Open Questions

Everything discussed but not settled. Grouped roughly by how soon it blocks the prototype.

Blocks or shapes the Tier 0–1 prototype

#	Question	Notes / leaning
1	Tier 1 build costs	Earlier proposals judged too low. Re-derive from the Tier 0 economy. Run 8 placeholders in balance-and-simulations.md.
2	Recipe rates and ratios for Refinery and Fabricator	Ratios should be deliberately uneven (2:3, 4:5). Run 8 used 5:4 refining and 3:2 iron:copper components.
3	One drill or two? Colite Drill and Mining Drill as separate buildings vs one general Colite-burning drill	—
4	Gun Bot ammo: physical ammo or infinite?	Physical ties military to logistics; infinite is simpler for the prototype.
5	Slots per Depot/Barracks (3 or 4) and slot cost per unit size	Concern that 3 free Harvesters per 100-Basalite Depot snowballs too fast. Proposed: Depot 3, Barracks 4 (tier0-proposals.md).
6	Truck respawn delay	60–90 s suggested.
7	Does an idle (starved or blocked) machine make noise / draw standby power?	Affects aggro. Proposed: idle machines are silent (tier0-proposals.md).
8	What does the Engineering Crew do before the first Depot? It idles for about 3 minutes because the ship holds no Basalite.	Starting stock (25–50 Basalite) is disfavored: Basalite is a native material the crew discovers, a ship wouldn’t carry stone, and building right away undercuts the survivor opening. Nearby Basalite shortens the wait but doesn’t fill it.
9	Exact list of what Engineering Crews can build remotely	Depot, walls and turrets at least.
10	Vision exploits: do only powered or active buildings give vision? Are some buildings base-only?	Proposed: Silos, foundations and walls give no vision (tier0-proposals.md).
11	Auto-foundation radius when placing near a network (5? 10?)	Proposed: 8 tiles (tier0-proposals.md). Auto-foundation itself is now wanted (base-network-and-inventory.md); also open: whether the path should be one tile wide or wider, and how the player opts out on a phone.
12	Order of the rest of Tier 1 research	Refining (free) → Logistics → Mining is set; Defense and Fabrication run parallel; the rest is open.
13	Overflow rule when a destroyed silo leaves the network over capacity	Lean: can consume, can’t add, until space frees up.
14	Deficit withdrawal order in the netting sim	Lean: output buffers first, then storage.
15	Day/night in the prototype or not?	Proposed for just after the first encounter.
16	Hive danger readability: visual size, labels, or let players find out?	—
17	Multiplayer in the first prototype?	Probably not, but keep the global Evolite balance and shared-world assumptions. The simulation is server-owned from the start (decided), so adding players later doesn’t change it.
36	Which recipes are chunked?	Each big recipe defines its chunks, or anything needing more than about one stack of an input is split automatically.
37	Where is stock in input buffers shown?	Not in network totals (decided). Probably on the machine’s own panel; maybe an “in machines” line on the inventory screen so iron doesn’t seem to vanish. This includes a chunk parked by a machine waiting for another input (see base-network-and-inventory.md).
38	Should Harvesters alone be the fastest way to Tier 2? Run 8 found that a Harvester-only rush beats drills under every setting tried, because drills arrive too late in Tier 1 to pay off.	Options: accept it (it’s the Doctor’s speedrun, and drills pay off in Tier 2+); gate the repair behind Mining (a drill-only material, or Mining as a prerequisite); or make big Harvester fleets cost something (noise, Depot slots, Depot price).
39	What is the electrical repair bill? It decides what Tier 1 is about: a component-heavy bill makes Fabrication the whole race, and a bulk-metal bill is the only kind that drills can speed up.	Run 8 placeholder: 2000 Iron, 1300 Copper, 50 Basic Components, 300 Colite, 100 Evolite.
40	Calcified Evolite rate and Tier 1 research costs	Run 8: 0.4/s and 60 / 80 / 100 / 150 put Mining at 25–27 min, but leave 300–500 Evolite unspent by Tier 2.
41	Does fuel take up a building’s storage? A small Outpost filled with Colite left its drill no room for ore, and the drill stopped for good (Run 8).	Either keep fuel separate from output room or make it a visible, fixable problem.
42	When does a crew’s material leave the network: when it loads at home, or when it builds?	logistics.md says crews draw materials from their home network; resources-and-economy.md says the cost is paid when the crew builds. The sync example in runtime-and-sync.md assumes loading.
44	When, if ever, to move the sim to Rust	Planned path, not a prototype need (player-and-platform.md). The parts of the ECS core that don’t map one to one onto Bevy (entity ids, next-tick messages, declared commands) are listed in ecs-core.md.
45	PixiJS or Three.js for the world?	Pixi for now (first-test.md). It depends on whether the final look is baked sprites (Pixi) or 3D models rendered live (Three.js). Switching later touches only the client’s drawing code.
47	Random numbers when several systems in one phase need them. The generator is one resource, and ResMut lets only one system per phase write it (ecs-core.md).	Options: a stream per system, seeded from the world’s seed and the system’s name; a stream per entity; or drawing all of a tick’s random numbers in one place. The first test needs little randomness.
