#arma3_feedback_tracker
1 messages ยท Page 14 of 1
At first i thought of putting it there, but someone told me thats only run for who the vehicle is local to, so not fitting for this.
Now i want to place it somewhere else, and try to find a working variant of the script.
Supposedly not, but certain commands can have global effects.
Had a glance at BI config inits, they do locality checks so no, its global
I was very strictly told they are not. But multiple people told me different things on this topic so far.
I'm not an addon maker, but looks like they're wrong
add some logging to test yourself
Something like diag_log [_this select 0, local(_this select 0)];
Im gonna figure out that part later.
I could also just do it per EH, when someone gets in, or changes seats.
The problem is, i think the code itself doesnt really work.
if (isServer) then
{
_this select 0 spawn {while {true} do
{
waitUntil {sleep 1; owner gunner _this != owner driver _this;};
(group (driver _this)) setGroupOwner (owner (gunner _this));
_this setOwner (owner (gunner _this)); sleep 1;
};
};
};
Add logging to see where it gets what what states of conditions are
Im unfortunately not experienced in some of this yet, esp testing in multiplayer.
Could you tell me how to do that?
Im also gonna join general voicechat and stream, but idk what else your up to rn.
Can't help you directly. Have this code in Init as I suggested originally.
You mean in eden/zeus clicking on the vehicle?
Also lets move this into #arma3_scripting
No, vehicle config
actually vehicle init in 3den should also work
Literally cfgvehicles eventhandlers init?
no idea about zeus, I don't use it much
Except for 3den init you'll need to change _this select 0 to this
no underscore
ok ๐
How to do that bit the most efficient way?
diag_log []; everything and check RPT
Im in %localappdata%\Arma 3 , how do i know which file it is? ๐ค
Nvm prop just the time. ๐
*ok just noticed, were gonna go to scripting instead, sorry for writing in the wrong channel
init event is raised on all machines where the object exists
Whoever said it's raised only where the object is local will have fun times when someone runs their code on mutiplayer
Is this also 100% the case in mods like:
class CfgVehicles
{
class newUGV
{
class EventHandlers
{
init="script";
};
};
};
yes
If I were you I'd be careful to take advice from this person in the future. It is global. And JIP
Thank you a lot!
I dont know who that is
From server arma 3 scripts and goodies.
still no idea
server where people who dont know how to script pretend they do
Do you know any working approaches to this locality/group issue?
In MP, even vanilla stomper doesnt react to player driving inputs from gunner if they didnt spawn it themself as zeus.
Im gonna spend the next few days testing how to achive it.
Any pointers would be very helpfull and shorten that time. ๐
I found them helpfull in most cases.
not sure what issues you are having
How to recreate:
- Join MP server with another person.
- Spawn UGV Stomper as Zeus.
- Other player connects to gunner, tries to drive.
= doesnt drive
While it works for who placed it.
Bypass:
4. The other player connects to driver.
5. Presses WASD until it reacts.
6. Connects to gunner.
= now drives for him
Problem is, im working on assets that dont have the driver available, or are semi unmanned, where this bypass only works for effectivecommander, aka who was first in the vehicle, which can be quickly messed up in the heat of action
Im banging my head against this since days or weeks. ๐
Sa-Matra suggests it has something to do with "owner" of the vehicle and "groupowner" of the ugv AI.
I think "effectivecommander" plays an additional role.
I'm not certain this is Zeus-specific. I've occasionally encountered this behaviour - temporarily non-responsive drivers when controlling a UGV gunner on DS - with Editor-placed UGVs too.
which branch?
For me, stable, but occasionally. Sounds like the other guy has it more often.
there were changes so curious what is it like on dev/prof
Does connecting to the gunner make the gunner local?
uh, remoteControl switched from LA to GA in 2.14...
should
This feels like a bug given the command description.
Main.
Sry at sports rn. ttys
offtop
My assets scripts has the same script EH config, it works if place via Eden, but script will not be executed on asset if Zeus placed
This works as designed? Dunno why scripts "init" EH won't work on Zeus placed objects
which inits are "script init"s?
class CfgVehicles
{
class Land_Laptop_unfolded_F
{
class EventHandlers
{
init="_this call compile preprocessFileLineNumbers 'path\my_script.sqf'";
};
};
};
my_script.sqf not work if Zeus spawn this laptop, but works if placed in Eden
My work when placed in zeus, for example spawning and attaching something. Or removing crew.๐
But you might as well mean something else.
On the topic of cfgvehicles EH init:
He wrote this:
ehs run on the vehicle where it is local.. if they are baked into a mod. loops with server checks when a ugv is spawned with zeus, and is possibly controlled by another person , prob not best approach as the locality is going to change.
Did i misinterpret, or is this false information?
He said i misinterpreted him, which is possible.
So i removed his name. He is a very helpfull person, and i dont want anyone thinking too bad about him.
Im wondering how would you interpret it?
if true, and I need to test this, this would mean zeus placed assets would not init properly on other clients and JIP. This is pretty major, unless intended, which I am trying to find out as well
this would be config based EH
To be clarify UD1E not the one who is "someone", just statement here for not blaming my friend
I don't see any reference here to init running only locally
He wrote me in dm.
We talked about it in detail and i think we calmed the storm.
I do not have any ill will against him, and also defended your server against badmouthing. ๐ค
Could you elaborate, im not a native english speaker, and want to prevent further misinterpretations. 
Do you mean with here:
A) Data about how the game works/ is supposed to work.
B) Leons screenshot.
(The context of the conv was me wanting to use cfgvehicles EH init.)
C) Or the copied quote earlier, which i might unfortunately have misinterpreted?
(The context of the conv was me wanting to use cfgvehicles EH init.)
Started using custom channels on live server, getting
2024/03/18, 22:15:04 Unexpected IsPlayerOnChannel 6
2024/03/18, 22:15:04 Unexpected IsPlayerOnChannel 9
2024/03/18, 22:15:04 Unexpected IsPlayerOnChannel 7
2024/03/18, 22:15:04 Unexpected IsPlayerOnChannel 9
```RPT spam regularly
what game version?
Original output filename: Arma3Retail_Server_x64
Exe timestamp: 2023/09/29 15:32:52
Current time: 2024/03/18 04:22:16
Type: Public
Build: Stable
Version: 2.14.151058
so this is before I did tweaks
Not exactly funny or true.
I always wondered why some people have the urge to throw lies even if they have 0 connection to the thing they are talking about.
<-
I was referring to the screenshot. A few pointers. If you want to know something for sure, test it. Sa-Matra explained you how. This is not a channel for drama. We would all appreciate more reports about the actual problems with the game so that it could be dealt with and less wasting time on pointless he said this I said this. Thank you for understanding.
Ok, i only refered to the copied quote i mentioned 22:53 CET.
Not at all to the screenshot.
ehs run on the vehicle where it is local.. if they are baked into a mod. loops with server checks when a ugv is spawned with zeus, and is possibly controlled by another person , prob not best approach as the locality is going to change.
As the whole topic of my conversation with them was using cfgvehicles EH init i likely misunderstood them refering that in this quote,
no drama was intended, just finding out how cfg vehicles EH init actually works, as i thought i got conflicting informations.
For that we ofc should have changed earlier to the scripting channel, as finding a workaround to the initial ugv bug is seperate from reporting it.
When i find a solution to the UGV bug, i will ofc share it. ๐๐
10:14:37 Unexpected IsPlayerOnChannel 6
reproed
Really juicy bug actually
there have to be tickets it is so obvious
Couldn't find any, only in unrelated RPTs
basically custom channel markers were not sent to JIP
oh, alright
How does it even work? You can't add JIP into radio channel until they join
No markers on custom channels for JIP ever?
not to first time joined not to lobby drop ones
Also been using custom channels for few days now, practice shows that even if I do radioChannelAdd only on server, players randomly get kicked off custom channels. Happens once every few hours, couldn't catch it myself yet but seen videos of it happening randomly.
now you can if you named unit and have ai controlling it when you are not home
Not mass kick apparently, happens mid player's life
you will have to retest it after I did changes to add remove routines
Are these changes on perf yet?
Just had a thought, maybe perf joining stable messed it up?
no they won't be it is networking changes need all clients and server the same
Didn't get it.
you have option to enable disable ai in role selection right?
Yes. We have AI disabled though.
My mission setup is following: Everyone joins as civilian Logic unit, then createGroup createUnit their own unit after they pick side in-mission, then you ask server to put your unit into custom channels.
if unit doesn't exist you obviously cannot add it to channel no?
Correct. I was wondering how custom channel markers could ever be sent to JIP players
Only if they join AI slot then?
Ooh, maybe I don't understand how it works. JIP get ALL markers in game and only show ones that are in their current channels?
no jip gets markers created that match condition
player being on the list is condition for custom markers
player being the leader is for command chat
etc etc
But being a leader and in custom channel is something that changes mid game, how does it work then?
that is how JIP works
Anyway I haven't tested JIP markers on custom channels yet, maybe it all just works and I have no idea how JIP functions in regards to markers
mod game != jip so there are other mechanisms
I just told you they are broken
Do you want me to write up a ticket about it?
I still haven't made a ticket about my local custom channel idea, I wanted to ask you if its technically feasible at all.
If I understand it right, clients list on custom channels is basically a list of recipients of your chat and voice messages?
not now please I have a lot of things to finish
Alright
Basic write up for now: https://feedback.bistudio.com/T180034
purged, thanks
But why they don't promote Pattaya
I kinda thinking about control (GUI elements) filters
- colorizeR, G, B
- hue switch
- saturation
- brightness/darken
- blur
- mirror V/H
And blend mode - Normal
- Addition
- Substract
- Multiply
- Overlay
or something, maybe transfer Alpha/transparency from other source is cool to have.
I know, I know. Say is easy, do is not. a Japanese proverb told me enough, just my wish
Have a need for this as well
At least extracting alpha from a picture
This way you could create custom shadows through ui2texture to fix poor visibility for many in-game icons
Darken texture and blur it, call it a dropshadow
Maybe this is unimportant, then just ignore it:
It would be nice if UAV_AI (which has different simulation) still reacts to cfg "forcehide" like "forcehidedriver", as it currently ignores that in certain cases, like when not all crew seats have UAV_AI in them.
More information:
https://www.reddit.com/r/armadev/comments/7vz67h/uav_ai_dilemma_config_editing_question/
when you say players get kicked off the channel, how does this manifest itself? Players losing access to send? receive? what happens to the units list for the channel, do they disappear from the list?
I don't know all details yet, couldn't capture it myself yet.
Apparently you get removed from the channel and it auto switches to the next one. You no longer hear anyone and you can't send as channel cannot be selected. As for players list, no data, but apparently not everyone gets kicked.
Here is a video of it happening. "Side chat" here is actually a custom channel, default side chat (channel=1) is disabled.
