Entry 313

Three player reports: a scroll's cast, the Book of Law, a story box's body

2026-09-30uibattleworld

Filescrates/piney-battle/src/world.rscrates/piney-battle/src/entry.rscrates/piney-battle/src/gimmick.rscrates/piney-world/src/combat/stage.rscrates/piney-world/src/lib.rscrates/piney-fieldui/src/lib.rscrates/piney-fieldui/src/ctrl.rscrates/piney-fieldui/src/menus/useitem.rscrates/piney-game/src/world.rscrates/piney-game/src/area.rscrates/piney-game/src/session.rscrates/piney-game/src/session/tests/fairy_orb.rscrates/piney-game/src/session/tests/boxes.rsBUGS.md

GitHub issues 10, 11 and 12 (Infection). Each case was played through the session first, as the report describes it, and each test fails without its fix. Addresses are INF gcmn unless marked.

Issue 10: a scroll cast with no hold and no animation

A spell scroll (category 11) goes ccUseItemRequest (0x0057aa80) → ccItemSkillRequest(pw, tp, sid, 1), stype 2. _ccSkillRequest stores skillID, skillStatus 9 and targetChar = tp (0x00572b80). Kite's AnimCtrl (0x005995c0) then plays act 17-21 by the skill's type bits (0x100 → 17) and holds him (restraintSW, stopFlag).

This is issue 8's cause: before 015c996 the port left targetChar as it was. Kite with no target of his own (his last one fell, or nothing struck him yet) dropped the skill before its act. The spell itself still landed, through the item's own skill run, which is what the report describes. The reporter's issue (15:46Z) predates that commit (16:31Z).

a_scroll_on_a_foe_is_cast_by_kite: itemTblD row 0 (skill 193) on the goblin from PERSONAL, Kite's targetChar cleared until the use. With 015c996's line taken out the goblin is hurt and act 17 never comes; with it he plays act 17, held, and the act ends.

Issue 11: the Book of Law did nothing

Key item 60 (61 and 69 too) at 0x0057c02c: OpenInfo of installWarnStr's three lines (main 0x00377e80), 8 frames, ccSeOnNote(196, 52) (0x0057c0c4), 8 frames, until OK, Close. The port had these steps, but:

• in a town (where the book is got) WorldMode answered every use with no steps: no window, no sound. Books (category 12) and epitaphs used in a town did nothing either.
• in a field the menu handed SeNote to the world, which dropped it.

Fixes: World::use_item runs the use on the town party's scene (Kite's stand-in the user; a book's stat lands in his record, stored at the frame's end); World::item_step sets his pauseSW; the menu plays the note (Request::SeNote). Tests: the_book_of_law_warns_in_town (before: neither), the_book_of_law_warns_in_a_field (before: the window only), a_book_read_in_town_raises_the_stat (before: the book spent, the stat unchanged). The text comes from the build's fieldui group.

Issue 12: boxes without collision

The screenshot: B2 of Δ area 18, Kite standing inside a treasure box. DUNGEON::SetItemBox (0x005beb30) makes each GIMMICKDATA type-0 row a box at (x, y, 250), entRoot 0, land -1. A box outside Kite's room can not be landed then; initObject (0x00430ec0) switches its body on when he enters (SetHitSW(1), body still at z 250), then sets the box's pos on the ground, not the body's. boxMain ends with bodyHit.pos = pos (0x0045410c; objectMain 0x004543e0, virusMain 0x004545a4). The game's list links the ccCharHit itself, so from that frame the body is on the ground.

The port's list keeps copies, and nothing refreshed a box's. So every story box outside the first room kept a body 250 up, which Kite's body never meets. Random boxes and a first room's are made on the ground and were fine. Fix: World::hit_sync, called by the boxes' frames (gimmick::body_follows); Stage refreshes the list's copy.

• area_18_s_boxes_stand_in_the_way walks every room of area 18. Before: three boxes loose, on floors 1, 2 and 4.
• kite_stops_at_a_story_box_on_b2 runs Kite at room 4's box on B2. Before: he came within 9.9 of its centre (radius 40).