Auto Warz — Space Layer Design¶

Document Type: Design Document Version: 0.6 Last Updated: 2026-09-06 Vault destination: 01_Design Documents Phase: Phase 1 — Space Update Related Documents: [05_World Generation & Planet Types](<../Content/05_World Generation & Planet Types.md>) · [19_Transport & Logistics](<../Logistics/19_Transport & Logistics.md>) · [33_Mothership & Fleet Design](<./33_Mothership & Fleet Design.md>) · [34_Ship & Mothership Editor Design](<./34_Ship & Mothership Editor Design.md>) · [30_Ship Construction System](<./30_Ship Construction System.md>) · [36_Progression Arc Design](<../Content/36_Progression Arc Design.md>) · [00_Canonical Design Facts](<../../04_Project Management/00_Canonical Design Facts.md>) Supersedes: the Unwritten/07_Space Exploration Design stub (OI-313)

Purpose of This Document¶

Space was the last layer in the game with a "NOT WRITTEN" banner where its design should be. That was defensible while space was Phase 2 content. It stopped being defensible on 2026-09-05, when the phase boundary moved and space became the second half of Early Access.

This document owns: what space is, how you cross it, why you would, and where the factory goes.

It deliberately does not own Genesis. See §8.

1. The Principle¶

[19_Transport & Logistics](<../Logistics/19_Transport & Logistics.md>) §2 sets a rule this document extends rather than breaks:

"Each carrier owns a different axis. None is an upgrade of another, and the ones that look similar fail differently."

Belt owns free to run, forever. Rail owns volume at distance, and districts. Drone owns no built path. Vertical structures own the only way between levels.

The freighter owns the only way between planets.

Space is the fourth carrier, not a second game.

1.1 Its honest cost¶

Every carrier in that table has a stated weakness. Rail's is latency: goods arrive in bursts, so both ends need buffering. The freighter is rail's weakness taken much further.

	Rail	Freighter
Arrival	Bursts	Rare bursts. A round trip is minutes
Ongoing cost	Fuel per run	Fuel per run, and more of it
Both ends need	Buffering	Serious buffering

A second planet is a batch relationship, never a flow.

You cannot belt-feed from another world. You stockpile there and you stockpile here, and the size of those two stockpiles is the design problem the whole layer hands the player. That single fact decides most of what follows.

2. The Gate Does Not Lead To Space¶

Correction, 2026-09-06. An earlier draft of [02_Scope & MVP Document](<../02_Scope & MVP Document.md>) said the gate ships in the Space Update because it is "the channel out of a private world". That was wrong.

Two facts settle it:

[05_World Generation & Planet Types](<../Content/05_World Generation & Planet Types.md>): your world already contains a galaxy. Galaxy, system, planet, terrain, all seeded, revealed lazily by a Scanner. Another planet is inside your world.
[32_The Gate & War Logistics](<./32_The Gate & War Logistics.md>): "One gate, one concurrent war. Maximum three." Gates are war slots, and they are powered down between wars.

You fly to another planet. You do not gate to it. The gate's only destination is a war, so the gate ships in the War Update.

Consequences, applied in [36_Progression Arc Design](<../Content/36_Progression Arc Design.md>): the First Gate Powered beat and Path B, Supply Chain Provider, both move to the War Update.

A peaceful first destination for the gate, a shared hub where private worlds trade materials before any war exists, is a real idea and a good one. It is not designed, so it is filed in 00_Ideas rather than assumed here.

3. Why You Go¶

Resources, and specifically the ones your world did not give you.

[05_World Generation & Planet Types](<../Content/05_World Generation & Planet Types.md>) already provides the whole answer: eleven planet types, each with guaranteed, uncommon and rare resource tables, and the starting planet is always Terrestrial.

You leave because the recipe you just unlocked needs Abyssal Crystals, and Abyssal Crystals are on ocean worlds, and you do not have one in reach.

