Runtime hygiene: fix RefCounted cycles (clean suite exit), full-length shifts, wand cooldown
Full build / Import + headless tests (push) Successful in 18s
Full build / Full desktop build + optional release (push) Skipped

- FSM holds owner via WeakRef; ShiftCoordinator is RefCounted (logic-only);
  persistence tests free the Save node -> test suite exits with zero
  'resources still in use' warnings (was ~13) and 57/57 pass
- GameManager uses each scenario's shift_duration_s (900s ~ 10-15 min) for
  real shifts instead of the 300s test duration
- Wand screening gets a 0.8s recharge cooldown with a wand.recharge message
- TODO/DONE/MEMORY/README synced
This commit is contained in:
zwitschi committed 2026-10-31 13:50:18 +01:00
1 parent ba76c0febf
commit 08b3292b80
10 files changed
+57 -17

No files matched your search

+7 -1
View File
@@ -62,12 +62,18 @@ Session memory for durable decisions, state, and gotchas. Keep concise, remove o
- Detailed graphics: textured facade/awning/skyline, humanoid NPC rigs + props + walk bob +
attention head-turn, inspection artifact cards, mode icons, escalation vignette, body-cam
overlay (REC + shift clock). Screenshot set regenerated (`docs/screenshots/`, day + night).
- Polish: wand scan cooldown (`WAND_COOLDOWN_S` + `wand.recharge` message).
- NOTE: collaborators occasionally add uncommitted work to this shared tree (e.g. the radio
easter egg); check `git status` before committing and leave foreign changes untouched unless asked.
## Known Issues / Notes
- RefCounted reference cycle: `NPCController` holds an `FSM` that holds a back-reference to its owner, so test suites report "resources still in use at exit". Benign (process exit), but worth breaking later (e.g. `WeakRef` or explicit `dispose()`).
- Reference cycles are fixed: `FSM` holds its owner via `WeakRef`, `ShiftCoordinator` is
`RefCounted`, and persistence tests free the `Save` node — the suite exits with no
"resources still in use" warnings (was ~13). If a future subsystem introduces a cycle,
repeat the WeakRef pattern.
- Shift length: `GameManager` uses the scenario's `shift_duration_s` (900s ≈ 10-15 min);
tests and the screenshot tool override elapsed time directly.
- Headless (dummy renderer) logs "Parameter m is null" from `mesh_get_surface_count`; it does not occur with a real GL driver and does not affect logic.
- Steps are visual-only (no collider) so NPCs can path to the entrance; the guard spawns on the landing so it never needs to climb.
- Persistence tests write into the real `user://shift_results.json` (harmless local pollution of the demo history).