Entry 118

"Smoke and dust: effSmoke, the enemies' feet, a gold goblin's run and an idol opening"

2026-09-25battlerendertest

Filescrates/piney-effect/src/dust.rscrates/piney-effect/src/lib.rscrates/piney-effect/examples/particle_probe.rscrates/piney-battle/src/enemy_motion.rscrates/piney-battle/examples/battle_probe/enemy_motion.rscrates/piney-battle/examples/battle_probe/spawn.rscrates/piney-world/src/combat/dust.rscrates/piney-world/src/combat/mod.rscrates/piney-world/src/combat/stage.rscrates/piney-world/examples/dust_probe.rscrates/piney-game/src/fx.rscrates/piney-game/src/session.rstools/test_effect_particle_rs.pytools/test_dust_ctrl_rs.pydocs/engine/effects.mddocs/engine/particles.mddocs/engine/battle.md

The battle's rules already raised three dust calls, and nothing drew them:

• Call::DustCtrl, from the races' exclusive() and their footstep notes;
• Call::Dust, a gold goblin running;
• Out::DustRing, an idol opening (116 left it).

All three end in effSmoke, which was not ported either.

effSmoke and the dust (piney_effect::dust)

• effSmoke (main 0x001ce300) is one particle on effect.cpp's static ccpgSmoke. Its life is shortened by up to half at random, and it takes the velocity, size and fades it is given, plus a random turn.
• ccEnemyEffDust (gcmn 0x0043a3e0) makes n puffs at random headings (ccRandF(pi)), 10 s out, flying out and up.
• ccEnemyEffDustRing (gcmn 0x0043a590) makes n puffs evenly round from a random start, r out. Under 4 it is the plain dust instead.
• The character overload (0x0043a7a0) places the ring at the character's feet (FW2LW) plus its turned offset.

The enemies' feet (piney_world::combat::dust)

ccEnemyDustCtrl keeps a node per ccEnemyDustInfo row with the last frame bucket it fired in. ctrl(obj, flag) fires a row when:

• the enemy is shown, its anmNum is the row's (2 and 3 match either), and its frameNum is inside the row's window;
• and either:
  • on a note, for rows with every 0;
  • every frame, when the window is one frame;
  • once per bucket of every frames otherwise.

The rules now send what ctrl reads with the call (DustAt: the rows, dispSW, eneSmoke, anmNum, frameNum, pos, dirc). The rules' harnesses print only the flag, so they still match.

The controller draws nothing random. The runtime runs it in the effects' pass over the frame's shows (Combat::fx_call), in order: a race constructor's Out::Dust sets the nodes, and each DustCtrl is followed by the Show::DustRings it raised. fx.rs starts:

• those rings;
• the gold goblin's dust (FW2LW, then ccEnemyEffDust(p, 1, size, 6, eneSmoke), arguments checked in its exclusive());
• the idol's ring ((0, 800, -530), 4, 100, 8, 30, 132).

Checked

• test_effect_particle_rs.py's test_smoke_and_dust. The game's effSmoke, ccEnemyEffDust and both rings run in eemu over the real particle system, beside particle_probe. The machine's genrand and the probe's are now the same MT19937 seeded alike. Ten cases in towns and fields: random sizes, lives, textures and ring counts, a character turned, the puffs run out. 904 frames of every particle's whole state match. The whole particle harness (15,263 frames) and the skill harness still pass.
• tools/test_dust_ctrl_rs.py (new). It lays a controller out in eemu over three races' real tables and a made-up one (rows of anm 2 and 3, a one-frame window), and runs the game's ctrl for 3000 steps each: animations changing, frames advancing, notes, dispSW off now and then. dust_probe runs the port's ctrl alongside. Every step's rings match, and every row fires somewhere.
• event_3_plays_its_fights now also asserts dust. The portal's goblin (row 130, type 1) has one row, its fall (anm 10, frame 35), and the fight raises exactly that one ring.
• The workspace's tests, clippy and fmt pass.