Remaining known DotA 6.83d bugs
-
Frank
- Treant
- Posts: 274
- Joined: Thu Sep 13, 2012 12:54 pm
- Has thanked: 18 times
- Been thanked: 35 times
Re: Remaining known DotA 6.83d bugs
Its usually happening when teamfights occur so I been box selecting my hero again quickly but i think it just deselects my hero not sure 100 percent ill try to pay more attention next time it happens
https://entgaming.net/findstats.php?id=12256907
a game where it happened I even said it in chat at around 28 minutes elapsed time as earthspirit
https://entgaming.net/findstats.php?id=12256907
a game where it happened I even said it in chat at around 28 minutes elapsed time as earthspirit
-
Hunter91
- Treant
- Posts: 437
- Joined: Sat Jun 16, 2012 1:27 am
- Has thanked: 5 times
- Been thanked: 11 times
Re: Remaining known DotA 6.83d bugs
h3rmit wrote:That's... very weird. Nothing in relation to unit control should have been affected. All changes, including Lanaya were very specific and very limited, and none of them had any distant relation to unit control.
How would you describe this "loss of control"? What is it that you're experiencing?
Also, I should clarify the following: Warcraft III is actually a "turn-based" game. In each turn, each player can choose to do one action, or pass. Afaik, the bot latency is currently set to 100ms. Meaning that each "turn" has a duration of 100ms, and you're allowed to send up to 10 actions per second. Issuing more than one command in the same turn results in overwritting the previous one.
What you see on your screen is a playback of your past actions, after you send them to the bot, the turn ends, the bot publishes what actions were taken by everyone, and you receive them. The playback of your actions starts then.
Is there any chance that you're temporarily exceeding the allowed 600 apm?
P.S.: The only other possible explanation I can think of is that some mysterious and undocumented bug or limitation of WC3 1.30.1 was triggered by recent changes.
We could verify if it's latency related or not, by testing the map on a bot with a very low latency (e.g. 20ms) that can fully cover your high apm, if your ping can afford it.
I think it is due to a few skills that cause this including Astral Imprisonment from Obsidian Destroyer.
-
h3rmit
- Armored Tree
- Posts: 87
- Joined: Sat Aug 27, 2016 12:15 am
- Has thanked: 9 times
- Been thanked: 63 times
Re: Remaining known DotA 6.83d bugs
Frank wrote:Its usually happening when teamfights occur so I been box selecting my hero again quickly but i think it just deselects my hero not sure 100 percent ill try to pay more attention next time it happens
https://entgaming.net/findstats.php?id=12256907
a game where it happened I even said it in chat at around 28 minutes elapsed time as earthspirit
Thanks for the detailed info! I watched that teamfight, been 27:00 and 28:00 in slow motion. Indeed, your hero gets deselected 3 times!
It's certainly not related to latency. It's right there, recorded in the replay and played all the same in everyone's Warcraft III.
I somehow need to find a way inspect the replay actions... DotaReplayManager doesn't seem to load WC3 1.30.1 replays.
-
EdteOfChaos2
- Treant
- Posts: 360
- Joined: Sun Sep 15, 2019 8:24 pm
- Has thanked: 68 times
- Been thanked: 108 times
Re: Remaining known DotA 6.83d bugs
Should be possible (and maybe even easy) to write a tool to parse replay actions based on this: http://w3g.deepnode.de/files/w3g_format.txt
I don't know if it will actually provide any valuable info though. It likely comes across as a 0x02 (Deselect) in the replay which wouldn't really explain anything. I think triggered deselects just look like deselects.
I don't know if it will actually provide any valuable info though. It likely comes across as a 0x02 (Deselect) in the replay which wouldn't really explain anything. I think triggered deselects just look like deselects.
-
dumpster_
- Forest Walker
- Posts: 111
- Joined: Sat Mar 09, 2019 9:51 pm
- Has thanked: 29 times
- Been thanked: 31 times
Re: Remaining known DotA 6.83d bugs
Bara's bash is triggered when he starts his attack instead of when the attack lands.
This on its own is pretty dumb and can be abused, but I don't know if it could be considered a bug.
But there's one interaction that I think does count as a bug.
Slark in ult is supposed to be untargettable.
When there are no other targets around except an ulted slark, bara will try autoattacking him.
Actually landing the attack isn't possible, so he keeps retrying, starting and cancelling many attacks.
The failed attacks can and do proc bash, often leading to permabash.
This on its own is pretty dumb and can be abused, but I don't know if it could be considered a bug.
But there's one interaction that I think does count as a bug.
Slark in ult is supposed to be untargettable.
When there are no other targets around except an ulted slark, bara will try autoattacking him.
Actually landing the attack isn't possible, so he keeps retrying, starting and cancelling many attacks.
The failed attacks can and do proc bash, often leading to permabash.
-
h3rmit
- Armored Tree
- Posts: 87
- Joined: Sat Aug 27, 2016 12:15 am
- Has thanked: 9 times
- Been thanked: 63 times
Re: Remaining known DotA 6.83d bugs
dumpster_ wrote:Bara's bash is triggered when he starts his attack instead of when the attack lands.