- Player was hearing people speak in custom channel
- Voices stopped randomly, it auto switched to global channel
- Cycles through all channels, custom channel (side channel) is not there
- Goes to settings which lets you join/leave russian side channel, flips there switch which asks the server to
radioChannelAddthem back into custom channel and it works, they're in custom channel again
Info so far:
- Happens randomly, not on death or respawn
- Channel auto deselects
- Apparently only some players get removed from the channel, not everyone at once
All this happens in a setup where only server ever executes radioChannelAdd and radioChannelRemove, clients never do
yes it was set to switch you to the next channel on the list when you are removed
so this could be due to some client adding himself but instead of adding to units it overwrites, this should be fixed in future updates and can be tested on dev
This case happened where only server ever use add/remove commands
but the client was still able to overwrite
in new version client sends request to the server and server add to the list then sends new list to everyone including itself
The one match for radioChannelAdd and radioChannelRemove, done on server side
Seems like there is some other bug that throws players out even if clients list is handled only by server ๐ค
but you can still execute on client if you have access
Sure, but I don't think there is anything in the game that uses custom channels
Unless you mean there is a cheater randomly kicking players out, I don't see how this could ever happen by itself
What's interesting is that this playing was living (and in channel) for quite a while before they got kicked off, they had to buy the tank and drive to action, they definitely shouldn't be first or last entry in that list
Something related to groups breaking maybe?
When server adds new client to the list, maybe it validates existing ones?
in order to get added removed there has to be a message I will have a look at simulation in case it has some sort of cleaning that could cause it, but really this should be tested on the new build
Since this happens fairly rarely (once every few hours), we won't be able to properly test until everyone updates to new version to have enough players to test.
This is the only place where client even asks server to add or remove itself from custom channels
You always only ask to add yourself into _side_channel, never ask to get removed from it.
Yet player got randomly kicked from it mid gameplay somehow ๐ค
Option one: You get added to custom channel 1 and 2
Option two: You get added to custom channel 1 and removed from 2
Gameplay: You get randomly kicked from channel 1
Feels like some deeper MP bug somewhere ๐ค
Are these things happening without mods?
phew Finally made it sync JIP custom radio markers properly with joining player
so if you have AI enabled and added AI unit which is placeholder for player to the channel, players joining into that slot will see markers on this channel
So wont work for me then
Is it a known and accepted thing that:
- If a player unit has a custom texture set by the mission maker for example:
this spawn {sleep 2; _this setObjectTexture [0, "g_orange_kakhi.paa"];};```
And the player uses the Arsenal UI with for example:
["Open", true] spawn BIS_fnc_arsenal
It resets the player's uniform textures to the vanilla textures of the uniform?
Its arsenal re-adding the uniform
does he just open the arsenal or also selects a new uniform?
Just opens it
Maybe Arsenal shouldn't re-add the uniform on load automatically unless the player clicks on another uniform? And do a getObjectTexture on load and try to re apply the custom texture if the uniform did not change when closed from what it was when the UI was opened?
You can script that
ArsenalOpenEH?
Thanks, will use that ๐
what won't work?
we have ai disabled, lobby slots are civilian logic units
yeah so what won't work?
getting markers for custom channels for JIP players
If I understood you right
if you give me rough repro I can test and make sure it works
is that a yes?
I'll make you a repro to test it for my case
is there any reason preproc support cannot be added for rvmats? you can do it through cmd with cfgconvert so wondering if theres a technical reason or just nobody's bothered yet
They should do. What's your usage?
speeding up importing things by bulk
wait i may be thinking of relative classes that dont work in rvmats
https://feedback.bistudio.com/p/jessica2024/
Spam account
Fault address: 2E5CD3E4 01:0009C3E4ll C:\Windows\SYSTEM32\ntdll.dll
Prev. code bytes: 20 00 00 00 F6 04 25 08 03 FE 7F 01 75 03 0F 05
Fault code bytes: C3 CD 2E C3 0F 1F 84 00 00 00 00 00 4C 8B D1 B8
worth to drop frozen.mdmp on FT with this?
could be related to our EMM extension fetching servers given the rpt context
crazy suggestion, but what about having (removed) TKOH options in A3?
https://community.bistudio.com/wiki?title=Arma_3:_Difficulty_Settings&diff=prev&oldid=96711
this would imply solo / server settings changes
Do you mean, uh, re-introduce aerodynamic things?
rough landing, wind effects, auto-trim, stress damage yeah ๐
can we bring back Super AI as well?
this was just max AI skills, was it not? aka can be easily done via sqf too
Bug in Arma 3 editor?
- Go into 2D view (by pressing
m) - Move camera by clicking the middle mouse button (dragging the camera icon doesn't cause the issue in question)
- Try rotating the camera with shift click
- Camera moves back to where it originally was before you middle-click teleported it
- Going back into 3D view and then back into 2D view fixes this though for each time you need to do it
Who rotates camera with shift click?
wasnt AI spot distance nerfed in tha tac ops dlc patch?
Can't test your fix as its not deployed yet
In repro you start as civilian, then selectPlayer createUnit a BLUFOR unit and then join a custom channel
Yeah doesnt sync markers
OWNER syncMarkers CHANNEL SE command?
So they can be script-synced when needed
@solid marten
cc = radioChannelCreate [[0.96, 0.34, 0.13, 1], "Custom channel", "%UNIT_GRP_NAME (%UNIT_NAME)", [], false];
cc radioChannelAdd [player];
diag_log(radioChannelInfo cc select 3);
deleteVehicle player;
diag_log(radioChannelInfo cc select 3);
=>
21:47:53 [B Alpha 1-2:1 (Sa-Matra)]
21:47:53 []

So deleteVehicle player actually deletes something, fucks up something over network so that custom channel no longer sees the player?
Wonder what other bugs this causes if it does
Doing deleteVehicle on remote player also removes that player from radio channel
It re-adds back properly though
no why script synced
what do you mean
deleteVehicle player ends up removing player from radio channel
ah yeh this is probably the cause of strange removal
it should remove units that are on the channel and deleted it has been fixed as well
but yeah looks like the cause of that bug
i will make it proper sync markers on add remove from channel just need to find where is the best place to put
How does deleteVehicle does anything if it can't delete active player unit though?
Is there a specific action to remove unit from radio channel within command?
it was removing real player from channel and then aborting because of bug
look we need to sync the testing as well, because there were fixes
Thought maybe deleteVehicle player messed something up over the network, causing unit to be deleted from radio channel
no i just told you it removed player unit from the channel then aborted because you can't delete player
now it aborts straight away
But yeah, I guess something somewhere tries to delete player and thats how they get removed from radio channel, mystery solved
the "Arma Files" ๐ฝ
Please fix this small config issue
https://feedback.bistudio.com/T178139
Matter of a fact, with such setup everyone sees civilian side markers for normal side channel
I guess there needs to be a way to swap markers when your side changes, probably through script of mission config setting, for backwards compatibility
you mean when you select player into blufor they still see civilian markers?
for backwards compatibility
you are not supposed to have access to markers you shouldnt have access to, there is no backward incompatibility, it is a bug
Yeah, new markers are also civilian
basically it uses lobby slot side for everything, regardless of your actual side
it is not, it just doesnt delete markers you have on your client when you change side and doesnt create markers that are on your new side
the marker control is done via create delete this is why you have local version of marker commands and thats why global marker command sync everything
Using same marker repro
Player 1: Two markers placed, one as civ, another as BLUFOR.
Player 2: Sees only one marker. Joining BLUFOR keeps civ marker, no BLUFOR marker
Player 2: Rejoins, sees both markers, one placed as Civ by Player 1, another as BLUFOR as Player 1
Looks like it broadcasts marker as BLUFOR (actual side) but saves it for JIP as Civilian (lobby side)?
its jip sync it works but was broken for custom. JIP is fine it is dynamic syncing that is...absent
broken for custom ...well technically it didnt exist either
Fine as long as you never change your side from your lobby side
Not sure what's the best way to solve it, have client request all markers each time they change side, or all custom channel markers each time they join/leave custom channel?
both
Considering the fact that you get removed from custom channel each time you die, won't it be too excessive to have server send you ton of markers each respawn?
tonnes? how many?
Yeah I'm exaggerating I guess, doubt it will be more than 100 markers or something
Unless somebody does a polyline painting somewhere off map
still not that much data
nothing
Sometimes I wonder, why/what is the condition to detect a synchronized object in-game (not in-Eden)? Even if in Eden it can successfully connect an object to an object, synchronizedObjects does not return anything if the object is e.g. a prop?
I know Eden connection is somewhat different with regular and old synchronization way, it actually confuses me why it doesn't work or why it is designed like this
I suppose for performance reason it only checks for "alive" units + waypoints, hence Modules having to be Logics
But it still is very confusing because you can sync in Eden but not in game
Very unreasonable IMO
removed spam, thanks

Vehicle commander can have issues with his weapon and Smoke Screens in some cases on dedicated server with full-player crew
https://feedback.bistudio.com/T180141
group eventhandlers dont fire when the player order 'combat' 'safe' 'stealth' etc
is this considered a bug/worth to make a ticket for?
only seems to trigger from auto combat
and waypoints I believe, but command menu doesnt
Command menu only sets unit behaviour, not group behaviour IIRC.
new record!
don't think I'll ever beat that one
Is that how long the account was active before being deleted or how long since it was created ?
created โ banned
That's short indeed
Very short
Some vanilla factions wipe out parent inheritances (MAN, CAMANBASE) for stuff like EventHandlers (Looters - Contact, Syndikat - APEX, FIA - Vanilla) down the line.
Is this documented somewhere?
Was testing yesterday but seems to only happen with Greenfor units, not sure about other sides/factions, as Vanilla Blue and Red units were not affected.
I can see why CBA has its own function for that after all.
Is this documented somewhere?
not really, it's just bad code on BI side. Some story units have broken inheritance too.
So, I just discovered an issue within the vehicle respawn module, that any vehicle of any type which has been synced to the module, is teleported to debug corner upon destruction.
How to reproduce:
Be on the latest arma 3 version 2.16
- go to eden and place down a helicopter, tank, wheeled, car or boat.
- place the module: Vehicle Respawn (ModuleRespawnVehicle_F) in the mission.
- Sync the vehicle to the module
- set the following parameters (I haven't tried if other parameters change the behavior other than the wreck parameter)
- delay is empty
- Deserted Distance is set to 1500
- Tickets is left empty
- Expression is left empty
- Position: starting position
- Position selection: Oldest
- Wreck: Can be both preserve or Delete or Delete with an explosion effect.
- Notification: Enabled
- Forced Respawn: Disabled
- Respawn when disabled: Disabled.
-
Place a player slot in the mission.
-
set up a way to destroy the vehicle either from a trigger on radio command. Or just spawn an A.I. from zeus with a rocket launcher.
-
Set the player slot up in such a way that it also has zeus access through a game master module.
-
Start playing the mission.
-
Destroy the vehicle and watch how its teleported instantly to the debug corner.
-
bonus: if lucky, the helicopter now gets stuck in a perpetual loop of spawn, take damage, be destroyed, respawn within 10 seconds. and it begins from the start.
I attached an rpt file in a zip file.
I have no crash dump.
Link to a video about the bug on my yt. it can be found through the following url:
https://youtu.be/4-DA_1urPAA
Another issue? I am not sure if this is part of the same issue or if this is something entirely different. Though the reproduction method is the same.
https://youtu.be/TTSfsQQMNds
During the testing of this, I had no mods loaded.
