Repository navigation
PokemonBDSP AutoStory: read movement times and taps from the shared table - #404
Conversation
Tap lengths, tile times per gait and axis, start delay and how open legs are pressed (BDSP taps, FRLG timed holds on foot), with an entry for BDSP and one for FRLG and a check of each. 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
route_ms_per_tile(), the walker's step taps and the route timing and turn tap constants now come from Pokemon::BDSP_MOVE_MECHANICS; values unchanged. 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
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
dayfine
left a comment
There was a problem hiding this comment.
Reviewed SHA: 88bb10a
Reviewed against bdsp/autostory-v4. The three Pokemon_MoveMechanics* files in the diff are #402's, merged in (byte-identical to #402's head), and are reviewed there.
Verdict
APPROVED (LGTM to merge, after #402)
Correctness: every replaced number has the same value (160ms/640ms -> step_tap/step_tap_release; 800ms chunk = 160 + 640; ROUTE_* aliases = table rows), and route_ms_per_tile() returns the same thing via TileTime::along() (EAST/WEST -> across, else up/down, as the removed across() helper). No planning code changes, so the plan golden is untouched. Code Health: 0 failures (the RouteTiming.h header-size bypass predates this PR).
Findings (optional, non-blocking)
- (Nit)
PokemonBDSP_AutoStory_RouteWalk.cpp:Pokemon/Navigation/Pokemon_MoveMechanics.his included afterPokemon_RouteLog.h; keep that block alphabetical. - (Nit)
RouteWalk.cppalready hasusing namespace Pokemon;, soconst Pokemon::MoveMechanics& game = Pokemon::BDSP_MOVE_MECHANICS;can drop both prefixes. Same inWalkUntil::press(), where a localgamealias like the one inrun()would read better than spellingPokemon::BDSP_MOVE_MECHANICStwice. - (Follow #402) If #402 moves the step tap onto
Gait(finding 1 there), the three reads here becomeBDSP_MOVE_MECHANICS.walk.tap…; a mechanical follow-up.
Notes
- Fork Draft CI is red from the base: the test run aborts at registration with
Duplicate Test Name: PokemonBDSP::ShortDialogDetector - …IntroRowanNameTab_True.png, the same onbdsp/autostory-v4itself. So no unit test ran in CI; the golden result rests on the local run in the description.
…e messages name expected and actual (#402 review) FRLG's Bicycle taps 40 ms at one per 100 ms, its walk 66 ms with a 700 ms release, so the step tap moves from MoveMechanics onto Gait. Every scalar gets a default, and the test reports which gait and which value failed. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
Follows #402 moving the step tap onto Gait; same 160/640 ms, so presses are unchanged. Also sorts the MoveMechanics include and drops prefixes the file's using-declaration covers. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss
|
Updated at 8c66bc8 (fix: 8c66bc8 on top of merging #402 7f63e8b and
Local headless run (Pokemon::, PokemonBDSP::AutoStory, PokemonBDSP::StoryData): 464 passed, 0 failed. That includes "plans match the golden". Code health: 0. |
dayfine
left a comment
There was a problem hiding this comment.
Reviewed SHA: 8c66bc8
Delta since my review at 88bb10a: the merges 44c5435 (base) and 7c4607f (#402 at 7f63e8b), and the fix 8c66bc8. The PR's own BDSP files differ from 88bb10a only by 8c66bc8: the base merge left RouteInput.h and RouteTiming.h unchanged. The Pokemon_MoveMechanics* files match origin/nav/grid-story-steps (#402 merged there) byte for byte, so they were reviewed in #402.
Verdict
APPROVED (LGTM to merge, after its parent chain, as before)
- Correctness: the three reads now use
BDSP_MOVE_MECHANICS.walk.tap.{hold,release}, which is{160ms, 640ms}in the table. That is the same as the oldstep_tap/step_tap_release, so the presses and the 800 msTAPSchunk are unchanged. Nostep_tapreference is left in the tree.WalkUntil::press()braces itscasefor the localtapand stillreturns inside it. - Earlier nits: both resolved. The include block is sorted, and
RouteWalk.cppdrops thePokemon::prefixes. Finding 3 (follow #402) is done. - Checks: Fork Draft CI on 8c66bc8 is green. The duplicate-registration abort is gone and the suite runs, including
Pokemon::MoveMechanics - BDSP/FRLGandPokemonBDSP::AutoStory - Simulation: plans match the golden. Code Health passes. The red "Runtime map reader" (native-schema) job fails on the base too and is not from this PR.
Findings (optional)
- (Nit) In
Walk.cpp,using namespace Pokemon;isn't in scope (unlikeRouteWalk.cpp), so thePokemon::prefixes there are correct. The sameconst Pokemon::StepTap& tap = Pokemon::BDSP_MOVE_MECHANICS.walk.tap;is now spelled in bothrun()andpress(). A private member or a file-local constant would say it once. Not blocking.
Depends on #402.
Base:
bdsp/autostory-v4, withnav/move-mechanics(#402) merged in; until #402 lands there the diff also shows its three files.BDSP AutoStory now takes its walking and running tile times, start delay, step tap, step tap release and turn tap from
Pokemon::BDSP_MOVE_MECHANICS.route_ms_per_tile(), the walker's tap chunks and the planned route's corner taps read the table directly;ROUTE_MS_PER_TILE_*,ROUTE_RUN_MS_PER_TILE_*,ROUTE_START_DELAYandROUTE_TURN_TAPare now defined from it, so their other readers are unchanged. All values are the same. Tested in the headless runner with thePokemonBDSPandPokemon::MoveMechanicssuites: the plan golden (#392) passes, and the only failures are image tests without fixtures.🤖 Generated with Claude Code
https://claude.ai/code/session_013VUubhHeYSMGSSDXZkKZss