perf: re-evaluate wart targets after each confirmed tool swing #95

Merged

gecko merged 1 commit from t3/toolmode-opportunistic into main

2026-11-02 19:02:04 +00:00

gecko commented

2026-11-02 18:40:22 +00:00

Owner

This targets the observed single-direction swing cadence without relaxing the one-unconfirmed-swing or packet/reach guards.

Validation: ./gradlew test --tests systems.gecko.mcbot.farming.WartToolSweepTest --tests systems.gecko.mcbot.farming.WartToolRouteTest.

After a confirmed nether-wart tool swing, rank every reachable ripe target again instead of letting the pre-aimed route target pin the next swing. The prepared target remains a fallback, so sideways and diagonal opportunities are used before moving on. This targets the observed single-direction swing cadence without relaxing the one-unconfirmed-swing or packet/reach guards. Validation: `./gradlew test --tests systems.gecko.mcbot.farming.WartToolSweepTest --tests systems.gecko.mcbot.farming.WartToolRouteTest`.

gecko added 1 commit

2026-11-02 18:40:22 +00:00

perf: re-evaluate all reachable wart tiles after confirmation

build / Java 25 build (pull_request) Successful in 2m57s

Details

f41b3f5cf6

After a confirmed tool swing, rank every reachable ripe target again instead of letting the pre-aimed route target pin the next swing. Keep the prepared target as a fallback when no other tile is usable, so sideways and diagonal opportunities are taken before moving on.

gecko merged commit 3eab6fad83 into main

2026-11-02 19:02:04 +00:00

gecko referenced this pull request from a commit

2026-11-02 19:02:05 +00:00

Merge pull request 'perf: re-evaluate wart targets after each confirmed tool swing' (#95) from t3/toolmode-opportunistic into main