This is recorded on version 2.16 of Arma 3, on 27-03-2024 at 09:40am local time (Amsterdam). This video is an explanation of how I set up the mission file with the Vehicle Respawn Module and it's behavior. After the video was taken, I also tried with a Marshall APC, A HEMTT car, The NATO Tank (Slammer), a Wipeout and the Scorcher Self propelled ...
This is recorded on version 2.16 of Arma 3, on 27-03-2024 at 09:55am local time (Amsterdam). This is mere minutes after the previous video I took to report an issue with the Vehicle Respawn module. And I started noticing behavior of the Ghost Hawk getting stuck in a perpetual destruction loop. After it being destroyed once by me, upon respawn it...
Posting this to bring attention to it https://feedback.bistudio.com/T180184
It's similar to a previous ticket I made:
https://feedback.bistudio.com/T175562
but I hope it's easier to implement as I think it would be a massive boon to people making 3d scopes in arma
update:
Narrowed the 2 base classes that wipe EH inheritances from camanbase:
I_soldier_base_F
I_G_soldier_base_F
If you properly redefine the EH on those 2 classes and on camanbase all the units in the vanilla game properly inherit again.
This affects quite a few units actually, basically all the ones that use randomized headwear in GreenFOR side
does this mean a thing is fixed?
In theory :P
when inside of a building submerged in water during rain, the rain sound shader will take priority over the house interior filter. Water inside of a structure makes the sound of the outside rain sound as if it was ignoring the structure altogether and going through it landing on the water inside, e.g. same as if standing outside. When distancing from water, the interior filter starts to apply. The rain also sounds louder when inside the structure in this case too.
So in BIS_fnc_fireSupportVirtual, there is a call to say3D with default parameters. The trouble is, that the sound played has a radius of 100m. The shells by default are created at 250m and have a speed of 150m/s. As a result, you can hardly hear the sound play properly. The sound is 3s long, the shell flight time is 1.6s.
This leads to a suboptimal effect when using the function with default parameters, or while using Zeus fire support which i think also uses the function (because the sound cuts of there too)
I had good results with making the shell spawn at 75m with a speed of 25m/s, but might be good to change the defaults?
Some say3D execution in vanilla scripts have issue when params are not written when executing a command (e.g. if audibly radius is more or less than 100m)
Just needs to add are necessary parameters (taken from the configuration of the SfgSounds sound class) when exec say3D
Same for jets ejection sound https://feedback.bistudio.com/T174022
Well I don't change the say3D in the function, just the shell spawn height and speed, and the sound plays correctly. So in this case i don't think it's the say3D itself is a problem, just how it interacts with the shell's timing. The mortar sounds used by default are 3s long, so the shell needs to be falling for that long, and players should be close enough to hear it
The radius is also the problem, because the sound is played from the shell itself as soon as it spawns (250 metres in the air). With the radius being 100m, for the first 150m of the shell's descent, it's too far away for the sound to be heard.
well yeah true that way it is also a problem. I worked around it by spawning the shell at 75m, since BIS_fnc_fireSupportVirtual doesn't expose the audibility radius, just the sound files/class names themselves. Maybe it should expose the var, along with setting the radius of the say3D call in the function to 250m and reducing the fall speed to 83.3m/s.
(But either way you want the sound heard for 3s, which can be achieved with or without modifying the sound radius)
Fall damage on units gets attributed to lastEntityCausingDamage by the engine 
Is this a recent change or it was always like this?
Collision from objects too
So basically if unit was shot by another unit and that unit then falls or gets hit by a car or some physics object, all damage will be attributed to shot instigator for some time, including subsequent death from that collision
But not to context>2 (fake) event handlers
I feel like 80% of me developing is building crutches around undocumented and unasked for engine bullshit
Fire (particle) damage doesn't
I encountered a bug last night where players got desynced in player-hosted Multiplayer. After leaving a vehicle, the player leaving it would seemingly stay inside the vehicle on the other players screens. If the player got into another vehicle and left it, another body of that player would stay in that vehicle on the other players screens.
I haven't looked further into it yet, is that a known thing that can happen or should i try to reproduce it?
It happened in 2 separate sessions with different hosts btw
test it in previous version
what version? Unless it is reprored on latest dev it is not good to us as it could have been fixed already
Is that bug in this file? functions_f\modules\fn_moduleinit.sqf for line 29 that having extra backslash at the end
_isPersistent = getnumber (configfile >> "cfgvehicles" >> _logicType >> "isPersistent") > 0 || getnumber (configfile >> "cfgvehicles" >> _logicType >> "isGlobal") > 1;\
_is3DEN = getnumber (configfile >> "cfgvehicles" >> _logicType >> "is3DEN") > 0;
is it in a #DEFINE?
just switch, no define/preprocessor marco
Looks like a leftover.
Stable 2.16
Not mine but looks useful:
High command needs a "GET IN " and "TRANSPORT UNLOAD" OPTION https://feedback.bistudio.com/T180240
Running also into the same issue today on a dedicated server of mine - latest Perf. version (2.16.151648)
Can't test it on dev sadly right now
I'll check looks wrong
Yeah this is fine preprocessor removes it
There were no changes in 2.16 in this regards but there were some changes in 2.18 so it has to be reproduced on dev branch
bumping dis because i just remembered about it would be veri nice
Just found a third duplication of sound shaders in weapon configuration. Leading to double triggering of sounds for Katiba, SPAR, and Mk200
https://feedback.bistudio.com/T178139#2589058
Bug #1 functions_f\modules\fn_moduleinit.sqf
_isPersistent = getnumber (configfile >> "cfgvehicles" >> _logicType >> "isPersistent") > 0 || getnumber (configfile >> "cfgvehicles" >> _logicType >> "isGlobal") > 1;\
Bug #2 functions_f\systems\fn_supplydrop.sqf
case 0: {_paraName = format["Parachute%1",_side],;}; //ParachuteMediumEast or ParachuteMediumWest not "FLY" (that spawns 150 m above ground) };
Bug #3 functions_f\variables\fn_variablespaceadd.sqf
if (isnull _object) exitwith {debuglog "Log: [variableSpaceAdd] ERROR: Null object! %1", [_variablename, _element, _global, _check]};
Bug #4 functions_f_contact\drone\fn_returnmattermodule.sqf
private
The first one is irrelevant, but I guess you want it fixed for some sqf parser or so?
Nah, just letting anyone here know about the bug exist, not even sure if those function was used in any other place
doesnt seem those are breaking the game
good report! did you make a ticket?
Affirm!
the bad EH definitions are annoying, but fixing those now may lead to mods breaking/reintroducing the problem. personally i would still go with fixing it. is there a ticket about those? pretty sure there is/was one for the civilian one
should be possible to compute the _maxDistance for say3d from the _altitude parameter of the function - just needs someone to update the fuction. time for a community upgrade project again ๐
imho BI should just put all configs and scripts on github and let ppl submit patches (plus hire someone to review the submissions and merge - so Dedmen/KK/reyhard can focus on engine/more complex work)
[Feature Request] Add eventhandler for issuing squad commands
https://feedback.bistudio.com/T180298
to account for unit state changes not handled by the group EHs (plus such EH would have a lot of additional utility)
i dont remember seeing anything specific and didnt create one myself
So... I recalled that AI weapon/turret sway is a thing for EVERY situations, even if it is meant to be an unmanned/computer-driven turret, like CIWS or such. Can we have a favor to... like, at least setSkill way to stop sway?
For players I know there is a mod that does it (remove sway) but individually or depending specific situation idk
Yeah removing from player is very possible and easy
But nobody wants to see a CIWS that is handled by a human operator who has trepidation
Do you mean them overshooting the target when they first aim at it?
Yeah
Happens for tanks as well..very annoying
Would be neat to actually aim any turret via scripting.
Actually, just realized Scorcher (NATO SPG) main turret has no such issue
...As far as I can tell
Maybe because it's artillery. No clue.
But Mk45 Hammer does have, which is also an arty
Hmm ok
maxHorizontalRotSpeed maxVerticalRotSpeed could be the issue actually. As far as I can tell turrets with < 1 has no such overshoot
I think the AI handles turret traverse a bit more like an actual turret rather than like mouse aim, and they hold it at full traverse speed until it's on target, rather than slowing down as they approach the target. So it's not sway/accuracy error but rather a flaw in how the AI aims.
Would be nice to fix but maybe a big change to how it works.
Yup seems like socpp class CfgVehicles { class Tank; class Tank_F: Tank { class Turrets; }; class MBT_01_base_F: Tank_F { class Turrets: Turrets { class MainTurret; }; }; class MBT_01_arty_base_F: MBT_01_base_F { class Turrets: Turrets { class MainTurret: MainTurret { maxHorizontalRotSpeed = 0.1; maxVerticalRotSpeed = 0.1; }; }; }; }; This won't repro the issue, make maxHorizontalRotSpeed and other bigger, it does the overshoot
It actually look like a PID overshoot issue
https://youtube.com/watch?v=HbhOUUAPvM4
I just want this smooth'n'fast turret ๐ฉ
United States Armed Forces. Counter rocket, artillery, and mortar, abbreviated C-RAM or counter-RAM, is a set of systems used to detect and/or destroy incoming rockets, artillery, and mortar rounds in the air.
00:00 C-Ram live-fire
00:57 C-Ram live-fire
01:45 C-Ram in Afghanistan
Become a channel sponsor and you will get access to exclusive bo...
Hello guys! Is there any news regarding https://feedback.bistudio.com/T180288 ticket?
thank you in advance!
Seems like something has changed in 2.16, because we have faced some new issues with Battle Eye now ๐ฆ
and without BatlEye?
without BattleEye we cannot reproduce this issue
Interesting, dont think we had BE changes in between
in 2.14 it has been worked well. So we had to downgrade to 2.14 ๐ฆ
Yeah we all do๐ข
With tanks it's even worse. Ai shoots before actually pointing at the target. And with like 20 seconds reload time that's pretty bad. Make commanding a tank a pain.
i just want setProxyModel so i can do modular vest ๐ฆ
I thought proxy in proxy is actually not possible except in a very few special cases?
you can proxy onto vests so shrimply change model of proxy to Things on it
in attributes for waypoints in init, Identified looks like it should be Identifier?
What's more important is, Identified doesn't even work
Is there an approximate ETA for 2.18?
soon โข๏ธ
OTOH feels like something that gives players a chance to shoot back?
Is it worth it asking to remove that built-in shadow from Arma 2 item icons? Its really ugly when used alongside non-shadowed A3 icons. I can do the removal myself too. Pictured: huge shadow on grenade icon.
Pretty sure everyone besides a few mods has a shadow on icons/items. That would also change the look and feel for Vanilla Arma 3 so I don't think it will get done.
