aussie/EscapistsReintroduced
1
0
Fork 0
forked from escapists/EscapistsReintroduced
Code Pull Requests Activity

Capture current mission, AI, map and administration development

Browse Source 
Include the current addon, terrain integrations, tools, regression tests and recorded development evidence.
This commit is contained in:
jakub
2026-10-13 22:52:53 +02:00
parent a951a96edc
commit 6a667d8deb
8104 changed files with 2124945 additions and 2061 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/radio-document-loss.md
+4 -4
View File
@@ -6,13 +6,13 @@ A conscious participant can also identify the actual radio station from within e
The server validates the currently controlled player entity, connected account and roster life, player faction, active mission phase, registered site, distance and visibility. AI, detached bodies and spectators cannot identify a site. Radio authoring accepts at most eight metres; optional sites retain their sixty-metre maximum; prison and extraction proximity discovery remain disabled. Escape cannot directly reveal the radio.
The extraction map marker and objective copy now explicitly describe the initial helicopter pickup site. It is a location marker, not live helicopter tracking.
The extraction marker now identifies the exact smoke-signalling pickup point. Transport is dispatched after the signal; the marker does not track moving aircraft.
## Native fixture, not yet run
## Native fixture and historical evidence
Run the Core fixture with `--smoke radio-document-loss` or `--smoke radio-document-loss-partial` after applying the integration snippets. The partial variant first transfers one real document into the real player's inventory with the isolated reveal chance set to zero, waits for ordinary discovery to produce knowledge 1, then destroys all four original document entities. The unknown variant destroys all four before any recovery. Both verify native weak references clear, cache/player inventories contain no target documents, and querying preparation does not replenish stock. A nearby unpossessed actor and a distant real player must not reveal the station. A wall inserted before a close approach must prevent discovery; removing it lets the normal site tick identify the station exactly once. The real participant must then pass ordinary tuning authorization and enter radio defence. The fixture ends with an explicit test defeat/reset; it does not claim a completed escape, manual play or multiplayer validation.
Run the Core fixture with `--smoke radio-document-loss` or `--smoke radio-document-loss-partial` using the integrated fixture. The partial variant first transfers one real document into the real player's inventory with the isolated reveal chance set to zero, waits for ordinary discovery to produce knowledge 1, then destroys all four original document entities. The unknown variant destroys all four before any recovery. Both verify native weak references clear, cache/player inventories contain no target documents, and querying preparation does not replenish stock. A nearby unpossessed actor and a distant real player must not reveal the station. A wall inserted before a close approach must prevent discovery; removing it lets the normal site tick identify the station exactly once. The real participant must then pass ordinary tuning authorization and enter radio defence. The fixture ends with an explicit test defeat/reset; it does not claim a completed escape, manual play or multiplayer validation.
Only disposable fixture stock and test entities are deleted. The production path neither spawns replacement documents nor changes inventories. No original assets or mission files were changed. Source-policy tests execute actual discovery-bound methods after mechanical translation; native participant/trace behavior still requires the fixture run.
Only disposable fixture stock and test entities are deleted. The production path neither spawns replacement documents nor changes inventories. No original assets or mission files were changed. Source-policy tests execute actual discovery-bound methods after mechanical translation; the historical native participant/trace coverage is recorded below. Those runs do not establish current multiplayer acceptance.
Native run `native-2026-09-19-070733` passes 58 combined assertions with zero engine or script errors for initially unknown radio knowledge. The previous failed trace diagnostic `070446` had clear sight but inconsistent offline identity lookup. Discovery now uses the audited roster’s canonical current-actor validation, shared with radio authorization. The partial-knowledge run remains separate.