[Resolved] ATTACK_MOTION_TIME_TERM is misconfigured

I've brought this up in a maintenance thread before and am aware that it has already been forwarded to kRO but I feel it deserves a thread of its own to raise some awareness.

Based on my testing, it would appear that iRO has its ATTACK_MOTION_TIME_TERM value set to 100. It's supposed to be set to 70 to enable true 193 ASPD for skills.

What is ATTACK_MOTION_TIME_TERM and why should I care?

AMTT is a configuration option for the server and is a part of determining the minimum possible delay between skill casts. To be more clear, it is added onto the character's ASPD delay. For example: your skill delay is 70ms at 193 ASPD, which then becomes 170ms after adding AMTT. This means your actual max skill casts/sec is 5.88/sec. AMTT has to be set to 70 in order for it to be 7 skill casts/sec.

Because many have misunderstood this before, it is important to note: you still need 193 ASPD to reach the cap of 5.88/sec. If you have 192 ASPD your cap is less than 5.88/sec.

Visual comparisons:

Rolling Cutter spam on iRO

Rolling Cutter spam on kRO (this person has 192 ASPD and is casting much faster than the clip from iRO)

Dragon Breath spam on jRO (no iRO equivalent to go with this but anyone playing DB RK can confirm that this is not possible on iRO)

Info from analyzing replays with different server side AMTT values. Special thanks to Dia from divine-pride.

NOTE: the client these were recorded on did not support hold-to-spam for Rolling Cutter and as such were recorded using an unoptimized keyboard macro which is why there are fluctuations

ATTACK_MOTION_TIME_TERM = 200
2024-02-06 18:10:46.476 (00:00:01)    +0 ms      HEADER_ZC_USE_SKILL2
2024-02-06 18:10:46.793 (00:00:01)    +317 ms    HEADER_ZC_USE_SKILL2
2024-02-06 18:10:47.109 (00:00:02)    +316 ms    HEADER_ZC_USE_SKILL2
2024-02-06 18:10:47.371 (00:00:02)    +262 ms    HEADER_ZC_USE_SKILL2
2024-02-06 18:10:47.648 (00:00:02)    +277 ms    HEADER_ZC_USE_SKILL2
2024-02-06 18:10:48.019 (00:00:03)    +371 ms    HEADER_ZC_USE_SKILL2
Average of 308 milliseconds between per skill cast or in other words, 270 + fluctuations.

ATTACK_MOTION_TIME_TERM = 50
2024-02-06 18:08:38.139 (00:00:03)    +0 ms      HEADER_ZC_USE_SKILL2
2024-02-06 18:08:38.276 (00:00:03)    +137 ms    HEADER_ZC_USE_SKILL2
2024-02-06 18:08:38.457 (00:00:03)    +181 ms    HEADER_ZC_USE_SKILL2
2024-02-06 18:08:38.608 (00:00:03)    +151 ms    HEADER_ZC_USE_SKILL2
2024-02-06 18:08:38.748 (00:00:03)    +140 ms    HEADER_ZC_USE_SKILL2
2024-02-06 18:08:38.919 (00:00:03)    +171 ms    HEADER_ZC_USE_SKILL2
Average of 156 milliseconds between per skill cast or in other words, 120 + fluctuations.

Info from analyzing replay of [Rolling Cutter spam on iRO]

Timestamp of first cast: 56:02.763
Timestamp of final cast: 56:12.121
Total amount of casts: 53

Time elapsed: 9.358 seconds
Skill casts/sec: 5.66 (53/9.358)

Some of the other regional servers also have it misconfigured. Confirmed ones include thRO and bRO.

Edited by Zayaan, 27 August 2024 - 08:49 PM.

please god fix this asap

The crazy part is it has been misconfigured on iRO since the very beginning of renewal.

Bump for this. It's very important for many skill cast which has no cooldown. What I saw in other server's video is that they can cast skill with no stuck but in IRO it is. 

@IROBeelzebub

The crazy part is it has been misconfigured on iRO since the very beginning of renewal.

I remember when Renewal came to iRO and I saw some kRO videos of people spamming skills and thinking, "the hell, iRO's spam speed is nowhere near as fast as kRO's."

The old management (CM Oda and CM Campitor) were made aware of this issue very early on after Renewal's implementation, but they had absolutely no idea how to fix it nor where to look. They did try; they simply didn't know where the setting was that made it different between kRO and iRO.

Now that it is known, can we please have ATTACK_MOTION_TIME_TERM adjusted to what it is supposed to be and have proper skill spam on iRO? We've essentially been waiting for nearly 14 years now for this to be corrected.