Only A2 icons have these large built-in shadows, A3 icons don't
A3 gun, A3 sight, A2 grenade icon
I see.
https://feedback.bistudio.com/T180338
The "SlotItemChanged" EH is partially broken when setUnitLoadout is used
I'll have a look
I should have specified: I'm using the latest profiling branch
it was added back with 2.14 so doesn't really matter, hasnt been touched since
Is it possible to create a server.cfg entry for BattlEye Query Timeout? Currently it seems to be hardcoded to around 2 minutes. And for reasons, weaker computers take some more time to connect to Servers and with heavy modset they are getting Query Timeouts more often.
f = {diag_log text (diag_frameno toFixed 0)};
call f;
0 remoteExecCall ["f", 2];
```=>
12:56:44 361233
12:56:44 361236
I'm trying to track engine events through RE and because of this delay RE arrives frame later than engine event actually happens.
(Trying to make sense of EPE damage, sadly remote crew engine collision damage arrives 2 frames earlier than my RE notification)
If only RE wasn't skipping frames I'd have it working correctly
to increase?
What is your value of 'disconnectTimeout' in server config?
Anyway I suggest to log the disconnects with available server tools first and see why this is happening
Arma 3 logo black.png
1.56
maxdesync = 150; unknown Max desync value until server kick the user
Arma 3 logo black.png
1.56
maxping= 200; unknown Max ping value until server kick the user
Arma 3 logo black.png
1.56
maxpacketloss= 50; unknown Max packetloss value until server kick the user
Arma 3 logo black.png
1.56
kickClientsOnSlowNetwork[] = { 0, 0, 0, 0 }; unknown Defines if {<MaxPing>, <MaxPacketLoss>, <MaxDesync>, <DisconnectTimeout>} will be logged (0) or kicked (1)
Arma 3 logo black.png
1.56```
https://community.bistudio.com/wiki/Arma_3:_Server_Config_File
arent these unrelated to BE timeout?
Dwarden might know more. may not be possible to have BE read from (arma game) files
Yeah, if this is handled by BE we have very limited say in it
there might be an existing interface, but would at least also have BE adjust their API and allow adjustment of the timeout
Dwarden would know more about it
Couldn't you adjust Arma to not have so much network-silent time?
Although I guess battleye has its own keepalive in another thread? Not sure why it'd go missing for two minutes.
is there a reason why Huron can't slingload any HEMTT other than the transport ones? I remember that back in the day it was capable of transporting even tanks so it's carryweight should be around 70t(?) but somehow I cannot slingload a flatbed truck or even the one with no trailer
Yes to increase the value. At least if its possible. From your answers it seems to be either yes or no. Hope Dwarden knows more.
Ok let me understand this, the problem is with player setUnitLoadout getUnitLoadout player as it detects removal but not addition right? Not sure what can be done about the different timing between CBA and this EH, engine does what egine does
draging manually from and to inventory correctly triggers everything for me and so is using linkItem/unlinkItem command. I can confirm that setunitloadout might not work as expected with this EH, is this correct?
hello guys! regarding our task https://feedback.bistudio.com/T180288
we have attached wrong report, but since the error is not leading to crash (and we're not able to create a following report) here's steppes we've done:
We're starting the server without mods. Mission on the Virtual Reality map without scripts and objects, signature checking disabled, battleye is enabled.
Then we launch ARMA v2.16 with CBA mod enabled (battleye enabled)
Then joining the server and kicking player with server-side command []spawn{uisleep 2; "pswrd" serverCommand "#kick 4"}
We go to the main menu and try to enter the editor.
If you select a VR map, ARMA freezes immediately. When selecting other maps, a loading screen appears, the slider reaches the end and also freezes
Thank you in advance
PS. When we do the same on 2.14 - it's all ok.
PPS. If we turn off BattleEye (on 2.16) - everything works fine
I saw similar issues while doing stuff on my Exile Test Server. I join the Server 1st time without problems, then I kill the task via Powershell. As soon as the Server started up and unlocked I joined it and whoop, game froze while loading and I had to kill it via Task Manager. When I joined then everything is fine.
Has it been started recently, or you have this error earlier?
Saw it first after 2.16 actually, also some slower loading screens after that update.
Pretty much.
If the order of linking and unlinking corresponds to what actually happens in game, I donโt really care if itโs earlier or later than CBAโs stuff.
Iโm not sure how well backpacks are handled in general, as it detected the removal but not the addition of a backpack when using the Vanilla arsenal. I didnโt change the backpack though, so I assume itโs the arsenal script that sets the loadout when leaving the arsenal.
Thanks for looking into it!
the EH fires at the end of both linking and unlinking method so it is consistent in this regard
This replication method is a pain in the arse but it does work. CBA not on server but on client does appear to be a requirement.
Shutting down the server rather than kicking the clients works.
I'd love to also have custom icons for vehicles. I'm turning Gorgon into unarmed APC by hiding its turret (why its not a separate vehicle in vanilla config already?), would love to be able to set it a custom icon along with custom name.
Doing that through scripts where possible but having it also work on in game UI would be great
Having turretless versions in vanilla would be nice yeah
Yeah, all 3 vanilla APCs support it too
no model changes needed, only config
new icons perhaps, but that's easy too
One issue though, Marshall has flag on its turret which also gets hidden
Easy new content
Speaking of flag, I wish there was standalone flag object that you can forceFlagTexture
Also isn't flag is still hardcoded to wave at like 30 FPS? I wonder how easy it would be to increase it on higher LOD
flagpole
Its a huge metal rod
its a standalone object
I was talking about invisible object that you can make display a flag
no pole or anything, just a proxy for engine-driven flag
So you can position it/attach to other objects to have the flag exactly where you need it
can you setObjectTexture the flagpole for it to be transparent ๐
Nope, tried that
Would still retain its geometry though
and shadow iirc
Yeah
Sadly forceFlagTexture doesn't work with simple objects so I can't make such flag-only object myself
It would be very useful to have these as true variants, but I wouldn't be surprised if it doesn't happen given how reluctant BI is to make the hidden variants of stuff available. Rotators made their own unarmed Marid for WS, maybe they'll do the others in a WS or RF update.
Not sure if they're reluctant, probably not caring enough.
https://feedback.bistudio.com/p/immediatealora/
Spam account
nuked
Yeah, pure money grabbers already to their next game
Gonna request a refund
Steam guy laughing seeing a refund request land after 10k+ hours gametime
Do you think Steam will let me?
it's barely fifteen thousand times too much but hey, worth a try
I'll say it didn't work past the title screen
"it's launcher time, I didn't play!!"
crashed continuously for 48 months
gotta respect that kind of dedication tho ๐
for 4 years straight : IT. WILL. WORK.
Titan AA and AT icons have slight non-transparent backgrounds under them

Not a thing with other vanilla weapons
death by firing squad for those responsible
Cancel the game
Wondering if this is worth reporting. I can fix the icons myself, but will it get updated in the game?
Speaking of icons, I really wish they would be all re-rendered, they're so dark
What I was trying to do: When a vehicle crashes into something and fires EPEContactStart, I notify all remote crew units that their vehicle crashed so I can associate their crash damage to crash event. Sadly actual engine damage arrives 2 frames earlier than scripted RE sent right from EPEContactStart event handler. Here is that 2 frames wasted distance seen by remote crew.
Frame 56517: Killed Both: B_Soldier_F
["B Alpha 1-3:1 (Sa-Matra (2)) (B_Soldier_F)","<NULL-object> ()","<NULL-object> ()",true]
Frame 56519: remoteExec EPE-EVENT: B_Soldier_F!
createHashMapFromArray [
["collider", 21c6ed66b00# 1669742: u_addon_02_v1_f.p3d (Land_u_Addon_02_V1_F)]
,["until", 410.234]
,["vehicle", B Alpha 1-2:1 (Sa-Matra) REMOTE (C_Hatchback_01_sport_F)]
,["hash_array", [
"#epe-crash"
,B Alpha 1-2:1 (Sa-Matra) REMOTE (C_Hatchback_01_sport_F)
,B Alpha 1-2:1 (Sa-Matra) REMOTE (B_Soldier_F)
]]
]
```Looks like engine damage travels instantly while remoteExec wastes 2 frames somewhere.
(Killed Both is Killed EH that unit died from a crash, unattributed to anything)
Not sure if this is fixable, but if it is, I'd be useful to me
what about mp event handlers do you use those?
I think MPKilled is executed 1 frame earlier?
Still its remoteExec that is the issue, it has delay of several frames
From what I gathered it is:
Frame 1: MPKilled
Frame 1: EntityKilled (Server)
Frame 2: Killed (Owner)
Frame 4: RemoteExec sent at the same time as death happened
Honestly a proper solution would be an event handler for collision damage or finally an attribution of collision damage properly instead of to null entities
is this on dev?
how much would i have to beg to get an option to disable modelsides stuff
You mean the thing that blocks picking up enemy uniforms?
yes
Might be a good thing for setMissionOptions but I don't know if there are ideological objections.
21:18:11 [B Alpha 1-1:1 (KK_DEV),"ItemCompass",609,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"ItemWatch",610,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"ItemRadio",611,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"ItemGPS",612,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"NVGoggles",616,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"G_Spectacles",603,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"H_HelmetB_grass",605,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"Rangefinder",617,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"U_B_CombatUniform_mcam",801,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"V_PlateCarrier1_rgr",701,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"B_AssaultPack_mcamo_Ammo",901,false]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"ItemMap",608,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"ItemCompass",609,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"ItemWatch",610,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"ItemRadio",611,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"ItemGPS",612,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"NVGoggles",616,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"G_Spectacles",603,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"H_HelmetB_grass",605,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"U_B_CombatUniform_mcam",801,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"V_PlateCarrier1_rgr",701,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"B_AssaultPack_mcamo_Ammo",901,true]
21:18:11 [B Alpha 1-1:1 (KK_DEV),"Rangefinder",617,true]```
will try to get it into prof as well
Pretty sure its been the same for years
what I meant was did you test it on dev?
No I didn't, all tests done on latest stable
What do you think can be different there? remoteExec behaviour changed?
dev has changes which may or may not be related but I guess the only way to find out is to wait for the next stable
I know you have no responsibility anymore but gotta report this anyways: @sinful kettlesqf a switchmove "HubStanding_idle1" and call something like sqf a switchMove ["AmovPercMstpSrasWrflDnon",0,0,false]makes the unit animation dead-end, it stuck at AmovPercMstpSrasWrflDnon and without any script input it stays as is, regardless of anything. Regular (older) syntax or ["",0,1,false] won't repro this
["Arma 3","Arma3",217,151641,"Diag",true,"Windows","x64"]
I was nearly sure there was no such issue when was introduced... but hey, the issue is there
Sidenote, there was no script involved other than those lines. No disableAI either
he's no longer an A3 dev? ๐ญ
No
๐, and my whole week goes to ruin. Not only he was going to fix the animation transitions which are broken for holstering with a pistol but he also had many amazing things in the oven... and I guess no only will continue those ๐ข
But he is still at BI right? Moved to Reforger/A4... BI can't be losing talent right now
Iirc he was just on a contract but that's just what I heard
Also leo was never mainly working on A3 iirc. His A3 stuff was just a bonus work AFAIK.