Variety, not volume. You do not go to a second planet for more of the same, you go for something you cannot make. That keeps the reason to travel a production reason rather than a scale one, and it means a player who has what they need is not being punished for staying home.

4. Where the Factory Goes Is the Player's Decision¶

Decided 2026-09-06, and it is the centre of this layer.

The game does not prescribe a settlement pattern. It prices them, and the player chooses.

There is no rule that a remote planet must be a mine, and no rule that it must be a factory. Little bits everywhere, one main factory with mines feeding it, two or three specialised mains, or a full factory on every planet, all are legal. Which one is right depends on the seed, and working that out is the game.

4.1 What prices each pattern¶

Nothing here is a new system. Every one of these already exists and already behaves this way.

Priced by	Effect on settling widely
Round-trip time	Every planet boundary crossed is minutes and a buffer at both ends (§1.1)
Fuel	Each run costs, so distance is an economic decision, not only a patience one
Power crosses only in a container	A grid cannot span a vacuum, but charge ships (§4a). A site needs its own generation or a freight route carrying Charge Cells, and the second is a route that can fail
Threats are per planet	Each site has its own natives, its own pollution and its own frontier to hold (`[37_Planet Threats & Defence Design](<./37_Planet Threats & Defence Design.md>)`)
Vertical levels are per planet	Each planet's underground and sky are its own, unlocked and built separately (`[06_Vertical Layers & Core Power System](<./06_Vertical Layers & Core Power System.md>)`)
Research is world-scoped	The one thing that is free to spread. Research once, it applies everywhere. This is the counterweight that makes multiple sites viable at all

4.2 How each pattern fails¶

Good design here is not making them equal. It is making each one fail in a way the player can see coming.

Pattern	Good at	Fails by
One deep factory, mines elsewhere	One grid, one defence frontier, one place to think about	Long routes and huge buffers. Every rare input arrives in bursts, and a missed freighter stalls a chain hours later
Two or three specialised mains	Each factory suits its planet's resources; short internal routes	Freight between mains. You now have the routing problem and several grids
Little bits everywhere	Nothing travels far inside any one site	Many small grids and many small frontiers. Power and defence repeat at every site, and the route count explodes
A factory on every planet	Self-sufficiency; nothing waits on a freighter	You are playing five factory games at once, and every per-planet system repeats in full

Research being world-scoped is what makes this a choice rather than a trap. If tech were per-planet, only one pattern would ever be correct.

4.3 What this commits us to¶

Stating it plainly, because it is a cost:

Every planet-layer system must work on a planet the player is not looking at. Power shedding, pollution, threats, production. That is a simulation requirement, and §7 is where it gets answered.
The UI has to make a second planet legible without being there. Which planet is starving, which is under attack, which has a full silo waiting for a freighter. Answered 2026-09-07: you build a satellite ([40_Genesis Design](<./40_Genesis Design.md>) §7.1). Remote awareness is paid for, not free, and without a satellite a site is genuinely silent — you find out what happened when you arrive. That prices settling widely honestly: many sites means many satellites.

4a. Power Ships, In a Container¶

Decided 2026-09-06. A grid cannot span a vacuum, and that does not mean power stays put.

Power crosses the same way a fluid crosses onto a belt: in a container.

[19_Transport & Logistics](<../Logistics/19_Transport & Logistics.md>) §3.6 already had the shape. A barrel holds a fluid and a filled barrel is an ordinary solid item. A Charge Cell holds energy, and a charged cell is an ordinary solid item that a freighter carries like any other cargo.

This settles [25_Power System Design](<./25_Power System Design.md>) §5.3, which had listed three candidate approaches and recommended filing them to 00_Ideas. Cargo wins; beamed power and a direct grid link are dropped.

4a.1 The Cell is not an Accumulator¶

The Accumulator is a grid buffer measured in seconds: 500 PU·s, five seconds of a 100 PU deficit, deliberately short so that it smooths a spike rather than replacing a generator. A freighter carrying five hundred of them would deliver about fourteen minutes of a modest 300 PU site.

