needed for full Solar System campaign

1. Jupiter and Saturn system dynamics.
2. Solar occultation and conjunction geometry.
3. Large interferometric sensor networks.
4. Beam-phased array coherence under node motion.
5. Surface and subsurface planetary continuity.
6. Comet and volatile logistics.
7. Solar weather as environmental variability.
8. Habitat population and life-support metabolism.
9. Political coalition graph and legitimacy.
10. Interstellar arrival trajectory uncertainty.
11. ISM erosion for high-\beta probes.
12. Long-duration civilizational recovery.

184. Priority C: optional extreme-engineering modules

1. Large-scale antimatter catalysis.
2. Fusion concepts across multiple confinement regimes.
3. extremely large sail and beam infrastructure.
4. statites and non-Keplerian swarms.
5. stellar lifting.
6. natural compact-object industry.
7. stellar engines.
8. solar gravitational-lens observatories.
9. neutrino communications or sensing.
10. quantum illumination at system scale.

185. Priority D: speculative physics modules

Each must begin with a rule document answering:

• What new physical entity or coupling exists?
• What conservation laws remain?
• What controls the coupling?
• What observable signatures result?
• What countermeasures exist?
• Why does it not trivialize propulsion, power, sensing, or defense?
• What resource or geometry limits it?
• What source supports the theoretical starting point?
• Which part is fictional?

Candidates:

• axion/dark-photon sensing;
• engineered dark-sector communication;
• artificial micro-black holes;
• exotic negative-energy phenomena;
• spacetime manipulation;
• wormholes;
• artificial gravity;
• gravitational-wave generation.

186. Source-quality rubric

No Grade D/E source may establish a hard-physics parameter without independent confirmation.

187. Claim ledger

Every research claim should be stored as:

claim:
  id:
  statement:
  scope:
  confidence:
  source_ids:
  direct_support:
  extrapolation:
  competing_evidence:
  implementation_effect:
  owner:
  review_date:

This prevents a true statement such as “photon–graviton conversion is theoretically possible” from silently becoming the unsupported implementation claim “a practical gravitational-wave cannon exists.”

188. Uncertainty treatment

Parameter categories:

• epistemic: lack of knowledge;
• aleatory: inherent randomness;
• scenario: author-selected;
• fictional: invented rule;
• numerical: integration/approximation error.

The UI and telemetry should keep these distinguishable.

Where appropriate, use distributions rather than point values:

radiator_specific_mass:
  distribution: lognormal
  median_kg_m2: 4.0
  geometric_sigma: 2.0
  confidence: low

189. Research notebooks

Maintain executable notebooks for:

• detection SNR;
• beam spot and pointing;
• thermal radiator sizing;
• rocket performance;
• orbital transfers;
• reachable sets;
• ISM column damage;
• relativistic energy;
• graph resilience;
• replication dynamics;
• confidence calibration.

Notebooks are validation companions, not production code.

Part XXIII — Risk register and model governance

190. Risk register

191. Model change governance

A change to the physics core requires:

1. issue with motivation;
2. source or fictional declaration;
3. dimensional analysis;
4. benchmark;
5. backward-compatibility assessment;
6. balance assessment;
7. provenance update;
8. reviewer sign-off;
9. physics-profile version bump if behavior changes materially.

192. Technology trivialization review

Before adding any P2–P4 technology, answer whether it trivializes:

• energy generation;
• momentum exchange;
• heat rejection;
• observation;
• light-speed delay;
• material scarcity;
• manufacturing time;
• defense;
• distance;
• death or continuity.

If yes, the technology needs narrower scope, a strong cost, a countermeasure, or a separate scenario universe.

193. Ethical and narrative constraints

The simulation may depict catastrophic civilizational conflict, but its primary analytical focus should remain:

• infrastructure;
• deterrence;
• continuity;
• coordination;
• uncertainty;
• political choice;
• ecological/industrial competition.

Do not reward extermination as the default optimal policy. Make irreversible harm visible through population, culture, archives, habitats, biospheres, and legitimacy.

Alien systems should not be presumed hostile solely because they are unfamiliar.

194. Scientific honesty contract

The product should never label P3/P4 features “scientifically accurate.”

Preferred labels:

• observed;
• demonstrated;
• established theory;
• engineering extrapolation;
• theoretical possibility;
• speculative particle model;
• fictional rule.

The simulation can be imaginative without laundering imagination into fact.

Part XXIV — Initial parameter library

195. Universal constants

constants:
  c_m_s: 299792458.0
  G_m3_kg_s2: 6.67430e-11
  sigma_SB_W_m2_K4: 5.670374419e-8
  h_J_s: 6.62607015e-34
  k_B_J_K: 1.380649e-23
  AU_m: 149597870700.0
  solar_luminosity_W: 3.828e26
  solar_mass_kg: 1.98847e30
  standard_gravity_m_s2: 9.80665
  tnt_J_kg: 4.184e6

Use a source and version for astronomical constants; do not scatter constants through code.

196. Physics-profile controls

physics_profile:
  id: hard_sf_v0
  gravity:
    model: newtonian_nbody
    relativistic_threshold_beta: 0.01
  communications:
    propagation_speed_m_s: 299792458.0
    ftl: false
  thermal:
    enforced: true
  momentum:
    enforced: true
  sensors:
    objective_visibility: false
    photon_budget: true
  quantum_illumination:
    enabled: false
  dark_sector:
    enabled: false
  artificial_gravity:
    enabled: false
  force_fields:
    enabled: false
  artificial_black_holes:
    enabled: false