I'm gonna say it's not a bug ๐
the problem is the interpolation time
I'm gonna guess because the move is static 
yeah
that one is just a simple config fix tho. you don't need to be a dev to fix it ๐
nvm I found the issue. it's related to move variants (I hate those)
I'll send the details to Dedmen
dedmen fired him
I trust Leopard's wisdom. If he left then it was the best for him and that is all I need to know, I am still sad thou but I will get over it eventually, or if he left to do his own CDLC then it is all forgiven ๐
for not washing coffee cups after himself
๐คฃ aren't most of you homeoffice anyway?
Huh. I assumed it was because of the laws about keeping exotic animals
in the cup
I hope you guys are not doing a late Aprils fools on me... ๐
That is my fear, that if we can't get a simple config fixed in the game then how can we ask for complex fixes? I guess that if a CDLC dev asks for it we might get it
you are very gaslightable
Looter clothes in contact DLC doesn't have the shemagh assigned as selection in the viewpilot lod, is it possible to get it fixed or is no model stuff happening any more?
["IsSpectator", [player]] call BIS_fnc_EGSpectator
Seems to be broken.
params [["_object", objNull, [objNull]]];
should be
_params params [["_object", objNull, [objNull]]];
Otherwise it accesses the mode which is "isSpectator"
If worth a ticket, let me know.
sure there will be some data fixes for 2.18 so might as well
Ahh, I think I see the logic there. Using that switchMove and playMoveNow makes the animation state nicely
@gaunt depot Revision: 151686 should sync custom radio markers with units on the channel
i.e. when unit is added the markers are created, when unit is removed, markers are deleted
When player uses a bipods, other players will not be able to hear the bipod sound
https://feedback.bistudio.com/T180412
Really need something like this so I can procedurally turn weapon icons into full white icons. All vanilla A3 weapon icons SUCK! Lots of padding, very dark, impossible to display without bright background yet game's UI theme has black background!
Looks okay-ish on a bright sky
Good luck seeing anything during the night
Or if it was on black background UI
Really could use a procedural texture or something to only use texture alpha with rest of components being full so its white
Thought about adding a white rectange under the icons, looks horrible
#(rgb,128,64,3)alpha('\A3\weapons_F\Rifles\MX\data\UI\gear_mx_rifle_X_CA.paa') Something like this?
=>
With this you could make a procedural texture that would place white versions into 8 directions around full color one to build custom white outline! (shadow=2 doesn't support colors, shadow=1 can't be colored white if icon is dark)

Thanks For Coming To My TED Talk
Sorry for wall of text
possible to get https://community.bistudio.com/wiki/inRangeOfArtillery to support pylon based weapons/magazines please?
internally there should be enough functions to make use of. ie from:
https://community.bistudio.com/wiki/getAllPylonsInfo
https://community.bistudio.com/wiki/getPylonMagazines
https://community.bistudio.com/wiki/ammoOnPylon
https://community.bistudio.com/wiki/getCompatiblePylonMagazines
do you have any new information about a possible server.cfg config entry for increasing the timer until BE Query Timeout appears?
I seriously still think having a filter to a control is more better way, reliable, comprehensive and straightforward, not a procedural texture like you suggest.
If we ever get such thing!
RscPicture properties?
ST_PICTURE*
At least Picture, but it also seems to be a very nice thing to overlay anything, even a BG
What if you want it say inside a structured text?
Or a texture?
Though on the other hand, what if you want to modify a procedural texture. You can't have procedural texture inside a procedural texture in single definition
๐ค maybe
Speaking of actually, like, serious question since A2,
Can we get a italic or bold filter for Structured text, even if it only twists the text picture to the right slightly or doubles the text to make it bolder?
I guess it was intended that you switch fonts with <t font=''> to bold or italic versions
Also I found it that configSourceAddonList ModList doesn't detect addon/Mod if the argument is a config parameter not class. Could be handy to have... pwease
https://feedback.bistudio.com/T128076
But what if we have is just this
I mean, having one font family is understandable. My intention after italic bold is more of comprehensive way to put something within a text, like a quote
CJK font issue is very stupid anyways. Why it can't be done?!
Guys you gonna laugh at me but how do i post a new subject on the feedback tracker ? I can't find the button to create a new post
top right star
make sure to pick the good one
Thenk you guys ๐
Even with a background to reduce contrast a little bit, unless you make it bright it looks terrible
I think you made your point :p
Just trying to make it look decent with current tools, I see no way ๐ฆ
make it a (0.5, 0.5, 0.5) rectangle background and call it a day :D
Yeah, white-ish background looks ugly and out of place during the night
Actually you know what, [0.5,0.5,0.5,0.5] makes it look the best out of everything I tried so far
The position of the scroll bar layered the mute buttons in the Role Assignment window in MP; it is impossible to scroll to the end of the list due centered on my nickname
https://feedback.bistudio.com/T180434
Oh yeah I remember seeing this too
oh, I tested on koth before ticket create... thought no bug in map list
Yeah I fixed it there
Can you share with me your config fixes? I've decided to drop my vanilla puritan approach and give modding a try, even if that means that my friends will stop playing with me ๐
I believe you fixed the pistol/launcher holster interaction bug and also the holstering when running sliding animation bug... but I can't remember exactly
PS: Your name is blue again!
(ohhh, that is Veterans, you used to be green right?) ๐
How are you finding the handeDamageEh in 2.16? Or should we wait for 2.18 for the full package?
_context is a huge help
all that I needed
Considering that RscPicture is CT_STATIC, why can't it support colorBackground? ๐ค
What do you expext if you set BG color?
Thanks for the idea yesterday, made something I don't hate
w00t w00t 
Anybody would find ARRAY = arrayIntersect ARRAY unary version useful, to only select unique values?
I use it all the time but it being binary command requires extra variable and most often I'd use it in a single statement.
Or maybe have a proper separate command ARRAY = uniqueValues ARRAY/HASHMAP?
Or arrayUnique
Use a macro?
But it still gonna need a variable creation
inside a macro
actually not sure how it could work with a macro ๐ค
only if its not inside a bigger statement perhaps
yeah I thought the problem was that it's binary
nah only proper method is keys (_array createHashMapFromArray []) anyway
Wow, nice!
Is it faster?
for large arrays, sure
Slower for small ones.
Similarly I want a hashmap batch delete command :P
because _arr1 - _arr2 also blows up for large arrays.
Regarding the BE Query Timeout config entry for server.cfg I have asked earlier (a couple of days ago), I've made a Feature Request: https://feedback.bistudio.com/T180455
Can this PLEASE be fixed for 2.18?
Its a data change, pretty simple one
It is most definitely a bug that appeared after lighting rework, NOT a feature
where can i report spam on the feedback tracker? ๐
https://feedback.bistudio.com/T180459
I have got a question regarding setMarkerDrawPriority
It seems that this command triggers allMapMarkers being sorted by the priority set for each marker.
However, adding new markers without setting their priority creates them with priority 0 at the end of allMapMarkers and thus these markers get drawn on top of markers with a higher priority.
Is this intended?
I guess new markers ignore priority?
For me it seems like engine doesn't care what the priority value is but only looks at the order of allMapMarkers
Yeah, that's what I meant, new markers simply get put into the end of the list without taking priority into account
Guess sorting by priority before drawing is too costly
I guess this was done for performance reasons? Ideally, creating a marker would trigger sorting of the markers array
Yeah, looks like it doesn't and simply gets put in the end
I guess I can just use an event handler
However, would be nice to get this behaviour confirmed by KK so we can document it on biki.
Hmm, okay, that's weird. I created a new marker without setting priority but now the array is sorted again.
Clouds on extreme graphics value still glitches https://feedback.bistudio.com/T175798
I don't think clouds were touched since alpha, could be FUBAR
just don't bank
setOvercast 0...fixed
How long has ambient been missing from the Virtual Reality map? It was since Bootcamp and now missing
Means CfgWorlds โ VR โ EnvSounds โ Default
wrong type name in command info:
https://community.bistudio.com/wiki/get3DENLayer
scalar_string-> should be object
ret nothing -> should be scalar
same goes for get3denParent
Land_SM_01_shelter_wide_F and Land_SM_01_shelter_narrow_F seem to be (semi?) impervious to 80mm rocket strikes. The structure is not destroyed. The ground below gets the decals for rocket hits but vehicles don't seem to take damage from direct hits, just splash damage for stuff that lands nearby (and not on the shed directly)
Does it have a FireGeometry?
i have no clue, will that info appear in the config viewer?
No. If you have diag exe you can use diag_drawMode "FireGeometry"
Or I guess you could also shoot it with a rifle and see if the bullet hits it ๐
too late, already switched to profiling ๐
Wait is diag exe different from profiling?
It's in dev branch 
ahhhh lol i don't have that
okay i'll just shoot it ๐คฃ
well bullets seem to go through (tried with a helicopter minigun from the top)
shooting from inside towards the roof seems to show me bullet hit effects (but no decals) on the roof
smaller one does have hit effects and decals
smoke grenade bounces off
going to try the rocket stuff again
okay for the small one, on the altis variant there are decals, for the tanoa variant there aren't
YES the rockets pass right through ๐
i have it on camera too lol
but then they explode right after passing though 
I present the evidence https://imgur.com/a/DLAvaHD
Rocket passes through, but the explosion is created on the exterior of the shed. Shed is able to withstand 4 80mm rockets. Trucks below still take damage though. And only when the trucks blow up do the sheds get destroyed.
The decals from the explosions are created on the ground beneath
apparently the fuel trucks can take more than one rocket to the face even without the shed ๐
...or sometimes they can't
it's so weird
Okay, more testing -- the big shed dies from 7 rocket hits (in my case, 5 direct and 2 indirect/near misses)
But the damage that goes through it is inconsistent
for example
However a third round that hit above the truck directly did 0.01 damage to the truck, but that did enough hit point damage to blow it up after a short delay
If possible for A4 can we have an official tool for converting WRP files please? I dont like the idea of paying 35 euros every year to a third party dev thats not really participating in the community spirit to convert a file thats advertised as open to the community to dev ourselves.(South Asia .WRP)
If I use BIS_fnc_addRespawnPosition is it possible to set a per player timeout for each individual one created?
Could've sworn there was an environment sound, and never realized it was gone
wrong channel, see #arma3_scripting
Thanks. Could you please make a ticket next time I almost missed this.
get3DENGrid - GameNothing -> GameScalar
get3DENLayer - GameNothing -> GameScalar
get3DENParent - GameNothing -> GameAny
@gray wharf
Revision: 151706
clear3DENAttribute - GameNothing -> GameBool
Revision: 151707
ye, was
there is a FT ticket https://feedback.bistudio.com/T180481
footsteps were different too unless that's the ambient we're talking about
now they are just the "normal" flat-sounding ones, however if you change your stance you can still hear the echoing footsteps
yes, configured the "normal" steps in VR, game data has full collection of VR-steps *P:\a3\sounds_f_bootcamp\characters\footsteps *
seems forgotten create a config of Sound Sets for them
gonna create FT ticket for footsteps too
CfgWorlds โ VR โ EnvSounds โ Default what about it? the config is still present and is not very different from original
I'm not sure if the problem is in the config...
sound is 3D with an audibility radius of 100 meters, who should this sound be attached to? Maybe to the player? Maybe something has changed and the object is now null?
