Skip to content

BDSP AutoStory rebuild 2/6: maps, planner and walker - #456

Draft
dayfine wants to merge 13 commits into
bdsp/rebuild-detectorsfrom
bdsp/rebuild-walker
Draft

dayfine wants to merge 13 commits into
bdsp/rebuild-detectorsfrom
bdsp/rebuild-walker

Conversation

@dayfine

@dayfine dayfine commented Nov 13, 2026 •

Copy link
Copy Markdown
Owner

Depends on #455

This PR holds the maps, the route planner and the walker: AutoStory/Data (story maps, world, planner, route records and goldens), AutoStory/Utils (walk, planned route, route walk, wall walk, location, leg battle) and ScreenEvents, along with the map tooling in devtools/maps and its workflows. It replaces the map and walker parts of #134, #127, #129, #375, #400, #427 and #432. Since the first round it also has the lab's walker fix for a wall leg next to a wander zone (probe with A, wait, replan, from #432) and the hook that pauses before each leg for breakpoints.

Built SerialProgramsTests locally. The PokemonBDSP, shared Story and LabChannel suites ran 270 passed and 28 failed, all 28 being CommandLineTests image tests. Code health: 1 failure, carried from rebuild/src (PlannedRoute.cpp is over 1000 lines).

🤖 Generated with Claude Code

https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss

dayfine and others added 2 commits November 13, 2026 01:57
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
The segment-01 plans come later in the stack (S01), which adds their blocks
back; until then the golden test compares segment 00.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
@dayfine

dayfine commented Nov 13, 2026

Copy link
Copy Markdown
Owner Author

New commit f07981f: Data/route_plans.golden.txt now lists segment 00 only, since the segment-01 plans arrive later in the stack (#462 adds their blocks back). Data-only change; BF2a's own tests are unchanged.

🤖 Generated with Claude Code

dayfine and others added 11 commits November 13, 2026 05:26
A wall leg by a wander zone tells the wall from the wanderer (A-probe,
wait, replan; lab #432 697e769), leg breakpoints through RouteIO's
leg_started, and route_bench.golden.txt's two 01.10 rows.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
…iew fixes

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
…iew fixes

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
…red leg trace and replay (#464)

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss

# Conflicts:
#	devtools/maps/tests/test_leg_trace.py
…code moves to #467-#470

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
Port of a7dcd77. PlayerHouse1F mid-room frames show only 10.6-14.4% black
on an edge strip, so indoors() read them as outside. On 2,669 lab
screenshots warm outdoor frames never exceed 3% and rooms never fall below
10.6%. The rebuild had no indoors() test, so this adds the suite with the
big-room case.
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