Entry 279

An enemy's condition effect outlived a Data Drain: clearConditionEnemy's ClearConditionEffect carried out

battlerender

Filescrates/piney-world/src/combat/mod.rscrates/piney-game/src/session.rs

The report: enemies' effects stay after they die, a floating staff and red marks after BATTLE MODE OFF. A character under a condition wears a ccConditionEffect (INF main particle.cpp). For a stat change (numbers 7-18 down, 23-34 up) it is:

  • particles that follow the character;
  • an effAbilityDown or effAbilityUp controller, the stat's icon.

Only two calls end an enemy's (xrefs, INF gcmn):

  • ccChar::DispConditionEffect (0x0056f950) kills it (it fades) or deletes it (it ends at once) as the conditions change. On a dead character (dead not 0 or 1) it kills it.
  • ccChar::ClearConditionEffect (0x00570180) deletes it and sets conditionNum (+0x30) to -1. Its enemy callers are clearConditionEnemy (0x004335d0), interruptThink and selectTarget.

clearConditionEnemy runs:

  • on the dying count;
  • on a freeze;
  • on a Data Drain (drainFlag), just before the enemy leaves the command lists (deleteCmnd) and its drained form is entered.

What the port did

The port's enemy AI returns Out::ClearConditionEffect from clear_condition_enemy, but no world code read it. The effect was left to the enemy's next DispConditionEffect.

A foe killed outright gets one: its frames run on through the death, and dead 2 makes it kill the effect in the same frame. That is why event 3's goblins, fought to the end and shot afterwards (after_a_fight_shots), left nothing.

A drained foe does not. It is off the lists at once, nothing of it runs again, and its effect was never ended:

  • the particles' generators, which follow the character's position and were never killed;
  • the icon's controller, unless its own target check (check_target) ended it. Whether it did was not checked.

With a goblin given attack down (temp[0] -10) and drained, its condition effect 7 was still live when the drain's windows closed.

The fix

Combat::disp_conditions_shown now takes, in order with the DispConditionEffect calls, each enemy's ClearConditionEffect. If the enemy has a live effect, it deletes it (deleteConditionEffect, a Show::ConditionEffect of CondFx::Delete).

a_drained_foe_keeps_no_condition_effect gives the goblins attack down in the fight of drain_in_a_fight. It checks that each one that wore the effect and left the lists has none. It fails without the fix: foe 52 kept effect 7. drain_in_a_fight's per-frame hook now takes &mut Session, so a test can change the fight.

Whether the report's staff and red marks were this is not confirmed. The icon of the attack-down change is one of effAbilityDown's rows, and its sparks are red. The report did not say whether the foe was drained.

Still unknown
  • Which enemy and area the report's staff came from, and whether it was drained. - Whether the icon rows of effAbilityDown include a staff (magic attack down): the rows were not rendered one by one. - A killed foe's effect now ends by the port's DispConditionEffect kill in the same frame as the game's delete. Which of the two runs first in the game's frame, and so whether the effect fades or ends at once, was not traced.