I also assumed the sound could be playing on the map on some 'grid', so I changed the audibility radius to 10 million km, even so there was no sound 
https://feedback.bistudio.com/T180485
Ticket VR footsteps
Grenade impact sound will plays on throw (in unit's hands) if unit aim in some angles
https://feedback.bistudio.com/T180487
https://community.bistudio.com/wiki/isNil
Syntax 3. why not Control, Display, Location, etc.?
good question
has this been forgotten? https://feedback.bistudio.com/T172248
21:47:15 Error position: <setParticleFire []>
21:47:15 Error 0 elements provided, 19 expected```What we know is `setParticleFire` does only have three params within the array... what does it mean 19?!
antispam, boom
i'll check it, could be typo in error message
23:37:30 Error in expression <Max == _fMin}) exitwith {
_fMin = _fMin / (1000^_logMin);
_fMin = _fMin tofixed >
23:37:30 Error position: </ (1000^_logMin);
_fMin = _fMin tofixed >
23:37:30 Error #: Zero divisor
23:37:30 File a3\Functions_F_Enoch\EM_Core\fn_frequencyToString.sqf..., line 50
23:37:30 Error in expression <Max == _fMin}) exitwith {
_fMin = _fMin / (1000^_logMin);
_fMin = _fMin tofixed >
23:37:30 Error position: </ (1000^_logMin);
_fMin = _fMin tofixed >
23:37:30 Error #: Zero divisor
23:37:30 File a3\Functions_F_Enoch\EM_Core\fn_frequencyToString.sqf..., line 50
Switching to the Spectrum Device instantly triggers that script error. Vanilla Game + Contact only. v217,151686)
2.17 cannot be a vanilla game
oh you mean no mods?
@untold sky another compiler issue?
well, relatively large negative power produces 0, not sure how to handle this in script, set result to infinity?
If I have time this week I'll try to play the contact campaign. As it uses some heavy math at places there might be more of these script errors.
Okay, I know, this is a very nitpick. And it is very stupid one. But, listen to me, Showcase Cultural Property is in June 24th, not canon cease fire date, August 10th
Would love to get an estimate
Have an important fix there, want to know if I should make a crutch now or wait for the fix
Is it not merged in prof?
Not sure. Not in stable, but in dev.
Looks safe to merge to prof
Still, any estimates? Month, two, late summer?
between 1 and 3650 days
Soon
Not sure if this was actual error or just fine, when getting in vehicle for Black Wasp Plane B_Plane_Fighter_01_Stealth_F, the Pose command will return "ERROR", btw the name of playing animation call pilot_plane_fighter_01_getin
which branch is this on
main one, 2.16
Out of curiosity. Is there a reason we cannot set marker colour in RGB format? The markers themselves seem to support any colour ,just that we are limited to what's defined in config.
All configs should have preproc, including rvmat ๐ค
useless, doesn't tell anything useful about the crash.
Why would there be a frozen mdmp with a crash? crash!=frozen
If it were EMM extension, then you could look at the mdmp yourself and see that
yes
Don't remember. Never tried.
But sounds like a bug
, can be used as a alternate for semicolon. it behaves similarly. Its ugly and shouldn't be done, but it doesn't cause issues in all these cases.
The returnMatterModule one should be broken ๐ค
The code in ther is also nonsense, as if someone stopped typing half way through ๐
I assume its unused.
no
That. Which is unlikely
And game does not create a freeze dump when that happens?
Is it actually freezing, or is the progress bar just going till the end and then it never leaves the loading screen?
Not reproable without CBA?
the progress bar freezes at 100% (or maybe less, at least it freezes), if you click into the game multiple times Windows opens a prompt that the program is frozen (you know which one). you can either wait (for nothing to happen) or kill the process. killing the process: then you can join onto the Server without problems.
Sounds nice. Needs a ticket and time from reyhard. You need a CfgVehicles class for it right?
If we just make it hidden in editor then maybe we don't need to do localization for it
Adding assets is quite alot of work
Gonna make one. I'm okay with no 3DEN entry as it is invisible without a flag anyway.
Sounds like a perfect thing for it. Ticket please
Speaking of new objects, having UserTexture that is a ground decal will be very useful too
Probably can also be hidden and/or use same texts as normal UserTexture1m_F?
Recent use case: #arma3_scripting message
Having scripted footsteps that follow terrain. Normal user texture won't bend at terrain cell edges
We only store that information per-class though. Not per entry.
So best I could give you is to let it return the parent class, which will return false-positives for mods that edited the class, but not the specific property you're looking for
Hmm, I could imagine why
what message were you replying to? Discord is doing a fucky wucky, reload doesn't help.
Same for reply to my post ๐ค
All same keys 5k elements
arrayIntersect: 0.115857 ms
hashmap: 0.252246 ms
All different keys 5k elements
arrayIntersect: 209.955 ms
hashmap: 2.71939 ms
Huh. Didn't expect that. I'd expect creating a hashmap to be more expensive
It is not a bug. Its a feature. The flares are signal flares. Not illumination flares.
There is no bug to be fixed here.
But we can probably add illumination flares, they've been on the backlog for quite a while.
But adding assets is alot of work
Correct.
setMarkerDrawPriority has a "sort" call inside it. if you call that command, ALL markers get resorted by their draw priority. If you don't run that command, then thats not done
Nice. Thanks for the info. Need to update biki
Yeah I already investigated further and they are indeed got changed to be signal flares, there are text descriptions with that. What's funny is that whoever did that also tried to increase its brightness brightness from 2 originally to 12 (which is a VERY bright).
It feels like they couldn't fix the flares not being bright anymore and instead of investigating why turned them into signal flares or something, lol
Performance I'd rather not.
But what we probably could do is, on insert skip all markers with higher priority and insert into the middle of the array.
Because the high prio ones are at the start, that should be somewhat simple.
I think that would be easy to do.
Actually new markers get added to the end of the array. Not the front?
And I assume markers are rendered from front to back? Or are they rendered in reverse?
Finding insertion point from end like I described would also work though
will not be fixed
this clouds library always felt to me like "nice for screenshots" but glitchy when moving, not "half-assed" but almost - like unfinished
I think I've ever seen a cloud that looked like Monster Energy (I mean having very same pattern three times) and uh, how long it was? That was actually strange looked
Iirc they get added to the end. That's why they are drawn on top of markers that even have a higher priority.
It's not a big deal as long as it's documented.
An easy workaround is to set priority on map event handler. So that order is correct when the map is opened.
scalar_string is the argument?
That command takes
// Object, Group, Waypoint, Marker, ID
Object, group, array, string, scalar
Yes. But also I don't know if we want to do that
the wiki was apparently ok already
get3DENParent - GameNothing -> GameAny
get3DENParent's page says "Object" though?
Worst thing is that it is desyncronized in MP and you can't use clouds as reference for directions ๐
nor hide during dogfights
just let's do same solution as for particles
extreme = ultra โ no bug 
I did some tests with the clouds and they can be synced in MP, but get desyncronized after a while
Plus weather is fucked in general and you never get same overcast values over the network reliably
And these clouds are very dependent on overcast value
unlike in Reforger where it is synchronised and clouds cast shadows
(reply to #arma3_feedback_tracker message)
The fix for first compiler issue wasn't released at that point.
I cannot reproduce on latest internal. Though I am just equipping spectrum device in VR
"relatively large negative power produces 0" that should not be possible here, there is no large negative power.
#define UNITS ["Hz","kHz","MHz","GHz","THz","PHz","EHz"]
_fMin = _fMin / (1000^_logMin);
UNITS # _logMin
The highest possible here is 1000^6 that is 1e+18, that is still well within limits. (max value is 1e+38)
_logMin must be between 0 and 6.
And even with its highest value at 6, we don't get a div by zero here.
I cannot reproduce the problem in debug console with the possible values.
1/ (1000^6) is 1e-18
There is no div by zero
Maybe our max/min are broken with extreme values
-1e39 max 0 is 0, all fine. (negative infinity)
1e39 min 6 is 6, all fine. (positive infinity)
_fMin = -1e39;
_offset = -1e39;
(floor ((log _fMin + _offset) / 3)) max 0 min 6;
There is the bug!
-nan(ind)
log of negative infinity, results in nan indeterminate. Which then propagates through.
I'd say we can make min/max handle nan.
-nan max 0 should be 0 (The bigger of the two)
nan max 0 should be nan (The bigger of the two)
months
legacy
"opens a prompt that the program is frozen (you know which one)"
Thats what I meant.
mh our freeze dumps might be disabled during loading screen.
If you go to task manager, details panel and right-click on Arma there, you can create a memory dump. Send me that and I can see what its been doing during the freeze
how do you spawn a model if there is no CfgVehicles entry and you cannot use simple object?
Had a brain derp. Meant to say no 3DEN entry, hidden from 3DEN.
Sounds nice yes. Needs ticket and run through our asset pipeline.
F discord #arma3_feedback_tracker message
Discord cannot even show the messages that I replied to, to myself
Yeah intercept shortens it to at most 2 types
Can you make a ticket for that insertion ordering bug?
So then it shouldn't be "object", and what it currently is. is correct
If I don't forget. Yes.
Anyway KK fixed it
Ah derp yes ofc. later drawn things are drawn on top.
Setting priority goes through ALL markers to resort them though. Not quite nice. And its an easy fix for us to do
Not that one, only the return values
It returns -1 if the item is invalid, or it couldn't find a parent.
Otherwise it returns https://community.bistudio.com/wiki/Eden_Entity
oh okay, nice & thanks
The ingame info for create3DENEntity/get3DENEntity/... also says it returns any, technically not correct but ugh
Actually yes that broke with the compiler upgrade.
Our doc even says so
That nan handling fact is not documented on wiki.
Documentation for the instruction also specifically says that should happen.
https://www.intel.com/content/www/us/en/docs/intrinsics-guide/index.html#text=minss&ig_expand=4523
https://www.felixcloutier.com/x86/minss
Doc specifically says
dst does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values.
If only one value is a NaN (SNaN or QNaN) for this instruction, the second source operand, either a NaN or a valid floating-point value, is written to the result.
But our compiler is so smart, and flips the left and right values.
instead if
a min b
it does
b min a
_nan min 0
is -nan
0 min _nan
is 0
darn you Microsoft :U
Write it in assembly 
I hope GCC on linux isn't making that same mistake, can't test that rn.
log (-1e39) min 0 should return 0
0 min log (-1e39) should return -nan
Maybe someone has time to check linux on prof or dev?
I already made a ticket about incorrect float optimization, guess I'll make another.
( a<b ? a : b) should definitely not be optimized into b minss a ๐
Did you check on compiler explorer?
typing as you type
It again only messes it up in /fp:fast ๐
Yeah its a MS bug. even with fastmath, it shouldn't just swap the order
Well microsoft gives themselves the opt-out in their docs
Special values (NaN, +infinity, -infinity, -0.0) may not be propagated or behave strictly according to the IEEE-754 standard.