A Charge Cell is bulk energy for shipping. Large capacity, slow to charge, slow to discharge, and useless as a grid buffer.

Two objects, two jobs, the same relationship a building's internal buffer has to a Silo.

4a.2 The rule that keeps generation necessary¶

You can ship the stock. You cannot ship the capacity.

A charged cell is a fixed amount of energy that runs out. A generator is a rate that does not. Imported power can bootstrap a site, cover a night, or carry a small permanent draw at a high freight cost. It can never replace local generation, because the freighter's throughput is the ceiling and every cell is one more thing on the manifest.

4a.3 What this buys, which is the point¶

A new site is viable before it is self-sufficient. Land, build the drills, run them on imported charge, and stand up local generation when the site has earned it. Without this, every remote site needs a working power plant before it produces anything, which makes settling somewhere an all-or-nothing commitment. This is what makes §4's freedom actually playable.

Hard planets stay hard, without being closed. [25_Power System Design](<./25_Power System Design.md>) §5.2 gives the Rogue Planet no star, dead solar and the lowest core rating, calling it genuinely hard mode. You can now run one on imported power, and the price is a permanent freight burden rather than a locked door. A cost is a better hard mode than a wall.

4a.4 Importing power is how a player opts out of the ferals¶

Confirmed 2026-09-06: power draw does not create a signature. Generation and grid density do. [26_Pollution, Waste & Organic Systems Design](<./26_Pollution, Waste & Organic Systems Design.md>) §6 attracts the Ferals to large generators and dense grids, and that is meant literally.

So a site running on imported Cells, with no generation of its own, is quiet. And that is not a loophole, it is a choice the game is offering:

A player who does not want to deal with the ferals can choose not to generate locally.

Generate on a rock with nothing worth defending. Ship the charge to the site you actually care about.

The threat is not removed. It is relocated and paid for.

What you still pay	Why it stays honest
A permanent freight route	Cells in, forever. It is on the manifest and it competes with everything else the freighter carries
The ferals still come, to the generating site	You chose a planet where you do not mind. You did not make them go away
The natives are untouched	They come for pollution, and a factory still pollutes whatever powers it
Freighter throughput is the ceiling	§4a.2. You cannot import the capacity, so the site's size is capped by the route

That is the shape every good option in this game has: a real answer, at a real price, with a failure mode you can see coming. A player who takes it and then loses the route finds out what it was worth.

5. Manual First, Then the Route¶

Canonical §17: manual first, automation is built. Space follows it exactly, and it maps onto the mothership's existing automation ladder ([33_Mothership & Fleet Design](<./33_Mothership & Fleet Design.md>)) rather than inventing a second one.

Tier	What you do
Manual	You fly it. Load, travel, land, unload, come back. This is First Passage (`[36_Progression Arc Design](<../Content/36_Progression Arc Design.md>)` §5)
Standing route	A named origin and destination, a manifest, and a ship assigned to run it
Rules	Run when the home silo drops below 40%. Hold if fuel is short. Skip the run if the destination buffer is full
Autonomous	Set intent, and the mothership manages the fleet's routes against demand

The first trip must be flown. It is the beat, it is where the player learns what distance costs, and automating it before feeling it would waste the only moment where minutes-per-trip is experienced rather than read off a panel.

5.1 A route is a train schedule with a longer headway¶

Deliberately the same mental model as [22_Trains & Rail](<../Logistics/22_Trains & Rail.md>): an origin, a destination, a manifest, a condition to depart. A player who has run trains already knows how to run freighters, and the only new terms are fuel and a much longer headway.

That is the payoff of §1's framing. Space is the fourth carrier, so it should be learned in minutes by someone who understands the third.

6. The Orbital Station¶

Decided 2026-09-06: it is a service station, not a factory.