This on its own is pretty dumb and can be abused, but I don't know if it could be considered a bug.
But there's one interaction that I think does count as a bug.
Slark in ult is supposed to be untargettable.
When there are no other targets around except an ulted slark, bara will try autoattacking him.
Actually landing the attack isn't possible, so he keeps retrying, starting and cancelling many attacks.
The failed attacks can and do proc bash, often leading to permabash.
Wow, that's an interesting one.
I can understand why the approach of triggering bash at the beginning of the attack was taken by IceFrog. The initiation of an attack is an event triggered by your unit, while the actual landing of the attack is an event triggered by any other unit on the map. Attaching triggers to every unit on the map to check whether they received an attack is tricky and error-prone.
Changing how this mechanic works would somewhat nerf Bara, so I'd rather avoid changing it.
However, bashing an untargetable unit is definitely a bug and should be fixed. I'll add it to the list.
-
EdteOfChaos2
- Treant
- Posts: 360
- Joined: Sun Sep 15, 2019 8:24 pm
- Has thanked: 68 times
- Been thanked: 108 times
Re: Remaining known DotA 6.83d bugs
h3rmit wrote:I can understand why the approach of triggering bash at the beginning of the attack was taken by IceFrog. The initiation of an attack is an event triggered by your unit, while the actual landing of the attack is an event triggered by any other unit on the map. Attaching triggers to every unit on the map to check whether they received an attack is tricky and error-prone.
This was the case back when the map was originally written, better systems exist now (Damage detection systems) to do this kind of thing. Axe's spin and Javelin's (1500 gold item) passive effect are other examples. I'd recommend just leaving it alone though, it's been that way for so long that everyone's used to it.
I believe the way they made Slark "untargetable" is by forcing enemies to issue stop when attacking Slark. This is a bit of a silly way of doing it. You can see this if you run into an enemy Sentinel tower with Slark - it repeatedly tries to attacks then stops. It also doesn't work for spells and Lina can ult Slark during his ult, if you click it before and then Slark ults its easiest to see this because of Lina's long cast animation.
-
h3rmit
- Armored Tree
- Posts: 87
- Joined: Sat Aug 27, 2016 12:15 am
- Has thanked: 9 times
- Been thanked: 63 times
Re: Remaining known DotA 6.83d bugs
EdteOfChaos2 wrote:This was the case back when the map was originally written, better systems exist now (Damage detection systems) to do this kind of thing. Axe's spin and Javelin's (1500 gold item) passive effect are other examples. I'd recommend just leaving it alone though, it's been that way for so long that everyone's used to it.
Indeed, unified damage detection systems exist in later maps. But 6.83d has remained in the past, as it was originally written.
EdteOfChaos2 wrote:I believe the way they made Slark "untargetable" is by forcing enemies to issue stop when attacking Slark. This is a bit of a silly way of doing it. You can see this if you run into an enemy Sentinel tower with Slark - it repeatedly tries to attacks then stops. It also doesn't work for spells and Lina can ult Slark during his ult, if you click it before and then Slark ults its easiest to see this because of Lina's long cast animation.
You were right on mark!
That is exactly how it has been implemented, I just checked.
Code: Select all
elseif GetTriggerEventId()==EVENT_PLAYER_UNIT_ATTACKED then
if GetTriggerUnit()==P7I then
call IssueImmediateOrder(GetAttacker(),"stop")
endifEdteOfChaos2 wrote:I'd recommend just leaving it alone though, it's been that way for so long that everyone's used to it.
I agree. I may just add an exception to Bara's Greater Bash to not proc on Slark during his Shadow Dance.
P.S.: Both Greater Bash and Shadow Dance's protection seem to be triggered by the same event EVENT_PLAYER_UNIT_ATTACKED, by the unit receiving the attack. Meaning means that the order in which they trigger is not guaranteed. Perhaps sometimes Bara's attack is stopped after triggering Greater Bash first.
Funnily, there's already an exception in place, protecting Roshan from getting affected by Greater Bash.
P.S.2: It looks like I can't reproduce it on my side.
EdteOfChaos2 wrote:When there are no other targets around except an ulted slark, bara will try autoattacking him.
I made a test setup with Bara waiting in the rune spot having skilled Greater Bash with 100% chance to bash on hit, then Slark coming in several times and either casting ulti there, or already having it on. I also removed the above "stop" command from Shadow Dance. Enemy towers were attacking Slark normally in his ulti.
In no case did Bara attempt to lift a finger. When Slark hit him, he turned his back and ran away instead. When I took control of Bara and tried to attack ground, Bara did nothing either. No hits, no bashes were triggered. Ofc, the moment Shadow Dance ran out, Slark got stun-locked by continuous bashes.
Do you happen to have a replay to check the circumstances under which I could reproduce it? Or do you remember any other detail I might be missing?
