-
Rydian wrote on 2015-06-06 19:24
^ What it says on the tin.
See below posts for testing.
-
X wrote on 2015-06-07 03:12
You should use Pale to do (new and more) tests. The SkillPrepare packet from the server contains the cast time as the second parameter (an int). You could also use the timestamps on the packet to measure lag. For example, on NA, I have about 200ms of lag. This leads to 400ms total when charging (once to start, once to stop).
Also since the combat system renovation, skills take all their MP right away, so the lag-based MP drain issue is no longer relevant.
To illustrate the effect of network lag on charging skills, this a video taken on an Aura server, with only one person, hosted on localhost, aka as 0 network lag as you can ever get:
[video=youtube;qhi7HSFxtxM]https://www.youtube.com/watch?v=qhi7HSFxtxM[/video]
-
Rydian wrote on 2015-06-07 06:53
hfgmjhksjk;sjs
Wish pale came with easy labels for known arguments (like naming packets).
Starting at the incoming SkillPrepare and ending with the incoming SkillStackSet.
Normal: 1395 from server.
20.821 - 22.315 timestamps.
1.494 (+0.099)
Vivace: 1196 from server.
38.232 - 39.526 timestamps.
1.294 (+0.098)
Wand: 1228 from server.
14.082 - 15.409 timestamps.
1.327 (+0.099)
Both: 1053 from server.
52.117 - 53.268 timestamps.
1.151 (+0.098)
Yup, much more reliable results than client-side animation timing, though with these tests I rebirthed into a different talent with 870 int right now.
So if the base time is 2 seconds and I have both a -15% (skill) and ~16% (int) reduction, I should cast at 69% base time if they add up. That of 2 seconds is 1.38 seconds, so it seems that the bonuses from Int and Magic Weapon Mastery are added together.
Now 1.395 with a 14.3% reduction from Vivace rounded makes 1.196. Nice, works out properly, Vivace is added after the skill+int reduction.
The wand bonus of 12% works the same way, 1.395 with a 12% reduction rounded is 1.228, so it's also added after the base values.
And it seems that the Vivace bonus applies after the wand bonus when both are used.
So it's...
1 - Skill+Int reduction.
2 - Wand/equipment reduction.
3 - Vivace reduction.
That seem right so far? I don't have a casting speed reforge to test.
Is there a page on the wiki that has an existing cast speed formula or something to check over and possibly edit?
-
X wrote on 2015-06-07 07:42
Technically, total lag is from the outgoing skillprepare to the incoming skillready. Between those two, the client thinks you're charging and won't let you fire. It should only be a few MS more than what you have, though.
-
Rydian wrote on 2015-06-07 08:44
Yeah but I'm more interested in what the game calculates, the lag thing was just to note why I did multiple tests and took the average before.
-
Darkboy132 wrote on 2015-06-07 08:46
Quote from Xcelled194;1277517:
Also since the combat system renovation, skills take all their MP right away, so the lag-based MP drain issue is no longer relevant.
I'd like to know why Lightning Rod's MP usage is "unstable" then...
-
X wrote on 2015-06-07 20:27
Does {insert skill} take all its MP at once? Then it is not unstable, or instability is built into the skill using a RNG.
If it takes MP over time, it's susceptible to lag based variation due to lazy coding and lag.
I don't have experience with lighting rod so I don't know which it is.
-
Rydian wrote on 2015-06-07 20:31
Lightning Rod takes the MP when you fire it, not while you charge (which is weird for a magic skill), not sure if there's a start and end fire period though.
-
Darkboy132 wrote on 2015-06-07 21:10
Lightning Rod uses MP for firing it, but its depleted "over time" and not instantaneously.
-
Rydian wrote on 2015-06-16 16:24
Alright, so in really laggy periods you will still use more MP than needed... client-side. The server corrects your current MP next time it's touched at all.
-
Esileruf wrote on 2015-06-16 19:21
Quote from Rydian;1277538:
Is there a page on the wiki that has an existing cast speed formula or something to check over and possibly edit?
I can attest to a No, for this. Because it really wasn't looked into and not readily found. I ended up finding it on someone's userpage and the reason why its not on any other page is because its excruciatingly simple.
Base Cast Time * (1 - Cast Speed Boost) * (1 - Cast Speed Boost) etc etc etc.
If you had all of them it would come out like this:
Base Cast Time * (1 - Intelligence Cast Speed Boost) * (1 - Vivace Cast Speed Boost) * (1 - Wand Cast Speed Boost) * (1 - Reforge Cast Speed Boost) * (1 - Potion Cast Speed Boost)
For example, for Firebolt:
1.5 * (1 - .25) * (1 - .3) * (1 - 0) * (1 - .3) * (1 - .2†)
1.5 * .75 * .7 * 1 * .7 * .8
Resulting in 0.441 with 1250 int, 30% vivace, No inherent wand boost, Cast Speed 20 Reforge, and a Casting Speed Potion.
†: I don't know if this is exact or not.
-
Rynn wrote on 2015-06-17 13:30
Since when does INT affect firebolt casting speed?
-
Esileruf wrote on 2015-06-17 15:46
Since when doesn't intelligence affect all casting speed? Unless I slept through a mage nerf that dictated that bolts somehow are exempt from the mana reduction and cast speed decrease that you naturally get from intelligence.
EDIT: I'd actually like to know for sure if it does/doesn't. My firebolts have always been faster than they should be or as fast as they're calculated with int bonus.
-
Rynn wrote on 2015-06-17 16:04
Afaik I've never seen intelligence bonuses be mentioned for anything besides INT~ADV magic.
[Image: http://i.imgur.com/gyQ0gap.png]
And my firebolt when using a wand that is not casting speed reforge feels like the slowest thing in existence, even with 1500 int so I've never
felt like it does.
I also would like to know for sure so +1 to your request.
-
callback wrote on 2015-06-17 19:32
Quote from Rynn;1278857:
Afaik I've never seen intelligence bonuses be mentioned for anything besides INT~ADV magic.
[Image: http://i.imgur.com/gyQ0gap.png]
And my firebolt when using a wand that is not casting speed reforge feels like the slowest thing in existence, even with 1500 int so I've never felt like it does.
I also would like to know for sure so +1 to your request.
It should be a noticeable difference going from near max to 0. Just berserk and prove bolts are the same.