48	How change detection works in TypeScript, for sending diffs instead of the whole state	Bevy marks a component changed when it’s written through &mut; TypeScript can’t see writes. Options: mark every item handed out by iterMut and getMut (coarse but automatic), an explicit markChanged, or comparing snapshots of a few busy components. Not needed until sync moves to diffs (runtime-and-sync.md).
49	How do players leave room so units don’t get boxed in? In the crowded hour, Depots packed around the ship closed off its dock at 12:27, and by the hour’s end 84 of 242 units had no path. canPlace keeps each dock tile free but not reachable; since units live inside buildings it keeps the whole entrance free, still without checking it’s reachable (first-test.md).	Options: refuse a building that leaves any dock (or ghost dock) unable to reach open ground (the “never stuck” rule, at the cost of a flood fill per placement and a refusal to explain); a margin or a clear side for each building (simple, and the gaps read as roads, but it wastes space, works against touching buildings forming networks, and can’t stop a ring of buildings); allow it and show a “No path” light, with demolishing as the fix. Units already retry every 3 s, so a way opened later frees them. Designer’s direction: no margin rule. Lean towards allowing it, with driveways (passable parts of a footprint) and foundation used as roads (base-network-and-inventory.md).
50	Should the Scout be flown directly, on a timed battery? After the first playtest the designer leaned toward yes, as the next gameplay change. Spokes are a chore even with a queue, and a long reach makes players want to see just past what a straight line shows.	The proposal: the battery drains per tick in the air, moving or hovering (a quadcopter), not per tile. Tapping the Scout is a shortcut to its building’s panel, and the player can send it somewhere, then somewhere else, while it flies. When the battery holds only the ticks needed to get home, it returns on its own and can’t be redirected. A “round trip” option on the panel keeps today’s one-tap trip. With control, the reach could go back up a little: it was cut for how dull spokes were, not for how much it covered. Harvesters stay as they are (a mission, no fuel, any distance; the wait is the cost). Later maybe: a navigation upgrade with queued waypoints for Harvesters, and light control for attack units. Clashes with “no direct unit control, ever” (player-and-platform.md).
51	Should a crew carry an inventory it fills for each trip from what the trip will build? Today a crew takes foundation in batches of up to 10 tiles, a rule for foundation alone (first-test.md).	The designer’s direction: give the crew an inventory, managed automatically from the trip’s jobs, so every job follows one predictable rule (several ghosts of any kind on one trip, up to what it carries). Open: how jobs are grouped into a trip, and what happens to a trip when one of its ghosts is cancelled.
52	Do the pacing targets in minutes halve with the game’s speed? After the second playtest the prototype runs twice as fast (balance-and-simulations.md): the first Depot at about 2:00, not 4:00. The targets (about 5 buildings at 12 minutes, a new capability about every 5 minutes, Tier 2 at 36–42) were set at the old speed.	Halve them: the same game in half the time, which is what 2× on screen felt like. Keep them: twice as much happens by each mark, so later tiers need more content or higher costs to fill the time. To settle by playing further than the first Depot.
53	What’s left of Investigate? The Scout scans an unknown deposit on site, Harvesters can’t mine one until it’s scanned, and scan time depends on the resource, about 15 s for easy ones (units-and-missions.md).	Still open: the time for each resource (Basalite 15 s, Evolite longest?), and whether Evolite’s scan replaces or comes before the sample-to-ship study (decision #81). Whether each later patch of a known resource is known at once, or needs a quick scan. How the scans fit the first-find checklist (narrative.md). The opening: the first Basalite’s 15 s scan pushes back the first Depot (now 1:58) by about that much plus the Scout’s flight, unless the patch the ship lands beside starts known. While the Scout scans it isn’t scouting, which matters less once it’s flown directly (#50), where hovering would drain its battery. Unknown patches need their own look and tooltip.
54	What’s left to decide about Gates? The rules are in combat-and-defense.md.	Cost and unlock: with Stone Walls at the first alien sighting, and whether a Reinforced Gate comes with the Reinforced Wall. What counts as near: how far from the gate an enemy keeps it shut, and whether that means seen enemies only, since a unit can’t react to what nobody sees. Turning it: the 2 × 1 piece runs along the wall, so it needs a way to lie across or along, like turning buildings. Opening and closing: how long each takes, whether it closes on a unit standing in it, and what happens when enemies arrive while it’s open. On screen: a waiting Harvester’s light and words (“Waiting: enemies at the gate”), and whether the gate shows open, shut or held shut. Paths: whether ground units path through a gate as though it were open, and wait there, or treat a held gate as wall and go round.

