jakub
1b1a8b5173
Integrate faction and random settings menus, generated objectives, scouts, convoys, air support, mortar stations and body-intel recovery. Add identified-body aggression, finite hunter UAZ deployment, dedicated objective defence and maximum AI skill with separate reduced aim accuracy. Include policy tests, native probes, evidence and documentation.
69 lines
3.9 KiB
Markdown
69 lines
3.9 KiB
Markdown
# Mortar intel, prison visibility and perimeter guards
|
||
|
||
The live Everon configuration contains four mortar sites (`everon-mortar-1` through
|
||
`everon-mortar-4`). Each is placed on surveyed clear, dry terrain, at least 1 km
|
||
from another mortar and 500 m from the prison. They use the existing physical gun,
|
||
four-person crew, finite initial ammunition, HQ targeting and permanent station
|
||
assignment. The gun's destruction still releases its surviving crew.
|
||
|
||
Each site registers a normal replicated discovery location. Untargeted recovered
|
||
intel can select any undiscovered or partly discovered mortar site. A partial
|
||
clue displays the native `mark-question` icon at the centre of its 500 m map
|
||
square, with an unknown-location label. Identification changes that same marker
|
||
to `artilleryFireMission`, moves it to the actual position, and names the site.
|
||
The prison never displays a discovery marker, even when its knowledge is complete.
|
||
|
||
The 3–5 opening prison guards are placed once, at separate random positions
|
||
12–24 m from the prison centre. Placement samples angular sectors with jitter,
|
||
checks supported dry terrain and standing clearance, and reserves all positions
|
||
before moving any guard. They remain frozen until the gate is unlocked. The
|
||
existing physical manpack-radio requirement is unchanged.
|
||
|
||
The ammunition courier previously required exactly one battery. It now accepts
|
||
multiple batteries and selects a depleted recipient between trips. A pending
|
||
report, command or native inventory move retains its original target. The
|
||
existing outpost still has one finite reserve of 12 shells, shared by all four
|
||
batteries; adding batteries does not duplicate that stock or courier group.
|
||
|
||
## Original intel messages
|
||
|
||
Wording is taken from the owner's
|
||
`escapists-intel/Scripts/Game/EscapistsIntel/EI_IntelService.c`:
|
||
|
||
- `New location marked. Type unknown.`
|
||
- `Hey! A [type] was identified in the intel!`
|
||
- `Every known location has already been identified. This intel doesn't help us.`
|
||
|
||
Successor site types supply the appropriate noun, including `mortar site`.
|
||
Notifications retain reliable delivery and per-round duplicate protection.
|
||
Multiple discoveries in one scan may be combined into a multiline message;
|
||
rough clues never disclose the location type. A spent folder cannot give another
|
||
clue. Separate untargeted folders can concern separate locations, so two pickups
|
||
may produce two question marks rather than one fully identified site.
|
||
|
||
## Validation
|
||
|
||
The full repository checks passed after the mortar/courier changes. A further
|
||
production-method regression covers selection among four batteries, unavailable
|
||
recipients and target preservation during pending work. The Workbench run in
|
||
`/tmp/erri-mortar-map4/logs/logs_2026-09-22_16-11-24` passed 30 assertions, admitted
|
||
the player, checked frozen perimeter guards and their radio, unlocked the prison,
|
||
and recovered seven physical folders to identify all four mortar sites. It
|
||
completed with confirmed cleanup and unchanged source/dependency inputs.
|
||
|
||
That run preceded the final restoration of question marks and original messages;
|
||
its additional validation is recorded below. Native reports still include the
|
||
existing RHS/WCS localization and stock UI/resource errors. These are local
|
||
Workbench tests, not dedicated multiplayer acceptance. No server deployment or
|
||
Workshop publication was performed.
|
||
|
||
The final question-mark/message run,
|
||
`/tmp/erri-mortar-map5/logs/logs_2026-09-22_16-19-19`, completed 35 assertions with
|
||
no script errors or VM exceptions. It verified an approximate question marker
|
||
before recovering physical folders and checking the four final artillery icons.
|
||
The immediate runner cleanup check encountered an exiting process and reported
|
||
cleanup failure; a subsequent host-visible `assert_prefix_idle` confirmed the
|
||
private prefix idle. Source and external-addon inputs remained unchanged.
|
||
Production-method tests also check the original message strings, mixed discovery
|
||
batches, reliable-delivery deduplication, and exhausted-intel feedback.
|