You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We found two apparent GCD discrepancies with the Festering Scythe talent in WoW retail 12.1.0.69933. We hope the maintainers can check these when possible. The evidence below is limited to two 180-second target-dummy runs; we have not measured the net DPS or cast-frequency impact of a fix.
In-game observations
An addon sampled C_Spell.GetSpellCooldown(61304).duration on UNIT_SPELLCAST_SUCCEEDED. This field is the reported total GCD duration, not the remaining time at observation.
Successful spell
Spell ID
Observations
Festering Strike
85948
39 of 40 snapshots: 0.750–0.845 s; one outlier: 1.218 s
Festering Scythe replacement
458128
39 of 39 snapshots: 1.090–1.236 s
The durations vary with combat state and are not proposed as fixed constants. The 1.218 s Festering Strike snapshot is included in the data rather than discarded; its cause is unknown. Combat-log SPELL_CAST_SUCCESS events corroborate 40 Festering Strike and 39 Festering Scythe casts across the two windows. The combat log does not itself record GCD duration, so the anonymized probe data is also included below.
Spell data and SimC behavior
Our pinned SimC source is b845947a34429874433d8e9362326894650dd20a with WoW spell data 12.1.0.69587. That spell data reports a base 1.5 s GCD for both 85948 and 458128. Talent passive 455397 applies a −500 ms flat GCD modifier through label 5916 to 85948 only; DK passive 137005 lists both spells under its hasted-GCD effect. Please verify these spell-data relationships against the current retail build, since the pinned data build differs from the client used for the measurements.
With the talent active, SimC first changes Festering Strike's trigger_gcd to 1.0 s. The generic melee-class rule then classifies it as non-hasted because the modifiedtrigger_gcd equals 1_s. In a one-iteration, 180-second debug replay, the Festering Strike entry action starts none GCD for 1.000. When the replacement is active, Festering Scythe executes as a background replacement action after that entry action starts the GCD, so it also effectively inherits the 1.0 s GCD. The same relevant code paths are present in the 2026-09-26 nightly source (a69b069) and later midnight source (7ffaabf).
Could you check whether (1) the generic 1 s non-hasted rule should distinguish an innate 1 s GCD from a 1.5 s GCD reduced to 1 s by a talent that remains hasted in-game, and (2) Festering Scythe should start its own hasted GCD rather than inherit the Festering Strike entry action's GCD?
The rows retain only run number, seconds relative to the combat window start, event type and spell ID. Player/realm names, GUIDs, targets, absolute timestamps, positions and unrelated events were removed.
The rows retain only run number, seconds relative to the first input, spell ID, reported total GCD duration, API modification rate and time elapsed between GCD start and observation. This table uses a different relative-time origin from the combat table, so their relative timestamps should not be paired row-for-row.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
[Unholy DK] Festering Strike / Festering Scythe GCD differs from in-game measurements
We found two apparent GCD discrepancies with the Festering Scythe talent in WoW retail 12.1.0.69933. We hope the maintainers can check these when possible. The evidence below is limited to two 180-second target-dummy runs; we have not measured the net DPS or cast-frequency impact of a fix.
In-game observations
An addon sampled
C_Spell.GetSpellCooldown(61304).durationonUNIT_SPELLCAST_SUCCEEDED. This field is the reported total GCD duration, not the remaining time at observation.The durations vary with combat state and are not proposed as fixed constants. The 1.218 s Festering Strike snapshot is included in the data rather than discarded; its cause is unknown. Combat-log
SPELL_CAST_SUCCESSevents corroborate 40 Festering Strike and 39 Festering Scythe casts across the two windows. The combat log does not itself record GCD duration, so the anonymized probe data is also included below.Spell data and SimC behavior
Our pinned SimC source is
b845947a34429874433d8e9362326894650dd20awith WoW spell data 12.1.0.69587. That spell data reports a base 1.5 s GCD for both 85948 and 458128. Talent passive 455397 applies a −500 ms flat GCD modifier through label 5916 to 85948 only; DK passive 137005 lists both spells under its hasted-GCD effect. Please verify these spell-data relationships against the current retail build, since the pinned data build differs from the client used for the measurements.With the talent active, SimC first changes Festering Strike's
trigger_gcdto 1.0 s. The generic melee-class rule then classifies it as non-hasted because the modifiedtrigger_gcdequals1_s. In a one-iteration, 180-second debug replay, the Festering Strike entry action startsnone GCD for 1.000. When the replacement is active, Festering Scythe executes as a background replacement action after that entry action starts the GCD, so it also effectively inherits the 1.0 s GCD. The same relevant code paths are present in the 2026-09-26 nightly source (a69b069) and latermidnightsource (7ffaabf).Could you check whether (1) the generic 1 s non-hasted rule should distinguish an innate 1 s GCD from a 1.5 s GCD reduced to 1 s by a talent that remains hasted in-game, and (2) Festering Scythe should start its own hasted GCD rather than inherit the Festering Strike entry action's GCD?
Relevant source: GCD classification, GCD start, DK replacement.
Anonymized evidence:
combat-casts.tsvThe rows retain only run number, seconds relative to the combat window start, event type and spell ID. Player/realm names, GUIDs, targets, absolute timestamps, positions and unrelated events were removed.
Anonymized evidence:
probe-gcd.tsvThe rows retain only run number, seconds relative to the first input, spell ID, reported total GCD duration, API modification rate and time elapsed between GCD start and observation. This table uses a different relative-time origin from the combat table, so their relative timestamps should not be paired row-for-row.
All reactions