Obliterate cleave also works the same way with aspect of the turtle and tranquility. This type of DK cleave going through is actually pretty game breaking, honestly. I’m pretty confident at this point that unholy probably works the same way with dnd cleave. New strategy just save dnd for aspect of the turtle or tranquility and kill through.
Told ya, now imagine with a frost Dk go with pillars
There is just no way that this is how it’s actually intended to work.
test to see if dchaks non primary, chain lightning non primary, bolas, and any other splash/chaining/cleave abilities fall into the same bucket
if it’s the case id say there’s a good chance that turtle only blocks targeted damage with chain effects bypassing it unintentionally, because that’s what it sounds like
if not
i dunno, spaghetti code is called that for a reason
idk if it’d give definitive proof of any kind for some of them (namely dchaks/chain) because they may be outright programmed not to even try to hit anyone with turtle active but im guessing if that’s the case you’ll just not even be hit
also check to see if hi explosive deals damage too, i know that tar entrapment works through turtle too for some reason, but it’s a weird spell overall because it’s got its own root dr for some reason
Edit: Whoops, wrong video.
Chain lightning gets deflected.
the consistency of this god damn spell is staggeringly bizarre
Pretty sure death and decay cleave on pet goes through turtle as well, cant remember.
(like when you death and decay and it makes scourge strike hits all targets while you’re in it)
2 target coil may as well but I don’t think so.
It’s like the secret menu at wendy’s.
so it appears to be abilities that have non-targeted direct damage that cleave based on a status or modifier, but not chaining ones
maybe they’re all using coding to always hit?
not sure, looks to be the case
where’s cruisr with the simc dump
The dummy variable is just a similarity in the metadata between blade flurry and trick shots. The similarity existing in and of itself doesn’t necessarily prove that that variable is the cause of the difference in effect. And if that is the cause of it, which is possible, that kind of furthers the point that it’s not intended/buggy. The difference in whether something chains into turtle/tranquility shouldn’t be whether it has a dummy variable in it’s metadata to accommodate talent modifiers. That doesn’t follow a logical pattern or category in terms of game mechanics.
That’s popping pustules and doing a crazy amount of damage if that happens through our only major defensive while also adding more heal reduc. What a joke. Hunter’s bad and horribly bugged. Why even bother at this point.
It seems to be abilities that are amplified via talents to cleave a different amount of targets. Blade flurry has dancing steel, trick shots has light/heavy ammo, and dk has cleaving strikes. Sweeping strikes, chain lightning, death chakram, etc, all cleave a specific, constant number of targets, and so that effect is seen directly in their spelldata. The other abilities use a dummy effect because they may be altered by talents. From what I can tell, the dummy effect calls a script related to the spell.
Hypothetically speaking, if you had access to a test environment and a debugging console and could look at an actual stack trace of the abilities interacting to prove that this common factor causes the difference in behavior, would you not still come to the conclusion that this is a bug? Like in actual software QA, you look at a design document/list of requirements/user story and then test behavior against those to determine “does this work properly”.
“Does this ability have talents that can modify it” does not seem like it would be an intentional reason for otherwise similar spell effects to behave differently in similar circumstances as it pertains to bypassing an immunity or deflection. Like this is not an obvious category of spell. It’s still just a random list in practice.
It’s very clearly a bug, though the warbringer bug has been in since it got added so I doubt it ever gets fixed
The more I test this, the more I am willing to believe that I was overly confident and maybe wrong about turtle animation. I think it’s possible that the scenarios I thought that I was dying during the animation were actually interactions with things like I’ve tested here that inexplicably just go through it. Honestly though, this is kind of worse.
And like the spell-in-flight mechanic is also something that you can’t actually prove and parse out instance by instance because you’d need timestamped logs from another character’s account and the server. It’s still kind of baffling to me that these two abilities are the biggest interactors with spell-in-flight mechanics and that we can prove out so many buggy scenarios outside of spells in flight, yet there is still so much confidence that the spell in flight mechanic ALWAYS works properly with them. That’s still kind of backwards to me. I think they’re just busted.
You can prove it rather easily by looking at your own combat log. Spells that finish casting before turtle is activated will land through turtle.
Edit: I said earlier that cloak removes spells in flight, but I may be misremembering vanish, because when I looked it up, I found a reddit thread from 2020 that coil would hit through cloak if cloak was used after the spell finished casting.
Turtle and Tranq aren’t really unique in the way in flight projectiles ignore them, as spell reflection and potentially cloak have the same problem. The majority of projectiles are just spell based or unable to be blocked/parried, and there just aren’t that many defensives that interact with spells.
The rub comes into play when it’s really, really close. Like they effectively happen within milliseconds of each other. I can pause the video on reddit where the resto gets globalled by the enhance between seconds 0:05 and 0:06 and it really looks like the last ability/thing to hit is a lightning bolt which drops it below 5%, then it gets a tranq tick back up to 5, then the following windstrike still goes through with tranq active. You can see a distinct period of time between that lightning bolt landing and the next spell queueing up while tranquility is being cast. Without the ability to compare the timestamps as the server reads for the windstrike and tranq, you really cant definitively say one way or the other. It feels like people watch content like that and just immediately default to this in-flight mechanic that we think we understand, but really the abilities have all of these bugs in addition to that mechanic. It’s hard to see it and believe that it works properly, or exactly as stated.
Literally just look at the combat log. Whatever shows first in the log is what happened first. You can enable timestamps, but they are down to the second so that doesn’t really help. You can also /combatlog to keep logs of combat into a text file and can upload it to warcraftlogs to attempt to parse the information.
So then look at these examples. In particular, sunfire, starsurge, and arcane barrage which don’t have long travel times. Turtle is cast first, it’s really close, the spell goes through.
That’s not the combatlog. The difference being that in the combatlog, you can see spell cast success, and not rely solely on the instance of damage being done and trying to determine if the projectile could have had a possible travel time outside the time of turtle.