That results the code
Print (a < b);
Print (a < b ? a : b);
to print out
false
a
Sure.
if false then a else b.
Is a!
Microsoft is also absolutely all over the place
Expected result: 0;0 | 0;0
VS2013 Release: 0;-1.#IND | 0;-1.#IND
VS2013 Debug: 1;-1.#IND | 1;-1.#IND
VS2017 Release: 0;0 | 0;-nan(ind)
VS2017 Debug: 1;-nan(ind) | 1;-nan(ind)
VS2022 Release: 0;-nan(ind) | 0;-nan(ind)
VS2022 Debug: 1;-nan(ind) | 1;-nan(ind)
GCC13.2 O2: 0;0 | 0;0
GCC13.2 O0: 0;0 | 0;0
GCC8.5 O2: 0;0 | 0;0
GCC8.5 O0: 0;0 | 0;0
That GCC there is with the -ffast-math flag too.
GCC gets it right with fast math, MSVC doesn't
Use GCC on Windows too 
for Tanks and APCs in config sets default sounds on collision Wood, Building & Armor, but data has special sounds for them
https://feedback.bistudio.com/T180603
https://feedback.bistudio.com/T180604 @untold sky (marker priority)
didnt touch it because of backward compatibility, it accepts a bunch of types and ignores unsupported quetly, we dont want code break and error instead
It handles Object, Group, Array, String, Scalar.
And it accepts Object, Group, Array, String, Scalar.
As of
Arma 3 logo black.png
2.16
createMarker/createMarkerLocal do not care about drawing priority and add the created marker to the end of the array, making it top priority
pretty sure this wasnt touched in 2.16
@gray wharf ^^^ dont think the create marker was automatically sorting markers ever, setmarkerpriority was doing that me thinks
correct, according to what Dedmen said here #arma3_feedback_tracker message
but createMarker/Local seem to add the newly-created marker "at the end" of the array, therefore making it high prio
what I think I understood at least ๐
but this is why you have separate command to set priority
create marker , set priority
if you needed it
yes, hence the note
although, it may not be seen as a "problem", IDK
i did not want markers get sorted for no reason every time marker is added or removed
dedmen said he might have a fast way of doing it. As I stated earlier it's not a problem really, just needs to be documented. However, if createMarker would handle that and it can be fast, why not.
Hashmaps are lovely. Please give batch delete so I can make even better temporary hashmap abominations :P
now im meant to beg for setproxymodel too arent i
That is correct we definitely shouldn't do that.
But inserting at correct position is alot cheaper than doing a full sort.
We just end up having to bump all elements after the new one (like how deleteAt 0 bumps all elements after the first one)
Which will be only markers that got their draw priority set, which should be very few
We have deleteAt with array of indices.
Same thing for hashmap with array of keys, ez
I never ask for difficult things, aside from AI fixes :P
Ah frick
no can't do same thing because deleteAt [array] might be a key = [array]
not an array of multiple keys
would need separate command. deleteRange would work but I don't like that :U
deleteRange makes the most sense though, so probably gonna be alt syntax for that, even though I don't like
but wonder if its even worth it.
Not much performance impact (number of hashmap lookups doesn't change).
With a simpleVM loop its probably not much difference
I was wondering. What's the method to test simpleVM these days?
diag_testScriptSimpleVM {code} on dev or profiling. To check if its applicable
{_hm deleteAt _x;} forEach _toDelete is probably good enough.
vs _hm deleteRange _toDelete
there is no way to turn SimpleVM off, without switching to non-profiling branch exe
so not without game restart
Profile/perf uses SimpleVM automatically and stable doesn't use it at all?
yes
ok, can work with that.
It's only enabled on 2.14+ perf and 2.17+ stable
so not on 2.16 stable
Right, SimpleVM is good enough for this. Hashmap creation cost is most of the expense for 300 elements minus 299 elements now.
99% inherited config from existing flares and no new models required, so hopefully not too much work
i can literally write the config for them
Could the #legacy be worked around by letting CfgMarkerColors work in mission config?
Don't know when they are loaded. That might work
Yeah, that would be splendid. Would reduce alot of limitations
that is assuming that markers are always sorted
They will be
If anything with priority is sorted to the end.
And anything without priority is inserted before the priority ones. It'll always be sorted
we just have to make all markers priority 0 by default
I thought they already are
Though I don't know what happens with the priority property when a network marker update comes in, could it reset the priority to zero? Such a thing probably wouldn't resort it correctly. But as the command has local effect only, I wouldn't expect a remote update to reset it
priority is part of network sync if I remember correctly
however here is a small problem. Say we have 20 markers with priority 0. they are drawn in order of creation. How do you create another marker that is on top of 10th but below 11th?
Wiki says they are already 0 by default
That'd be weird, if the command with local effect, can have global effect if you run another command after it 
local markers are not synced so no problem there
You'd have to set 11-20 to priority 1. And then when you create another marker, it'll be between 10 and 11 (at the end of the priority 0 partition)
yeah 10 times sorting
If you want 10 markers with priority, that already is resorting for each right now. Wouldn't change.
We could optimize it though, we don't need to resort the whole markers array when we change priority of one. We just need to move that one around
we can ++ priority with each marker creation
๐ I don't get it
whn you create next marker you create it with higher priority then if you want to insert marker in between then you can. priority is float you +1 it for each next marker. sort of like priority sorting in action menu
I don't think we have a usecase for inserting marker specifically inbetween two others, atleast so far we didn't
maybe because it is a pain in the rear to do it now
since you will have to kinda set priority on bunch of markers to make space
creating markers with increasing priority shouldnt break anything
except existing scripts that expect their 1 priority to be ontop of all other normal markers. But thats only since 2.14
If a mod required that all markers are ordered, they could use markerCreated EH and handle incrementing ID by themselves
bool true ++ priority false or 0 set to 0, a number - specific priority?
GameBoolOrScalar
I don't see the need for us to solve that.
So far noone seemed to have that use case, and if they want it its already solved with this
GMarkerCounter = 0;
addMissionEventHandler ["MarkerCreated", {
params ["_marker"];
_marker setMarkerDrawPriority GMarkerCounter;
GMarkerCounter = GMarkerCounter + 1;
}];
In this case I am not sure why are we even discussing it, don't get @daring wagon point, everything is doable with current tools
The bug is that the sorting doesn't stay
marker with priority 0 will draw over top of marker with priority 500
I'll just turn the append into a insert_sorted and thats solved
yeah do that
yeah, but since creating a new marker (that is supposed at zero prio) makes it take precedence over markers with prio 5000โฆ you may understand the end user can be confused ๐
That's what I said in the beginning...
I just asked for an explanation so I can document the behaviour on the wiki.๐
too late, now you did it
Any chance to get HandleHeal to be fixed in regards to the return value being ignored?
Might need HandleHealEnd or something, as it healing might not end in success (unit ran off and such)
Its all script fixable but a huge mess and a crutch right now
I only want it for this (option to override engine healing), meaning the fix EH should (in theory) be much smaller than adding an extra EH
Can we just have a general system for disabling vanilla actions :/
I'm still scarred by that ACE engineer flip.
Had some write up about it year ago, can't remember if its in any ticket
Sent, hope it helps
https://feedback.bistudio.com/T180633 Script error in Eden editor caused by a missing define in vanilla function. Perhaps that could get fixed as I get reports quite often because ppl think its my mod. Steps to reproduce are added as well.
Clients not knowing owner ids of remote entities is the weirdest design decision ever
Would it be possible for an alt syntax/change to setAperture for a coefficient of the current auto setting? Negative numbers maybe but could have bwc problems
bwc?
Backward compatibility I think
Dedmen said bcomp, KJW said bwc, I believe we should agree on one explicit term please ๐
e.g 'โcp' ๐คฃ
But Why I Care
๐ maybe cp isnt the one
compatibility
emojis ๐ not ๐ optional ๐
backwards compromise
The marker insertion sorting issue is fixed next dev
_createMarker = {
params ["_name", "_offset", "_color", "_priority"];
_marker = createMarker [_name, getPos player vectorAdd [_offset, _offset]];
_marker setMarkerType "hd_dot";
_marker setMarkerShape "RECTANGLE";
_marker setMarkerBrush "SolidFull";
_marker setMarkerColor _color;
_marker setMarkerSize [100, 200];
if (_priority != 0) then { _marker setMarkerDrawPriority _priority; }
};
["A", 0, "ColorBlack", 0] call _createMarker;
["B", 10, "ColorGreen", 1] call _createMarker;
["C", 20, "ColorRed", 0] call _createMarker;
Could improve performance of setMarkerDrawPriority, but removing from list and reinserting at correct position is more than the 5 minutes I'm giving this fix
I HATE that Respawn copies events and variables from the corpse. There is a solution to clean them up on Killed, but this won't work if you want to keep EHs and variables on the corpse
Can we have some kind of flag to stop that?
And another issue is that EntityCreated and EntityRespawned on server order is not even guaranteed
14:12:08 1573420 :: EntityRespawned :: ["B Alpha 1-3:1 (Sa-Matra (2)) REMOTE (B_Soldier_F) LOCAL=false","1b972acf700# 1780782: c_nikos.p3d REMOTE (B_Soldier_F) LOCAL=false"]
14:12:08 1573420 :: EntityCreated :: B_Soldier_F sim=soldier (B Alpha 1-3:1 (Sa-Matra (2)) REMOTE) local=false
...
14:14:56 1597262 :: EntityCreated :: B_Soldier_F sim=soldier (B Alpha 1-3:1 (Sa-Matra (2)) REMOTE) local=false
14:14:56 1597263 :: EntityRespawned :: ["B Alpha 1-3:1 (Sa-Matra (2)) REMOTE (B_Soldier_F) LOCAL=false","1b8095be800# 1780880: c_nikos.p3d REMOTE (B_Soldier_F) LOCAL=false"]
So much headaches because of this behaviour
ENTITY setNoEventsAndVarsReplication BOOL or something like that
So if you think you can setup new variables in EntityCreated you end up with them being replaced by respawn behaviour
I'd suggest having respawn logic check if variables weren't already set, but this could break backwards stuff
Nevermind, you can't even do that because setVariable ["var", nil] makes it copy the nil over your EntityCreated set variable after respawn
setNoEventsAndVarsReplication could break mods/addons that rely on this behaviour though
I guess there is no solution
Only suffering
ENTITY deleteVariable STRING maybe
IIRC this is because the respawn process is actually carried out on the client, and then the server gets told about it afterwards.
Yeah, some network BS
Which is oddly flipped from the connection process, where the unit is created on the server and then moved to the client.
Although maybe with a respawn the new unit still gets created on the server. Not sure I checked.
ENTITY deleteAt STRING by analogy with arrays and hashmaps
I'm not following what this is for.
_object setVariable ["string", nil] already deletes?
Not quite, respawn logic still copies this nil
I guess key still remains or something?