Repair, hangar, refuel. A place to service the fleet that is not at the bottom of a gravity well.

Function	Why it belongs off the ground
Refuel	A ship that must land to refuel spends fuel landing and taking off again. The station is where a route tops up without paying that twice
Repair	Ships come back damaged from mining hazards and, later, from wars. Repair is a materials cost, consistent with `[37_Planet Threats & Defence Design](<./37_Planet Threats & Defence Design.md>)` §8
Hangar	Somewhere to park a fleet that is not in use, and to hold ships between routes

Added 2026-09-20: it is also the cargo hub. The station is where the Mass Driver delivers, where freighters load and unload, and where the Drop Pod Bay sends from. That is not a change of character — §6a makes orbit the place goods change carrier, and the station is the only structure there.

What it is not: it is not a second factory and it does not manufacture. Its inputs are fuel, repair materials and whatever is in transit, and they arrive the same way everything else does, by being shipped there or thrown up to it.

That makes the station itself a destination with a supply problem, which is the correct shape: it is a consumer, like a turret, like a war.

Whether the station can build hulls too large to launch from a surface is open, and it is the natural home for that if the answer is yes. See Open Questions.

6a. Ships Do Not Land¶

Decided 2026-09-20. The largest single decision in this document's economics, and it came from noticing something true:

Getting to orbit is most of the energy. Going between orbits is comparatively little.

§6 already half-knew it — "a ship that must land to refuel spends fuel landing and taking off again". Taken to its conclusion: ships never land at all. They are orbit-to-orbit vessels, and the surface is reached by something else entirely.

6a.1 Three stages, three carriers¶

Stage	Carrier	Cost
Surface → orbit	Mass Driver. A building, not a ship	Expensive. This is where the energy goes
Orbit → orbit	Freighter hulls	Cheap in fuel, long in time
Orbit → surface	Drop Pod Bay. Also a building	Cheap. You are falling

Why the lift must not be a ship. If a shuttle flies cargo up, the whole fuel story collapses back into the fleet and every hull is designed around its worst leg. Making it a building turns reaching orbit into an infrastructure problem, which is the right shape for this game: it is a factory building that happens to point upward.

You do not fly cargo to orbit. You throw it.

And coming down needs no engine at all. You are not fighting gravity, you are surviving it.

6a.2 The asymmetry is the point¶

Importing to a planet is cheap. Exporting from it is expensive.

That is physically true, and it pays off the cargo mass decision in [34_Ship & Mothership Editor Design](<./34_Ship & Mothership Editor Design.md>) directly. If mass matters and the expensive leg is the lift, then:

Refine before you lift. Sending ore up costs power that could have refined it first, so a mining planet grows a factory rather than staying a field of drills. Nobody has to teach that — a player watches the Mass Driver's power draw and works it out.

It also makes a second planet a place rather than a quarry, which is what §4's where the factory goes is the player's decision was reaching for. And it rhymes with the ambient rule in [41_Fluid Packaging & Carriage Design](<../Logistics/41_Fluid Packaging & Carriage Design.md>): you already ship plate rather than molten metal, and now there is a second reason why.

6a.3 The three buildings¶

Building	Sits	Does
Mass Driver	Surface	Throws cargo to the station. ⚠️ Very large power draw. Batched, not continuous
Drop Pod Bay	Orbit, part of the station	Sends cargo down to a landing zone. Cheap, one-way, pods are consumed
Space Elevator	Surface to orbit	A later upgrade. Continuous rather than batched, far cheaper per unit, and a very large capital cost

The Elevator is the tier, and it is the right shape for one. A Mass Driver is cheap to build and expensive to run; an Elevator is brutal to build and cheap to run. That is the same trade as belts against drones — free to run forever, once the corridor exists — applied to the hardest corridor in the game.

6a.4 What this does to fuel¶

Transit fuel stops being the dominant number and becomes what it should be: the cost of propelling and maintaining the ship over a long, slow crossing. Range is still the constraint that shapes a design, but it is no longer inflated by a launch the ship never performs.

