Skip to content

FRLG AutoStory: breakpoints pause before each step and leg - #450

Draft
dayfine wants to merge 38 commits into
frlg/stage19-victory-roadfrom
frlg/story-breakpoints
Draft

dayfine wants to merge 38 commits into
frlg/stage19-victory-roadfrom
frlg/story-breakpoints

Conversation

@dayfine

@dayfine dayfine commented Nov 13, 2026

Copy link
Copy Markdown
Owner

Depends on #448 (merged in; base is #446's branch).

FRLG AutoStory now pauses at breakpoints: the BREAKPOINTS option (#448) reaches the FRLG run, which pauses before each step and, with "legs", before each leg of a route.

  • StoryExecutor gets the run's pause points from the panel, so each checkpoint pauses before its steps (<id>/<n> <step>).
  • run_route() pauses before each leg through RoutePlanning::pause, labelled like 12.03/5 pallet_shore leg 2/5 east 4 -> (10, 10).
  • Only a lab build waits; elsewhere the run goes on.

Testing

  • Unit tests: PokemonFRLG::OverworldRecovery - pause before each leg, Pokemon::StoryStepExecutor - pause points ..., StoryEngine, LabChannel suites pass in the lab build.
  • Not run on a console yet.

🤖 Generated with Claude Code

https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss

dayfine and others added 24 commits November 12, 2026 22:51
…tion, engine)

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
…kpoints (set_pause_points)

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

StoryExecutor sets the run's pause points on its checkpoints; the
BREAKPOINTS option reaches them through StoryEnv, and run_route() pauses
before each leg.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
…03 Pallet shore at (10, 16)

Every leg's "ended at" line (re-find legs too) is followed by a line with the
screenshot saved under Screenshots/AutoStory/route/, as BDSP does; the Fly
landing saves one too. 12.03 pallet_shore now targets (10, 16) and the surf
is declared after (10, 17): south, east, south, no west leg.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
The flight's black band, and Fuchsia before the fade, are quiet with
nothing showing and passed for the landing (console B, 2026-11-12 23:59,
12.03: "landed in PalletTown" 3 s before the Pallet banner). Landing is
now a fade that rose after the A, then a quiet overworld with a readable
facing; else fly_no_takeoff / fly_no_landing. turn_to() on a walking
sprite no longer takes an unreadable facing as turned: it rereads for
about 2 s, then fails the leg.

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

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
…on a map's edge tile isn't failed for an unread facing

12.03 failed at Fuchsia's Center (0, 5): after the turn tap east the
escalator's rail hid the player's legs and the facing couldn't be read.
The checkpoint now walks from the nurse straight south to the door warp
(user, 2026-11-13), without the corner event. turn_to() takes an unread
facing as turned where the route knows the player stands on a map's edge
tile, and fails the leg elsewhere as before. turn_to() takes its sprite
as an enum and moves with read_facing() to OverworldFacing.cpp.

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

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
…p-edge test as a case table

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

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
Console B 2026-11-13 02:42/02:46, 12.03: the fly map's close fade, then Fuchsia
quiet with the player on screen, was taken as the landing in Pallet Town.

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

The fly map's close is a fade, and the departure town after it is quiet with
the player on screen (console B 2026-11-13 02:42 and 02:46, 12.03: "landed in
PalletTown" on Fuchsia; the first leg then walked nothing during Pallet's
fade-in and every later tile was 2 rows off). The landing now also needs the
game screen to differ from the overworld shot taken before the Fly menu.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
dayfine added a commit that referenced this pull request Nov 13, 2026
Conflicts with #453's party cache: run_field_move_step takes both the player (Fly landing reads the facing) and the PartyCache; the departure snapshot is taken before the party is opened; StoryEnv keeps party/field moves/master ball and the pause points; keep_field_move_users and set_breakpoints both kept. 12.03 plan comes from #450 unchanged (Pallet shore (10, 16), straight out of the Center).
dayfine added a commit that referenced this pull request Nov 13, 2026
Conflicts: HealingSteps includes keep both sides; PlanGolden12.h is the lab's resolution of #450 + #463 (lab/dev-frlg-stage8 ea2a0c9), which merged the same two tips. 12.03 plan now matches the lab's console version (a3a3e01).
dayfine added a commit that referenced this pull request Nov 13, 2026
… the party menus for FLY

Conflict: run_field_move_step's comment in FieldMoves.h; kept #453's menu scan
and the Fly landing judgement from #450.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
dayfine added a commit that referenced this pull request Nov 13, 2026
…egs that land

12.03 takes #463's Route 21 legs (sand bars, then the beach) from Pallet's
shore at (8, 16), and keeps #450's walk straight out of Fuchsia's Center.
RouteCalls and the facing reader's tests keep both sides; the wanderer check
keeps 13.03's to_elixir and incidental fights next to #463's Swimmer Roland.
OverworldSim.h is LF on this line since the gates merge; #463's change is
applied to it without changing its line endings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
# Conflicts:
#	SerialPrograms/Source/Pokemon/AutoStory/Engine/Pokemon_StoryEngineRuns_Tests.cpp
dayfine and others added 12 commits November 13, 2026 14:52
# Conflicts:
#	SerialPrograms/Source/Pokemon/AutoStory/Engine/Pokemon_StoryEngineRuns_Tests.cpp
#	SerialPrograms/Source/Pokemon/AutoStory/Pokemon_AutoStory_Tests.h
#	SerialPrograms/cmake/SourceFiles.cmake
# Conflicts:
#	SerialPrograms/Source/Pokemon/Navigation/Pokemon_LegPlanner.h
Moved from the FRLG stack, where 0ac33f9 added it inside a game PR (FRLG
names its route screenshots by the step under way).

(cherry picked from commit 0ac33f9, Pokemon_StoryBreakpoints.h only)

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
TEXT_BOX -> DIALOG_BOX, BATTLE -> BATTLE_STARTED, CAN_MOVE ->
PLAYER_CAN_MOVE, STILL -> SCREEN_STILL; Pokemon:: screen event names
are now in PokemonAutomation (CommonTools/ScreenEvents).

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

Conflict in the FRLG observer: both sides dropped the Pokemon:: prefix;
take the incoming side, which also includes ScreenFades.h. The merge
moves this branch's FRLG screen event users off Pokemon:: (ScreenEvent,
ScreenEvents, ScreenEventLog, ScreenEventObserver, ScreenEventWatcher,
fade helpers), uses NintendoSwitch::SwitchHomeObserver from
NintendoSwitch_HomeScreenObserver.h, and calls the shared describe().

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

# Conflicts:
#	SerialPrograms/Source/VideoGames/PokemonFRLG/Overworld/PokemonFRLG_OverworldSim.h
#	SerialPrograms/Source/VideoGames/PokemonFRLG/Programs/AutoStory/PokemonFRLG_AutoStoryScreenEvents.cpp
#	SerialPrograms/Source/VideoGames/PokemonFRLG/Programs/AutoStory/PokemonFRLG_AutoStoryStepInterrupts.h
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