--- EntityCreated ---
21:49:21 ["5678563","koth_ehs_hitparts_history",<null>]
--- New var set ---
21:49:21 ["5678563","koth_ehs_hitparts_history",[]]
21:49:21 ["5678564","koth_ehs_hitparts_history",[]]
21:49:21 ["5678565","koth_ehs_hitparts_history",any]
21:49:21 ["5678566","koth_ehs_hitparts_history",any]
21:49:21 ["5678567","koth_ehs_hitparts_history",any]
21:49:21 ["5678568","koth_ehs_hitparts_history",any]
...
My setVariable lives for two frames and then respawn logic replaced it with nil I attempted to set on unit's death to stop this
Yep
ah, it has addendum there for objects.
Maybe respawn variable replication logic can stop copying variables that are nil?
Has a slight potential to break something somewhere
Being able to delete variables for good could be a better solution
HandleRespawnVariables, HandleRespawnEvents

ENTITY setNotRespawnableVariables ARRAY ex:["var1", "var2"]
ENTITY setNotRespawnableEvents ARRAY ex:[["Respawn", 1], ["Killed", 2], "HandleDamage"]
Or better add/remove so its mod-compatible
So solutions so far:
- Stop respawn logic from re-adding
nilvariables (only truenilreturned by the command so other nils are still there?) ENTITY/ETC deleteAt STRINGto delete variables from variable namespaces- Add/Remove variables and events that shouldn't be replicated on respawn (most complex one)
What about a global setting to disable var/EH transfer on respawn?
Could break mods that expect this to happen, thus an idea to stop transfer selectively
ah I see, can't make it mission-specific.
(4. Make an ugly crutch that adds variables again on respawn too which I'm going to do for now)
Deleting variables for good instead of setting nil sounds like the best solution to me
Doesn't solve the issue for case where you want to keep events and variables on the corpse though
If there is a chance for this to get looked at, I'll make a ticket
no, gets created on the client
Please explain what problem you are trying to solve rather than asking for crutches
Problem: I want to get the target position selected by the player in the artillery computer
Current solution appears to be loops and an EH trying to interpret stuff from the UI (apparently MapSingleClick EH doesn't work with artillery computer?)
https://forums.bohemia.net/forums/topic/191406-get-artillery-firing-position-in-computer/?tab=comments#comment-3259454 (from 2017)
Request: Let getArtilleryComputerSettings return the target position, or allow MapSingleClick work with artillery computer, or new EH for artillery computer that returns fire position and settings either when the map is clicked or the artillery is fired (or two separate EHs that do the respective tasks)
Display EH mouseClick something?
It would definitely be useful to be able to get an artillery vehicle's current target position without going through the UI, so it can work with AI artillery and not just players
- You can't stop respawn logic from re-adding same variables to newly spawned unit
setVariable ["name", nil]on the corpse makes the engine copy thatnilas variable is not really deleted- Overwriting over re-added variables is hard because order of MEH
EntityRespawnand MEHEntityCreatedis not guaranteed #arma3_feedback_tracker message
Speaking of artillery UI. It would be nice if it had an IDD so we can actually identify it without looping though all displays and controls to find it.
https://feedback.bistudio.com/T171670
Been 1 year, can somebody have a look at it please?
please make a ticket and tag me with the link
I will have a look after I dig myself out of this rabbit hole I'm in
Thanks, would help us a lot
Went back to this, turns out setGroupOwner is the only safe way of doing it, otherwise locality may break and forever get stuck at some client for the vehicle itself, even if driver and group changed locality
So that warning Script command setOwner cannot be used for object 'B_UAV_AI'. Use setGroupOwner instead. is indeed very much true, if you don't follow it, shit breaks
Here! https://feedback.bistudio.com/T180721
Added Nikko's suggestion to make it more generally applicable (both AI/player). And the note from R3vo too in additional comments.
There might be technical inaccuracies but i hope it's clear from the ticket what the desired end goal is!
I wonder if we can just get general engine-driven calculation for predicted projectile hit position? This way you could run it in Fired to find out where AI was firing
It will differ a bit from wanted fired position though
Add the suggestions to the ticket please
Well at least in my case I would like the intended position -- not opposed to extra stuff of course!
there are some cases in which the rounds fall far away from the desired position (poor accuracy but otherwise normal precision/dispersion), wanted to try diagnosing/repro-ing that
(seems to happen at particular azimuth and range with respect to the unit -- i observed it with one of the GM MLRS, but want to make a repro)
I can't post on FT right now but my general feeling for this would be unit EH for artillery firing that gives info about the target (possibly could skipped as can be replicated with Fired EH + following items), a command to get what an artillery unit is currently aiming at, and a command to get the AI's current fire mission
Finally went to check this, you did change roadkill damage source.
Dev/Perf:
18:18:33 ["B Alpha 1-2:1 (B_Soldier_F)","",2.5833,"B Alpha 1-4:1 (Sa-Matra) REMOTE (C_Hatchback_01_F)","",-1,"B Alpha 1-4:1 (Sa-Matra) REMOTE (B_Soldier_F)","",false,0]
Stable:
18:30:19 ["B Alpha 1-2:2 (B_Soldier_F)","",3.06911,"B Alpha 1-2:1 (Sa-Matra) (B_Soldier_F)","",-1,"<NULL-object> ()","",false]
I do like this change because it makes more sense BUT it could be a huge backwards compatibility mess if somebody expects roadkill damage to be setup exactly like it was for 11 years.
In summary, roadkill damage is attributed:
- Dev/Perf:
source=vehicle,instigator=driver - Stable:
source=driver,instigator=null
Would you check first for instigator anyway?
Define โhugeโ because I donโt see it
Well, maybe I exaggerated it a bit but it broke my scripts, I always assumed that roadkill damage is:
ammo == "" + source == unit + instigator == null + !isNull objectParent _source
Who knows who else does it like this
Looks like its not an issue in ACE: https://github.com/acemod/ACE3/blob/0e2c7167a5f83026e9407a77fa2043bc46daabb7/addons/medical_engine/functions/fnc_handleDamage.sqf#L171
Maybe others can chip in and see if they assume roadkill damage to be setup exactly like this
Wish you could also change ammo to something like #roadkill or something
but this is even more backwards compatibility mess as there is no such CfgAmmo entry
11 years too late for such things
So what happened to your script you got error?
Maybe, make a ticket
Can add fake ammo to cfgammo
Not error but it used to be correctly attributing damage as roadkill and displayed a cool icon. Now its unknown damage.
Hardly game breaking
Who knows, maybe it will break for somebody else who messes with HandleDamage
Considering how unreliable it was before hmmm
If we go this way, I'd also add dummy entries for other damage kinds - Fall damage, Collision damage (physx barrel hitting you), drowning damage, fire damage (particles)
roadkill meaning run over by vehicle right?
Yes
Back in OA there was no damage source, in A3 it started to be driver (currentPilot), now it got changed to be vehicle and driver as instigator
Maybe add another context instead?
We just have this
_isMainSelection = _selection isEqualTo "";
_return = if ((_projectile isEqualTo "") and {not (_unit isEqualTo _source) and {(isNull _source) or {((side _source) getFriend (side _unit) >= 0.6)}}}) then
{
//reapply previous damage state
if (_isMainSelection) then
{
(damage _unit)
}
else
{
(_unit getHit _selection)
};
}
else
{
if ((_isMainSelection) and {(_damage >= 1)}) then//auto-removal if unit is dead
{
if (!(alive _unit)) then
{
_unit removeEventHandler ["HandleDamage",_thisEventHandler];
};
};
_damage
};```
Yeah that what I was suggesting before, but I'm not sure how much work this will entail to also send damage kind alongside other stuff, thus my thought that sending fake ammo class names could be easier.
Personally I rename _ammo to these # monikers after decyphering where the game got the damage from
and this
// units from playerSide cant be run over
if
(
(_isInfantry)
&&
{
(SPE_CadetMode)
&&
{
(_projectile isEqualTo "")
&&
{
(_unit isNotEqualTo _source)
&&
{
(isNull _source)
||
((_sideSource getFriend _sideUnit) >= 0.6)
}
}
}
}
)```
should be ok here
_selection isEqualTo "" is far from ideal as hit parts can have empty selection names, better check for _hit_index < 0
yep from the looks of it. if easier to identify, would be handy I'd agree
getArtilleryComputerSettings actually doesn't look in the artillery computer at all, it has no reference to any artillery computer or UI.
It just finds the current selected weapon.
It basically just gets the current weapon and loaded magazine of the vehicle you are sitting in.
Even sitting at a MG would return that, no matter if artillery computer. How fun.
its a combination of currentWeapon/currentWeaponTurret/currentMagazineTurret and some config lookups to find the displayname.
What a weird command, but its from OFP era so.. I guess that justifies that..
shownArtilleryComputer actually can look up the UI. And it could pull the.. targetETA... eh..
Mh, it seems it also doesn't know what position is selected. But you could read it back from the text in the UI controls ๐
Mh yes SetArtilleTarget Ah yes sure SetAtilleryTarget
oh wow the artillery target is not stored in the artillery UI, its stored in the global UI thing.... The same global UI that handles all the AI commanding, because clearly AI commanding should be in UI code, not in AI code. Ahhhhhhhh

Adding that to the get settings command is easy then, it can just pull it from global UI. So ticket for that.
But omg ugh I hate this
I wonder whos idea it was to hardcode two entire menus (arty and UAV terminal) when 99% of it could be scripted?
I could be scripted nowadays
back then that would've meant adding script commands to make it possible to script that, which was too much work.
And after the script commands were there, transferring it over was too much work.
Interestingly, roadkills from empty vehicles are also now registered with source=vehicle vehicle that hit you
["B Alpha 1-2:1 (Sa-Matra) (B_Soldier_F)","",0.847368,"1e46fe1e040# 1791203: mbt_02_arty_f.p3d (O_T_MBT_02_arty_ghex_F)","",-1,"<NULL-object> ()","",false,0]
Doesn't work with generic PhysX objects though, still source=self
["B Alpha 1-2:1 (Sa-Matra) (B_Soldier_F)","",0.528751,"B Alpha 1-2:1 (Sa-Matra) (B_Soldier_F)","",-1,"<NULL-object> ()","",false,0]
I still detect it though
So much effort just to see this
Would it be possible to expand getArtilleryComputerSettings command to return the display?
yes thats what i proposed on ft ticket
Very nice.
must have missed it
Are you gonna change it?
might as well to keep it consistent, is there current ticket?
will need repro for that as well
I guess we can use this one as that roadkill change derived from that
v = createVehicle ["Land_ToiletBox_F", player modelToWorld [0,10,0]]; pos = getPos v; onEachFrame {v setPos pos; v setVelocity [0,100,0]}
```and touch it
could you add it to the ticket plz
Done
v = createVehicle ["Land_ToiletBox_F", player modelToWorld [0,10,0]];
pos = getPos v;
onEachFrame {v setPos pos; v setVelocity [0,100,0]};
u = group player createUnit [typeOf player, v, [], 0, "CAN_COLLIDE"];
```Expanded it a bit so it auto kills a unit with collision damage
well this is tricky, maybe box need to be moving in order to be considered source?
It is moving, I'm just keeping it in place for easy debugging
Otherwise flying toilet could kill units
if it just has velocity value but not moving it is suicide