⚠️ It also removes a trap. Had ships landed, every freighter design would have been dominated by thrust-to-weight, and the editor's interesting decisions — cargo against range against protection — would have been crushed by one number.

7. Exploration, and What Runs While You Are Away¶

7.1 The Scanner reveals, the seed generates¶

Already decided in [05_World Generation & Planet Types](<../Content/05_World Generation & Planet Types.md>) and unchanged here:

The universe is infinite but lazy. Systems do not exist until a Scanner reveals them.
Revealing generates from seed plus coordinates, so nothing is stored for undiscovered systems and two players with the same seed see the same galaxy.
Star density rises toward the galactic centre, giving direction to exploration.

So exploration is a building you invest in, not a minigame. You build a Scanner, you get more of the map, you choose where to go.

7.2 The unattended-planet problem¶

§4.3 flagged it, and it is the largest technical consequence of letting players settle anywhere.

A planet the player is not standing on must still run. Its drills extract, its power sheds, its pollution spreads, its natives wake. Otherwise a second planet is a diorama.

[19_Transport & Logistics](<../Logistics/19_Transport & Logistics.md>) §4 already has the mechanism: Simulation Level, a global default with per-system overrides. A remote planet is the obvious candidate for a lower level. What "lower" means precisely is OI-285, still open, and this document makes it more urgent rather than answering it.

The rule this document does assert: whatever the fidelity, a remote planet's behaviour must be honest. If the player comes back and finds a stalled chain, the reason must be a real one they could have predicted, not an artefact of it having been simulated cheaply.

And a remote site can genuinely stall, which is the point. Its bots run on power, and if that site's grid fails the bots stop and there is no Genesis there to carry on ([40_Genesis Design](<./40_Genesis Design.md>) §4.0 and §5). A site left alone long enough with a marginal grid is a site that has done nothing for an hour. With a satellite you would have heard about it.

8. What This Document Defers, and What Has Since Been Answered¶

Genesis — answered 2026-09-07 in [40_Genesis Design](<./40_Genesis Design.md>). §4's freedom survives, and here is why:

Genesis founds. Bots continue. A site with no Construction Bot Bay needs Genesis to place the first one. After that the bots build there and Genesis is free to leave.

So a remote site is buildable while the player is elsewhere, which is what §4 needed. And founding one is a trip you physically make: while your Genesis is away, home can continue everything queued but cannot start anything new where no bots reach. A real cost with a name, and it matches the batch rhythm in §1.1.

There is also one Genesis per player, so co-op players found and build in parallel.

Also deferred: all numbers. Travel times, fuel per distance, station costs and buffer sizes belong to the item, building, recipe and flow pass (OI-363).

9. Vocabulary¶

Term	Meaning
Freighter	The carrier that crosses between planets. Owns that axis alone
Route	A standing origin, destination, manifest and departure condition. A train schedule with a longer headway
Orbital station	Repair, hangar and refuel, off the gravity well. Not a factory
Charge Cell	The container that lets power leave a grid. Bulk energy, slow both ways, an ordinary solid item when charged. Not an Accumulator
Site	Any place the player has built, home planet or not. Deliberately neutral about whether it is a factory or a mine

Note that outpost is not used here for a remote site, because Outpost Mode is a separate product on its own track.

Open items for this document are tracked in docs/open-items.md, area space.