Edited by Jerbear, 11 June 2024 - 05:34 AM.

[Edited] As of right now the issue its on a decent state, as the spikes happen once 5-8 minutes. Nowhere near the frequency at which happened before, so i guess you can call it "fixed". I mean it still occurs its just noticeable if you taking deltas between CZ_PING in long periods of time.

Edited by SpookyFenix, 14 June 2024 - 05:59 AM.

Bumping to show agreement and support. 

Hello ,

I've forward this issue to HQ , 
I hope this can be fixed soon ,

Regards,
Beelzebub

actually based if this gets fixed 

please fix the other broken and missing stuff too god please

Hello ,

I've forward this issue to HQ , 
I hope this can be fixed soon ,

Regards,
Beelzebub

It's been over 3 months since you initially forwarded it. HQ sure is taking their sweet time for something that is trivial to test and confirm. Would a clear visual demonstration of the difference it makes help move things along?

Edited by Zayaan, 24 June 2024 - 02:58 PM.

Test results from a local server, in order of highest to lowest:

ATTACK_MOTION_TIME_TERM = 500

ATTACK_MOTION_TIME_TERM = 100

ATTACK_MOTION_TIME_TERM = -100 (to emphasize, that's a negative value!)

The one and only difference between these 3 clips is the ATTACK_MOTION_TIME_TERM value. That being the case, we can safely conclude that it directly affects skill delay, made even more evident by the third clip as a value of -100 completely removed the delay.

Hello,

Finally we win after arguing with HQ,
Will fix this soon ,

Thank you

Hello,

Finally we win after arguing with HQ,
Will fix this soon ,

Thank you

Thank you so much! Excited for this update.

According to this week's patch notes this will be changed to match kRO/twRO. I will post a comparison before and after when maintenance is finished along with a replay analysis.

And at the risk of celebrating early, I would like to extend a massive thank you to Dia and someone who probably wishes to stay anonymous. Without them this would likely have stayed the way it was for much longer.

Hello

This should be fixed now ,

anyway lemme know if you find something about iRO that seems wrong or not correct compared to KRO ,

Regards,
GM Bee

Hello

This should be fixed now ,

anyway lemme know if you find something about iRO that seems wrong or not correct compared to KRO ,

Regards,
GM Bee

Doesn't seem to have changed anything.

Before maintenance
Timestamp of first cast: 50:33.154
Timestamp of final cast: 50:47.155
Total amount of casts: 78

Time elapsed: 14.001 seconds
Skill casts/sec: 5.57 (78/14.001)

-----------------------------------

After maintenance
Timestamp of first cast: 22:58.519
Timestamp of final cast: 23:11.598
Total amount of casts: 71

Time elapsed: 13.079 seconds
Skill casts/sec: 5.42 (71/13.079)

Edited by Zayaan, 03 July 2024 - 10:29 AM.

Hello,

We will check again tmr , im quite sure the patch is done by HQ

Hello,

We will check again tmr , im quite sure the patch is done by HQ

Unless the testing methodology is flawed (which doesn't appear to be the case), it's quite weird that patching it did not show any changes. Based on ZP decompilation provided by Dia, there shouldn't be anything else affecting it. Plus testing on local servers showed promising results.

Edited by Zayaan, 03 July 2024 - 10:43 AM.

Hello,

We will check again tmr , im quite sure the patch is done by HQ

Well, be sure to check the configuration file to see that it's set to 70 correctly tomorrow.

We're still experiencing only 5 skills per second on iRO right now as opposed to the correct 7 skills per second as kRO does. So either the patch didn't go in correctly to the server or ATTACK_MOTION_TIME_TERM is still configured at 100 and it wasn't actually changed.

Oh good, i'm not crazy. Was testing if I felt any faster and was like "this seems the same as before"

Hello,

We will check again tmr , im quite sure the patch is done by HQ

Hello,

Confirmed by HQ its patched to our server and the setting was changed from 100 --> 70,
I will investigate this,

Anyone have tried it on Thor or Freya server?


*Note the delay on 2nd July MA was because we are fixing the MD disconnection issue , not the attack motion


*i apologize if my English is not good

Hello,

Confirmed by HQ its patched to our server and the setting was changed from 100 --> 70,
I will investigate this,

Anyone have tried it on Thor or Freya server?


*Note the delay on 2nd July MA was because we are fixing the MD disconnection issue , not the attack motion


*i apologize if my English is not good

If there were any tests done it would be helpful to know what's causing the discrepancy