Skip to content

ntre_apparatus_dm.nav: cut the helipad fence drops and mark the fence tops - #167

Open
sunzenshen wants to merge 1 commit into
NeotokyoRebuild:masterfrom
sunzenshen:navmesh-apparatus-dm-fence-drops
Open

sunzenshen wants to merge 1 commit into
NeotokyoRebuild:masterfrom
sunzenshen:navmesh-apparatus-dm-fence-drops

Conversation

@sunzenshen

Copy link
Copy Markdown
Contributor

Follows the review of #157 (sunmachine).

What the review found. Three one-way drops off the helipad fence, 4671 / 4674 / 4672 -> 415, cross the open kill
slot west of the pad (~76 u down; the slot is 28-98 u wide). A bot steers at the first point past the fence tip,
which is over the slot. The fence-top areas carried no CLIFF / PRECISE, so a bot on them could also dodge or
side-step into the slot. No spawn-to-spawn or spawn-to-area shortest path uses the three drops; they carry roam and
chase traffic only. Separately, 3401 lacked the PRECISE + CLIFF its mirror 3404 has.

What changed (area ids unchanged):

  • Cut 4671 -> 415, 4674 -> 415, 4672 -> 415. The way down stays 93 -> 8436 -> 415.
  • PRECISE + CLIFF on the 11 pad-edge areas whose footprint is at least 30 % over the open slot (BSP + static prop
    collision, no solid from z 572 down to the trigger_hurt): 4671, 4672, 4674, 4676, 4679, 4733, 4731, 4677, 4673,
    8286, 1819. That is the review's three plus eight.
  • PRECISE + CLIFF on 3401.

Evidence.

  • Offline (ntnav review against master): connections -3 (the three drops), nothing else; no pocket added or removed,
    no area unreachable or entered with no way back, no route longer by more than 10 %. Each fence top still links
    back onto the pad.

  • nav_analyze rerun: same graph, analysis data unchanged.

  • In game (Linux dedicated, Release build of neo master f426c222b plus harness logging; 30 bots, 1 round, low damage;
    CTG forced with the ghost pinned on six pad-edge spots so bots converge on the fence; XP tier rotated;
    48 matches per mesh, interleaved 8 wide):

    master this PR
    samples on the fence-top areas 255 465
    falls into the slot from the pad 2 (one from 8286) 0
    world deaths a match, whole map 4.98 4.96
    moves pad -> 8436 / 415 (the way down) 29 / 116 26 / 71

    Bots still reach the pad and still get down off it by 93 -> 8436; no fall from the fence tops in 48 matches.
    The counts are small: the edit removes the drop links and the side-step, it does not claim a rate change.

Review of NeotokyoRebuild#157: three one-way drops off the helipad fence, 4671 /
4674 / 4672 -> 415, cross the open kill slot west of the pad, and the
fence-top areas carry no CLIFF or PRECISE, so a bot on them can also
dodge or side-step into the slot.

Cut the three drops; the way down stays 93 -> 8436 -> 415. Mark
PRECISE + CLIFF on the 11 pad-edge areas whose footprint is at least
30 % over the open slot (4671 4672 4674 4676 4679 4733 4731 4677 4673
8286 1819), and on 3401, as its mirror 3404 is.

Gates: connections -3, no pocket, reachability or route change.
Pad probe, 48 matches a mesh: falls into the slot from the pad 2 -> 0,
world deaths a match 4.98 -> 4.96.

nav_analyze FRESH; same graph.
@sunzenshen sunzenshen changed the title ntre_apparatus_dm: cut the helipad fence drops and mark the fence tops ntre_apparatus_dm.nav: cut the helipad fence drops and mark the fence tops Oct 14, 2026
@sunzenshen
sunzenshen requested a review from sunmachine October 14, 2026 08:16
@sunzenshen
sunzenshen marked this pull request as ready for review October 14, 2026 08:16
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant