OpenCode/Flight-Ops
2
0
Fork 0
Code Issues Pull Requests Actions Activity

Idle fleet freezes world clock — market stops generating jobs; slow to real-time instead of freezing #128

New Issue
Closed
opened 2026-09-22 19:08:38 +02:00 by kamilkrzywda · 1 comment
kamilkrzywda commented 2026-09-22 19:08:38 +02:00
Owner
Copy Link
Copy Source

Reported by user (2026-09-22): "when game is paused no new jobs are appearing - game shouldn't stop at all - it should slow down to 1/100 or something - 1/24 would be fun number - let's use our speed constants."

Problem

Since v0.52.0, when the whole fleet is grounded (idle-pause via shouldAdvanceTime in lib/sim/engine.ts), the world clock freezes completely. Package generation/expiry only run on advancing ticks → while idle, no new contracts appear and nothing expires. From the player's perspective the game is dead until they dispatch a plane.

Change

When all planes are grounded, advance the world clock at a minimum rate instead of freezing: real-time (1 game-second per real second) = WORLD_TIME_SCALE / 24 — derived from the existing speed constants in lib/worldClock.ts, no new magic number. When any plane is active, behavior is unchanged (full scale; test-hook set-timescale overrides as today).

Tests to update

  • e2e/sim.spec.ts "a fully grounded fleet freezes the clock exactly across polls" → now: advances at ~real-time (bounded tolerance, clearly below full 24x)
  • e2e/state.spec.ts:42 exact clock equality across restart relies on idle-freeze → tolerance-based bounds instead
  • Audit unit tests (lib/sim/engine.test.ts etc.) + other e2e specs for freeze assumptions; the activateWorld/repair pattern in actions.spec.ts stays valid

Process

Branch off master after v0.55.0 (PR #126) merges. Version 0.56.0. No API change (no OpenAPI delta).

Reported by user (2026-09-22): "when game is paused no new jobs are appearing - game shouldn't stop at all - it should slow down to 1/100 or something - 1/24 would be fun number - let's use our speed constants." ## Problem Since v0.52.0, when the whole fleet is grounded (idle-pause via `shouldAdvanceTime` in lib/sim/engine.ts), the world clock freezes completely. Package generation/expiry only run on advancing ticks → while idle, no new contracts appear and nothing expires. From the player's perspective the game is dead until they dispatch a plane. ## Change When all planes are grounded, advance the world clock at a minimum rate instead of freezing: **real-time** (1 game-second per real second) = `WORLD_TIME_SCALE / 24` — derived from the existing speed constants in lib/worldClock.ts, no new magic number. When any plane is active, behavior is unchanged (full scale; test-hook set-timescale overrides as today). ## Tests to update - e2e/sim.spec.ts "a fully grounded fleet freezes the clock exactly across polls" → now: advances at ~real-time (bounded tolerance, clearly below full 24x) - e2e/state.spec.ts:42 exact clock equality across restart relies on idle-freeze → tolerance-based bounds instead - Audit unit tests (lib/sim/engine.test.ts etc.) + other e2e specs for freeze assumptions; the activateWorld/repair pattern in actions.spec.ts stays valid ## Process Branch off master after v0.55.0 (PR #126) merges. Version 0.56.0. No API change (no OpenAPI delta).
kamilkrzywda commented 2026-09-22 20:59:06 +02:00
Author
Owner
Copy Link
Copy Source

Fixed in v0.56.0 (PR #130, merge commit 9b2938a).

A fully grounded fleet no longer freezes the world clock — it now advances at real-time: new constant IDLE_WORLD_TIME_SCALE = WORLD_TIME_SCALE / 24 (= exactly 1) in lib/worldClock.ts, and tickOnce (lib/sim/engine.ts) uses it instead of skipping advancement when shouldAdvanceTime is false. Packages keep spawning/expiring while idle, so the market stays fresh; an active fleet keeps its full scale including any FLIGHT_TEST_HOOKS override (grounded + hook scale 1000 → ×1, not ×1000), and lastRealTickMs always tracks real time so resuming activity never fast-forwards the idle duration. The client mirror (advanceSimClock) extrapolates at the same idle rate between polls instead of freezing.

Tests: unit — engine.test.ts pins clamped real-time advance, resume-without-catch-up, hook-scale-ignored-while-grounded and the constant itself; useSimClock.test.ts mirrors it (366 total). E2E — sim.spec.ts freeze test became a bounded real-time delta check; state.spec.ts restart test uses derived bounds instead of exact equality; right-pane.spec.ts penalty preview no longer races a frozen clock.

Deployed: https://flight.dev.kkhost.pl (/api/version = 0.56.0, verified).

Fixed in **v0.56.0** (PR #130, merge commit `9b2938a`). A fully grounded fleet no longer freezes the world clock — it now advances at real-time: new constant `IDLE_WORLD_TIME_SCALE = WORLD_TIME_SCALE / 24` (= exactly 1) in `lib/worldClock.ts`, and `tickOnce` (lib/sim/engine.ts) uses it instead of skipping advancement when `shouldAdvanceTime` is false. Packages keep spawning/expiring while idle, so the market stays fresh; an active fleet keeps its full scale including any FLIGHT_TEST_HOOKS override (grounded + hook scale 1000 → ×1, not ×1000), and `lastRealTickMs` always tracks real time so resuming activity never fast-forwards the idle duration. The client mirror (`advanceSimClock`) extrapolates at the same idle rate between polls instead of freezing. Tests: unit — engine.test.ts pins clamped real-time advance, resume-without-catch-up, hook-scale-ignored-while-grounded and the constant itself; useSimClock.test.ts mirrors it (366 total). E2E — sim.spec.ts freeze test became a bounded real-time delta check; state.spec.ts restart test uses derived bounds instead of exact equality; right-pane.spec.ts penalty preview no longer races a frozen clock. Deployed: https://flight.dev.kkhost.pl (`/api/version` = 0.56.0, verified).
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
bug
Something is not working
 duplicate
This issue or pull request already exists
 enhancement
New feature
 help wanted
Need some help
 invalid
Something is wrong
 question
More information is needed
 wontfix
This won't be fixed
No labels
Milestone
No items
No Milestone
Assignees
Clear assignees
kamilkrzywda (Kamil Krzywda) smysakowski
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: OpenCode/Flight-Ops#128