Entry 179

"The hacked areas' noise, the game's DEG2RAD, and the event pages' unknowns"

script

Documented in Event scripts.

Filescrates/piney-game/src/area_host.rscrates/piney-game/src/field_host.rscrates/piney-event/src/vm/exec.rstools/evscript.pydocs/engine/events.mddocs/engine/event-vm.md

Going down the "Unknown" lists of events.md and event-vm.md.

The first play pass's noise. event-vm.md said play_pass_done's worldman +0xf0 == 3 case (ccMenu +0x104 = 1) was not modelled, and inferred that a town never has it.

  • ccThEvent (0x001b5eb0) was read: after the first play pass, WORLD_MAN.hackFlag == 3 sets ccMenu.interNoiz (+0x104) to 1, the menu's noise (field-ui.md).
  • The writes of hackFlag in WORLD_MAN::GO were found: before GO branches on the area, it sets 2 for every area, or 3 when the save's crisis byte (+0x6772) is set; a field or a dungeon then takes its area's own.
  • So the inference was wrong: in the crisis a town has the noise too.

The area host now turns it on from WorldMan.hack, which already holds 3 for a hacked area and for any in the crisis. The town host turns it on from the crisis byte.

DEG2RAD. Also an unknown: whether the event VM's deg2rad (event positions' headings) equals the game's. It computed pi * v / 32768 in the host's IEEE floats. The game's DEG2RAD (0x001dabb0) was run in tools/eemu.py for all 65,536 shorts:

  • The IEEE form is one bit high for 32,680 of them.
  • The EE-arithmetic form the world and battle crates already use (piney_data::field::ee mul and div) equals the game's for all of them.

The VM now uses the EE form, and deg2rad_is_the_games pins four of the values that differed.

Numbering and meanings that were already known. Several items only needed a pointer to where the answer lives:

  • The board states: 0, 1 new, 3 read, 7 the player's own post waiting to be written out. The top page's port has them.
  • emode: the low byte picks how the window ends (0 closes, 1 chains, 2 stays up, 3 asks); 0x100 shows the text untyped; 0x200 is a typing test with no effect. The desktop and field UI pages have these.
  • The camera: its angles are DEG2RAD shorts, its distances tenths, and cam_mode4 (cpCtrl 4) holds the camera point while the look-at goes on. piney_world::evcam, checked against the game, has this. The opcode table's "mode not traced" is gone, in tools/evscript.py too.
  • WORLD_MAN::SetEventData "not read": dungeon.md documents it and the area mode runs it.

events.md's unknowns are now only the later volumes' (their BSS message groups, OUT's rewritten cases, the KFED thread).

Still unknown

The crisis town's noise was not seen in a picture; it follows from the code. Whether any Infection field is hacked outside the crisis (EVENTAREA_INFO.flag 3) was not surveyed.