# The service you keep not taking
Every hull carries a maintenance posture: preventive, balanced, run to failure. Until this week the posture was a multiplier on wear. 0.7, 1.0, 1.4. Preventive was the only sane pick, because less wear also means a lower daily maintenance charge, and the insurance premium it was meant to trade against doesn't do anything yet. One right answer and two worse ones, and I'd been showing it to players as a decision.## A threshold, not a rateProactive maintenance repairs steel. It doesn't make steel wear more slowly. The posture sets the condition at which a service falls due, and nothing else. 85, 70, or never. Cross the line and the hull raises a claim against itself. There's no defer button. Deferring isn't a moment, you just keep flying. The claim stays open, a surcharge lands on the daily running cost where you're already looking, and the hull stays on its wear curve, so the drydock you eventually take is a bigger one.Forty runs from 88%, against a 95% repair ceiling:
preventive services at 85 14 services never below 88
balanced services at 70 5 services dips to 71
run to failure never 0 services slides to 0Call it 48 hours a visit. 672 hours in the yard against 240. Preventive buys a higher floor for about three times the grounding, which is the trade I wanted and didn't have. Provisional pending playtest, like every number here.## The graph had nothing to sayShip Details draws a condition trend. It projected wear forward under the posture, and the projection only ever subtracted, so it drew a monotone line between two endpoints the bar directly above it already stated. Delete it and nothing's lost.
The Ship Details condition trend: a sawtooth forecast against the hull's repair ceiling, with the service rule at 70 and the condemned band under 30
It books the services now. The sawtooth says how often this hull comes off the line, which is what the posture buys and what a bar can't draw. It runs against the hull's own repair ceiling, and that ceiling erodes with flight hours, so an old hull's forecast climbs back to less than a young one's. When to renew it, not only when to service it.Then the bill, in the header: *Plan: 3 services over 30 runs · first in 9 · 6 days in the yard.* Counted off the forecast's own service runs, so the picture and the price can't disagree about how many visits it takes. A readout that showed only the benefit would recommend preventive all over again, which is the thing I just took out.The flight card says it too. The hull row already drew condition now and ghosted where the plan leaves it; what it couldn't say was whether that lands past the line. So the line is a tick on that bar, amber, with the number written under it.
Flight authoring. The hull row reads 72% to 65% with an amber service 70% mark on the bar
Amber and not red because crossing it owes you a drydock. It doesn't refuse the flight.## Planned against actualWhat I want out of that graph is the asset-management reading. Here's when we should renew it, here's what ops actually booked in. The cost is downtime as much as it's repair bills.So the hull keeps a log, and it records two things. A service was raised. A service was taken. Deferral is the gap between them, derived rather than logged, because there's no moment to attach it to. Under the graph: taken of raised, what's outstanding, the longest a service ever went unattended. A hull too new to have owed one reports no data. Not perfect adherence.Condition on its own can't tell you any of that. A hull in good shape was either looked after or barely flown.## Old and well keptNone of it compares against anything if every hull in the world got the same treatment.The used-hull catalogue was seven profiles hardcoded in the shipyard panel, a UI file deciding what exists in the world, and condition was a strict function of age. 100 at new, 94 at two years, 86 at four, 78 at seven, 72 at nine, 64 at eleven. Monotone all the way down. An old hull in good condition wasn't rare in that table. There was nowhere to write one.Ten profiles now, authored in the tuning YAML, and each rotating set holds a pair that argues with age. A twelve-year hull at 91% that took nine services out of nine, sitting next to a four-year at 86%. A three-year at 68% that took one of four, next to an eleven-year at 64% carrying two refits on top. The careful ones have fewer refits, not more. A refit is a class change and this economy prices it as damage.
The Shipyard catalogue: seven listings ranked by price, with condition and an expiry countdown, and an 11-year 64% hull selected
Buy one and the record comes with the paperwork. The counts are authored rather than derived from age. Deriving them would make adherence a function of age again, one layer down from where I'd just taken it out.## Two things that were brokenChecking that a bought veteran could draw its own graph turned up its whole service history stacked on hour zero. Campaign time starts at zero, and a twelve-year-old hull was flying before the company existed, so its services genuinely fall at negative hours. The clamp put all nine of them on the same morning. The suite that exists to catch exactly that buys its hull at hour 100,000, where the clamp never fires.The other one's worse. Hulls were being delivered free. Purchase debits the price and nothing else, and the panel was handing its "Deliver to" port straight through, so the hull relocated to wherever the selector pointed at no cost, while the confirmation described an empty flight — "arrives at Ceres in 6.1 d, empty. §84k fuel + opex" — that the sim never flew. The hull stays berthed at the yard that sold it now. The column is the sticker price. Repositioning is a thing you choose and pay for.The catalogue has always expired its listings on a shelf clock. It just never said so, so a player saving for a hull had no way to know it was going. Both catalogues carry the countdown now, read off the same primitive that does the pruning.