Entry 215

Kite's, the streams' and the effects' tables generated

2026-09-27volumes

Filescrates/piney-gen/src/manifest.rscrates/piney-gen/src/layout.rscrates/piney-gen/src/dtype.rscrates/piney-data/src/tablescrates/piney-battle/srccrates/piney-stream/srccrates/piney-effect/srccrates/piney-game/src

The Infection boot got through the desktop but stopped entering The World. Every run-time image is GCMN.PRG over an empty executable, so every read of main's data failed:

• DamActuTbl, in Kite's tables;
• streamTblE, in every stream;
• effectCCSTbl, in the field's effects.

Three systems now read generated tables instead, all four discs'. An inventory of every run-time image read left in the crates (world towns, fields, dungeons, streams, effects, battle, the story announcements) set the order: what the boot reaches first.

The groups

The readers take of(volume): KiteTables, MotionTables, RideTables, SkeithData, table::Def, Subtitles, the stream effects' and ending's tables, Effects::new and StreamEffects::new with every spell's tables. The port no longer builds an image for any of them.

What the later discs showed

• Outbreak and Quarantine have no Japanese-voice streams. Their ccStreamInit builds streamTblE alone, and their gate hack has only the E tables. The Japanese-voice entries are absent there, and the readers take the E ones as the game does.
• Reading past a table's end. hitRot0120a, hitRot0240, particleTbl and the thunder's time lists are read on past their ends, as the game does. On a later disc the bytes there are that disc's own, so the vote by likeness to Infection chose wrong places. hit_rot_0120a landed a row off on Outbreak and Quarantine. Such an entry now takes the carry's place alone (.carried()).
• Only the thunder's time lists (their tail) differ between the discs. Every other effect table is identical on all four.

The generator

• Bitfields. ccParticleGeneratorParam packs gType, rType and dType into one halfword. A member with AT_bit_size is a Bits layout, read from its storage unit. Its write sets only those bits. DWARF 1 says AT_bit_offset counts from the most significant bit, but here it counts from the least.
• An end row kept. array_through keeps the row until finds: strSndTbl's BGM tables end on a record the cursor rests on.
• All-capital names. A DWARF name in capitals becomes CamelCase (FOOD_PARAM, STREAMDATA, CCSND_STR_BGM).
• Decoded code. ccParticle::Setup's switch is decoded from each volume's code in the generator, no longer at run time.
• piney-gen global NAME prints a global's DWARF type and its struct's members. Every table here was laid out with it.

Particle rows keep their identity as the volume's own address of the row (*_va entries), which the harnesses compare. The boss effects' rows stay keyed by Infection's addresses as their names.

Checks

• Kite's, the members', the ride's, Skeith's and the navigation's battle harnesses pass.

• The stream crate's tests pass: the stream fixture, the subtitles, the ending and the effects. So does the audio crate's stream test.

• piney-effect's tests pass, and effects_on_every_disc builds the field's and the stream's effects on all four discs.

• The effect harnesses all pass against the game run in eemu:

  harness	cases
  effect	6
  particle	8
  hit	3
  misc	9
  draw	2
  skill	6
  spell	5
  stream	1
  boss	4

• the_world_top_page_and_into_mac_anu passes. The World enters Mac Anu again.