Open Questions¶

Whether the orbital station can build hulls too large to launch from a surface. It is the natural home for that, and it would set up the Master Rill later
What Simulation Level actually does to a remote planet (OI-285), which §7.2 makes urgent
Whether a route can be interrupted, by hazard or, later, by an enemy, and what happens to the cargo
Whether fuel is refined on the home planet only, which would make every remote site dependent on home for its own return trips
Charge Cell capacity, charge and discharge rates, and shipping loss. The loss should exceed the Accumulator's 10%, for the reason 25_Power System Design §5.5 gives: storage is a service you pay for. Data pass
Whether Charge Cell is the right name, and whether an empty one ships back or is consumed
How many planets a player can reasonably hold before the UI stops coping, which is a playtest question rather than a rule
Whether the Scanner has an ongoing cost or is a one-off reveal per system

Revision History¶

Version	Date	Changes
0.6	2026-09-20	§6a added: ships do not land. From El's observation that getting to orbit is most of the energy while going between orbits is comparatively little — which §6 already half-knew, noting that a ship landing to refuel spends fuel landing and taking off again. Ships become orbit-to-orbit vessels and the surface is reached by buildings: a Mass Driver up (expensive, batched, very large power draw), a Drop Pod Bay down (cheap, one-way, you are falling), and a Space Elevator as a later upgrade (continuous, brutal capital cost, cheap per unit — the belt-against-drone trade applied to the hardest corridor in the game). ⚠️ The lift must not be a ship, or the fuel story collapses back into the fleet and every hull is designed around its worst leg. The asymmetry is the point: importing to a planet is cheap and exporting from it is expensive, which pays off the cargo-mass decision — refine before you lift — so a mining planet grows a factory rather than staying a field of drills. §6's station gains the cargo hub role, since orbit is now where goods change carrier.
0.5	2026-09-07	§4.3 and §7.2 answered by `40_Genesis Design`: remote awareness is a satellite you build, so an unmonitored site is genuinely silent, and a remote site can genuinely stall because its bots run on power and no Genesis is there to carry on.
0.4	2026-09-07	§8's Genesis deferral resolved by `40_Genesis Design`: Genesis founds, bots continue, so a remote site is buildable while the player is elsewhere and §4's settlement freedom holds. Founding a site is a trip that costs the home planet its ability to start anything new. One Genesis per player.
0.3	2026-09-06	§4a.4 rewritten and confirmed. Power draw does not create a feral signature; generation and grid density do. So importing power is how a player opts out of the ferals, and that is offered rather than tolerated: a player who does not want to deal with them can choose not to generate locally. The threat is relocated and paid for, not removed — a permanent freight route, the ferals still visiting the generating site, the natives untouched since they come for pollution, and freighter throughput capping the site's size. Charge Cell added to the vocabulary.
0.2	2026-09-06	§4a added: power ships, in a container. A grid cannot span a vacuum, so charge crosses the way a fluid crosses onto a belt, extending `19_Transport & Logistics` §3.6 to a third container. The shipped object is a Charge Cell, not an Accumulator — an Accumulator is 500 PU·s, five seconds of one generator, and five hundred of them would deliver fourteen minutes of a modest site. You can ship the stock, you cannot ship the capacity, which keeps local generation necessary since the freighter's throughput is the ceiling. Two payoffs: a new site is viable before it is self-sufficient, which is what makes §4's settlement freedom playable, and hard planets like the Rogue stay hard as a permanent freight burden rather than a locked door. Records as intended that a site running on imported charge is quiet to the ferals, so a player may deliberately move their noise to a planet they do not care about. §4.1 pricing row corrected.
0.1	2026-09-06	Initial document, superseding the `Unwritten/07` stub. Space is the fourth carrier: the freighter owns the only way between planets, and its cost is round-trip time and fuel, making a second planet a batch relationship, never a flow. The gate does not lead to space — your world already contains a galaxy, and the gate's only destination is a war, so it moves to the War Update along with First Gate Powered and Path B. You go for resources you cannot make, variety rather than volume. Where the factory goes is the player's decision: the game prices four settlement patterns and prescribes none, with research being world-scoped as the counterweight that makes multiple sites viable. Manual first, then the route, mapped onto the mothership's existing automation ladder. The orbital station is repair, hangar and refuel, a consumer with a supply problem rather than a second factory. Genesis deliberately deferred to its own conversation, and §4 written to survive either answer.