Design direction

#	Question	Notes
18	Captain died vs player is the captain	Leaderless is better for co-op and character; captain makes it personal. The “acting commander” variant was rejected.
19	Should the advisors react to the player’s play style?	Pinned (“maybe”). Not a reputation system.
20	Basic deposits: infinite or very slow depletion?	Latest statement: “patches don’t deplete.” Earlier agreement: Factorio-style very long depletion (around 18 hours). Calcified Evolite currently infinite.
21	Should ordinary mines ramp up noise over time like the drill?	Interesting (remote sites drawing new enemies, on/off and timers) but possibly tedious.
22	Uncollected dropped Evolite turns into new hives?	Deferred. If yes, probably needs critical mass to avoid exponential growth.
23	Bosses: roaming or dormant until disturbed? Names and types?	—
24	Roads/terrain: do trucks need roads, or do roads just speed them up?	Never answered.
25	Command Center role	Liked, but as an optional high-tier building. Possible: management UI, quiet mode, inter-base coordination, other planets.
26	Ship restoration list and the exact capability each repair unlocks	Only “electrical → Tier 2” is decided.
27	Ship systems visible early with unknown requirements?	Proposed, not confirmed.
28	Can materials committed to ship repairs be reclaimed?	Proposed to avoid deadlock.
29	Network-wide quiet mode (cut power plus recall vehicles plus stop Colite machines) and automated triggers	Proposed as later unlocks.
30	Do repair crews have dedicated buildings later, or stay in Depots?	—
31	Throughput limits on long foundation corridors?	Not for now.
32	Intermediate products for Tiers 2–4 (motors, circuits, composites…)	Needed for 50 hours of depth.
33	Ground spill / salvage piles when storage is destroyed	Set aside; needs a collection mechanic.

Naming

The game title, the planet, the alien species and creature types, the bosses, turret types, the final name for Outpost/Truck Stop, and whether Deepcore replaces or classifies the Evolite Drill. See naming.md.

Future expansion (explicitly later)

Rockets and interplanetary trade; other planets and bootstrapping planets in the spirit of Space Age; trains; the Doctor’s alien-research tech tree; a second act or sequel after launch.