Tighten Undertow Run's early attack pacing

This commit is contained in:
raptorreplica committed 2026-10-18 01:37:13 +11:00
1 parent 1bbf78c78b
commit 1d0b2dda36
9 files changed
+78 -18

No files matched your search

+57 -3
View File
@@ -4,6 +4,57 @@ The 14 Iron Coast missions now use difficulty-scaled attack groups, clearer
weapon feedback and distinct enemy/boss patterns. Original campaigns and their
content definitions are unchanged.
Mission II 2, Undertow Run, version 1.0.3 moves one existing armed classic
curved pass from tick 1,650 to 1,530. The unarmed aircraft stays at 1,650, and
all other spawns and their fields remain unchanged. This count-neutral timing
correction leaves objectives, enemy statistics, equipment, prices, rewards and
the boss warning/arrival unchanged.
## Mission II 2: focused version 1.0.3 checks
On 18 October 2026, the frozen SourceE library (`3b04e473`) ran the unchanged
recorded paid-route controller. Four authentic post-Mission-1 states fed paired
version 1.0.2 and 1.0.3 continuations through Missions 2–8 and the finale replay:
68 ordinary mission executions in total. All completed with full pilot,
script-state and RNG save/load roundtrips. The changed Mission 2 version/hash
was reported and accepted through normal save compatibility; inventory and
other gameplay state were not silently rewritten.
Mission 2 completion times, shields and paid refill counts were unchanged at
all four difficulties. Ending shields were 79 / 52 / 44 / 12, preserving the
Elite route's narrow 12-shield margin. Veteran defeated one fewer optional
interceptor (two → one); Elite earned 100 more credits through the changed
encounter timing. Required skimmer and boss objectives were unchanged. These
are actual outcomes for the recorded controller and carried supplies, not a
guarantee for every loadout or play style.
Separate shield-replenished observations recorded the earlier Training/Rookie
quiet interval at ticks 1,448–1,650: 8.70 seconds with no visible enemies or
hostile shots. The retimed armed pass split it into a longest piece of 3.73
seconds and produced four observed hostile projectile births. The later quiet
intervals at 1,664–1,762 and 1,804–1,948 were unchanged, so the correction did
not merely move that lull later. Quiet time remains elsewhere, including the
deliberate boss recovery window.
Veteran/Elite arsenal observations absorbed six more replenished-shield points
each; their ordinary paid routes kept the same shields and purchases. No-fire
observers already had surviving required targets and do not establish ordinary
survival. These focused SourceE results are separate from the earlier
full-cohort measurements and do not claim a new SourceF playthrough or renderer
check.
The same seven maintained campaign, enemy and weapon suites were rerun against
the installed SourceF source epoch and version 1.0.3 definitions: **65 passed,
zero failed, four existing integration tests ignored**. The installed Linux
validator also accepts all 45 packages with one explicit override.
A separate SourceF presentation check rendered the same paid Veteran mission
state at tick 1,540 in SD and HD4 at 1920×1080. The retimed classic aircraft was
alive and entering from above the screen; partial top clipping and HUD overlap
are part of that entry pose. Both views and the actual Mods diagnostics were
reviewed at full size, with version 1.0.3 visible. This verifies that scene,
not every encounter or a complete SourceF mission playthrough.
## Wave schedules
Enabled scheduled enemies, including required targets and bosses:
@@ -25,6 +76,8 @@ Enabled scheduled enemies, including required targets and bosses:
| II 7 — Siege of the Iron Fleet | 57 | 65 | 77 | 93 |
| II 8 — Black Tide | 61 | 69 | 81 | 97 |
### Earlier version 1.0.2: a different middle window
Mission II 2 version 1.0.2 adds a small number of classic aircraft in its observed middle lull. Paired current-C paid continuations from authentic post-Mission-1 saves complete Missions 2–8 and the finale replay at all four difficulties, with unchanged ending shields, completion timings and refill counts. The added kills and pickups earn normal credits. In the separate shield-replenished arsenal fixture, the longest threat-free stretch within that middle window decreases from 8.91 to 5.87 seconds; the lull is shortened rather than eliminated. The full-cohort pressure tables below remain the recorded version 1.0.1 measurements. These focused checks were recorded on 17 October 2026 and are separate from the shield-replenished observations.
Counts are not simultaneous enemies or difficulty quotas. The 27 original
@@ -74,9 +127,10 @@ save-pin roundtrips at every difficulty. Gameplay measurements use the frozen
runtime definitions/scripts; the final version-number, README and blank-line
whitespace corrections do not change gameplay tokens.
Seven maintained campaign, enemy and weapon test suites pass against the same
runtime: 65 tests passed, none failed and four existing integration tests remain
ignored. Updated fixtures check the revised attack clocks, offscreen firing
Seven maintained campaign, enemy and weapon test suites passed against the
frozen final-B runtime on 17 October 2026: 65 tests passed, none failed and four
existing integration tests remained ignored. This is the recorded result, not
a new SourceF test run. Updated fixtures check the revised attack clocks, offscreen firing
suppression, reward-free aircraft exits and short contact effects. Synthetic
native-enemy placeholders in portable mission tests are explicitly labelled;
the original-catalog playthrough checks above cover actual native enemies.