-
EdteOfChaos2
- Treant
- Posts: 360
- Joined: Sun Sep 15, 2019 8:24 pm
- Has thanked: 68 times
- Been thanked: 108 times
Re: Remaining known DotA 6.83d bugs
h3rmit wrote:
I made a test setup with Bara waiting in the rune spot having skilled Greater Bash with 100% chance to bash on hit, then Slark coming in several times and either casting ulti there, or already having it on. I also removed the above "stop" command from Shadow Dance. Enemy towers were attacking Slark normally in his ulti.
In no case did Bara attempt to lift a finger. When Slark hit him, he turned his back and ran away instead. When I took control of Bara and tried to attack ground, Bara did nothing either. No hits, no bashes were triggered. Ofc, the moment Shadow Dance ran out, Slark got stun-locked by continuous bashes.
Do you happen to have a replay to check the circumstances under which I could reproduce it? Or do you remember any other detail I might be missing?
You will need some sort of Truesight since the ult is also invisible - give him a gem or do it under a tower
-
h3rmit
- Armored Tree
- Posts: 87
- Joined: Sat Aug 27, 2016 12:15 am
- Has thanked: 9 times
- Been thanked: 63 times
Re: Remaining known DotA 6.83d bugs
EdteOfChaos2 wrote:You will need some sort of Truesight since the ult is also invisible - give him a gem or do it under a tower
Alright, that did it, thanks! Confirmed and fixed! An exception was added for Greater Bash to not trigger on Slark during Shadow Dance, just like it doesn't trigger on Roshan.
The fix will be included in the next version.
-
h3rmit
- Armored Tree
- Posts: 87
- Joined: Sat Aug 27, 2016 12:15 am
- Has thanked: 9 times
- Been thanked: 63 times
Re: Remaining known DotA 6.83d bugs
Frank wrote:if you die under with the boat buff from kunkka ult and buyback instantly u go down to 1 hp very fast
https://entgaming.net/findstats.php?id=12255524 replay razors bbs and loses 99 percent of hp (55min)
I was able to reproduce it multiple times in single player, it looks like all delayed damage can come back even after death and revival
Not even death can save you from the hangover after getting drunk in CoCo's Rum
-
h3rmit
- Armored Tree
- Posts: 87
- Joined: Sat Aug 27, 2016 12:15 am
- Has thanked: 9 times
- Been thanked: 63 times
Re: Remaining known DotA 6.83d bugs
While testing Shrapnel, I noticed that half the animation still plays when you're out of charges, before the "No more charges" message is displayed. While it's only a visual glitch, I added it to the list.
The interference of Shrapnel with Last Word has already been fixed and will be released in the next version.
CoCo's rum delayed damage after reviving has been fixed as well.
I also noticed that Abaddon's skill description displays wrong damage values in its description in the Learn Skills screen. Probably a leftover from a previous version. Fixed that too.
The interference of Shrapnel with Last Word has already been fixed and will be released in the next version.
CoCo's rum delayed damage after reviving has been fixed as well.
I also noticed that Abaddon's skill description displays wrong damage values in its description in the Learn Skills screen. Probably a leftover from a previous version. Fixed that too.
-
RichardCoffee
- Treant
- Posts: 358
- Joined: Wed Aug 10, 2022 1:49 am
- Has thanked: 25 times
- Been thanked: 37 times
Re: Remaining known DotA 6.83d bugs
In one of my games, there was a Sat that was having dust quickly dispelled on him far before it should disappear. Near him was a Tauren Chieftain ally casting spells.
can TC remove things like dust at level ~2? I don't play TC, and never saw it before. It happened twice. Not sure if it is a bug or hero feature.
can TC remove things like dust at level ~2? I don't play TC, and never saw it before. It happened twice. Not sure if it is a bug or hero feature.
-
h3rmit
- Armored Tree
- Posts: 87
- Joined: Sat Aug 27, 2016 12:15 am
- Has thanked: 9 times
- Been thanked: 63 times
Re: Remaining known DotA 6.83d bugs
No, there isn't such a spell. Tauren Chieftain's skills only affect enemies. Do you happen to have a replay so that we could inspect more closely what happened?
-
RichardCoffee
- Treant
- Posts: 358
- Joined: Wed Aug 10, 2022 1:49 am
- Has thanked: 25 times
- Been thanked: 37 times
Re: Remaining known DotA 6.83d bugs
Thanks for reviewing in PMs.
Yes, it appeared to be some kind of weird bugged interaction going on. Unless some 2,000 ELO korean here knows some obscure way to remove dust at min 4 with pretty much nothing — it would seem it was indeed a bug. Thanks for adding it to your list.
Yes, it appeared to be some kind of weird bugged interaction going on. Unless some 2,000 ELO korean here knows some obscure way to remove dust at min 4 with pretty much nothing — it would seem it was indeed a bug. Thanks for adding it to your list.
Return to “Defense of the Ancients”
Who is online
Users browsing this forum: CCBot [Bot] and 0 guests
