SKID.NC_ARMSCANNON skill no cast time. Not sure about KRO official behavior, but it has cast time before I updated to the lastest version of ra.:
Modifications that may affect results:
This is the oficial update of Arms Cannon Cooldown.
https://ragnarok.gamepedia.com/RO_Patch_(2018_Oct._17)
https://replay.divine-pride.net/Replay/d3e47f30
Made a recording for each level on a Mech.
Skill cool down added Oct17,2018, as noted by above patch link (verifiable using HEADER_ZC_SKILL_POSTDELAY packets):
- 암즈 캐논 스킬의 쿨타임이 추가됩니다.
1레벨 > 0.15초
2레벨 > 0.20초
3레벨 > 0.30초
4레벨 > 0.45초
5레벨 > 0.65초
After cast delay appears to be 1s for every level, can be verified by HEADER_ZC_MSG_STATE_CHANGE3 packets.
HEADER_ZC_USESKILL_ACK2's delaytime value appears to be the cast time (variable+fixed), but it's rather confusing to me. The Mech I used for that recording was 2 DEX, 3 INT. Yielded the following cast times:
https://replay.divine-pride.net/Replay/c77e63ad
Contains a replay after I upped my DEX to 131 and INT to 37. Delaytime values became:
Dunno if that's enough to figure out the Fixed+Variable of each level. I'll try to math it out later.
Scratch later, couldn't sleep until I solved it. I abused WolframAlpha by plugging in my cast times and stats as so:
F+V*(1-sqrt(7/530)=result1, F+V*(1-sqrt(299/530)=result2
Where F is fixed cast time, V is variable cast time
sqrt(7/530) was my stat-based cast time reduction on the first trial (DEX=2,INT=3, DEX×2+INT=7), while sqrt(299/530) was the stat reduction for the second trial (DEX=131,INT=37, DEX×2+INT=299).
result1 is the first trial result for that level, result2 is the second trial result for that level.
Doing that for all levels gave me the following cast time results: