#arma3_feedback_tracker
1 messages ยท Page 20 of 1
https://community.bistudio.com/wiki/localize
//currently
localize 'STR_TAG_key1'
//needed functionality
if (isLocalized 'STR_TAG_key1') then {localize 'STR_TAG_key1'} else {'default'};
//proposal
localize ['STR_TAG_key1','default']```
Isn't that what BIS_fnc_localize do
dunno
it doesnt provide a default, just returns the key
BIS_fnc_localize = {
if (isLocalized _this) then {localize _this} else {_this};
};```
Ah that's the point
At least a function would be cool
https://community.bistudio.com/wiki/lineIntersectsSurfaces
LOD2 pareameter dosent seem to work
Why doesn't AI use disposable RPGs if he has them in his backpack?
disposable launchers isnt a vanilla feature
arma is not just about vanilla
I'm sure there is some AI overhaul mod that allows AI to use it, but idk why you would expect vanilla AI to be able to do something only possible with mods
Well, technically you can put a non-disposable launcher in a backpack, and they won't use that either :P
I don't think they'll draw a pistol that you put in their vest either, for example. They just don't look there.
weird, could/should they?
That is very low on my list of AI fix/feature priorities.
Also you could workaround it in script if you really wanted to.
little bump for this, mentioned account literally spams nonsense into the ticket
LOD2 works ONLY if LOD1 wasn't found in the model
only a single LOD is checked for intersection. LOD2 will only be checked if LOD1 is unavailable.
warning line in wiki
Which is counterintuitive and sucks and makes you run the same command twice when you want to test against several LODs
Pretty much how usage works in practice:
_distance < 3 && {
count lineIntersectsSurfaces [_from_asl, _to_asl, _zombie, _target, true, 1, "GEOM", "", true] == 0
|| {count lineIntersectsSurfaces [_from_asl, _to_asl, _zombie, _target, true, 1, "VIEW", "", true] == 0}
}
``` and LOD2 is close to be useless
Thanks for the info, I ended up doing two calls aswell
dealt with, thanks
silly question, yet you could still script it
I assumed that this should be built into the game mechanics... but alas... no item - no problem
i guess i disagree. i dont know any game where the AI is set up to use mechanics not already in the vanilla game
a modded reloadable launcher id expect them to be able to use if its modded correctly by the authors
but thers no disposable launcher in vanilla so i wouldnt expect AI to be set up to use one
dont think im defending the arma AI though, i was arguably its most outspoken critic for many years, specifically on the (lack of) dev resources assigned to developing it
I'm pretty sure the AI can actually use disposables; as far as the AI knows they're the same as reloadables, they just get removed by scripted systems after firing.
The question was actually about the AI pulling weapons from their backpacks into their active slots. I suppose they mentioned disposables because that's the weapon type you're most likely to need to do that with, but in theory such a mechanic could apply to any weapon.
Sanity check before post a ticket, we have no Mission EH to detect cameraView update yes?
Let's say fire one whenever you click Num Enter
https://community.bistudio.com/wiki/Arma_3:_Event_Handlers#OpticsSwitch theres this but obviously not a mission EH, and Im not sure if it triggers when switching from internal/external or just gunner/whatever
Only happens when you ADS or left from ADS, not changing view mode
Oh, and also, I recalled something... what about an idea about Curator EHs, that you can add EHs into curator-placed objects (not module) so like
{
_x addEventHandler ["CuratorObjectEdited",{
params ["_curator","_entity"];
if (typeOf _entity == "MyAwesomeModule") then {
};
}];
} forEach allCurators; //meh
_myAwesomeModule addEventHandler ["CuratorObjectEdited",{
params ["_curator","_entity"];
}]; // much smoother```The fact I can't add EHs to objects botheres me so much
Can we get a up to date comref ? 
that would be #community_wiki ^^ is the "old" comref that bad?
isn't it "old" ?
from โ 17.08.2022 17:21
it lacks maybe v2.14+ commands, but it should cover 98% of commands still
2022-08-17 indeed #community_wiki message
I'll try and update it if absolutely needed
Would it really be hard to add some connection/disconnnection logging in the RPT :/
No I can't.
Because the issue is trying to support people with connection problems on this discord.
Also I think it usually fails too early for any of those event handlers to fire.
You end up with an RPT full of "normal errors" and absolutely no indication of whether the client ever tried to connect and how far it got.
Didn't realize you were talking about why a disconnect happens, yeah that would be really nice to have
Log the user message, or the technical reason (or both)?
Well, for a start there are 13 different client states and absolutely no logging of the progression through them on either client or server.
The closest thing we get to a disconnect indicator is the sodding EPE release :P
Just realized return value for drop is not implemented in 2.18
http://feedback.arma3.com/view.php?id=23847 (AI will be stuck to a spot after disableAI "FSM"; enable AI "FSM")
@thick hare Any news on this? ๐
@tulip wasp I'm sorry, not really. If you have more situations where the AI would hang after disabling/enabling FSM let me know
Thanks for the quick answer! I just "accidentially" reproduced the bug in editor/zeus. I was testing if disableAI "FSM"; .. make unit run somewhere ... enableAI "FSM"; would be a way to reduce the "combat dance" for units that are retreating or charging.
The results looked pretty promising ...except that once I re-enabled the FSM's all the units where stuck to a spot. More specifically they would not follow waypoints orders and when zeus-moved away would just walk back to their spot
Actually, I can't get the "walking back to original spot" to happen again. Might have been a one-off fluke.
"enableAI "FSM"; wold be a way to reduce the "combat dance" for units that are retreating or charging." -yup, disabling the FSM is definitely a way around. It will disable all the formation routines (bounding overwatrch, covering...) in formationFSM and allow the AI to move through urban areas with ease ๐
Mh, one of the AI (group leader) does in fact always walk back to his spot, no matter where I move him. He doesn't even do pathfinding or collision checks, just walks on a straight line through walls/etc. Mh, he still appears to be doing some pathfinding since when I move him far enough away it's no longer a straight line
Ok... so I place down a fire team, put them in combat, disable FSM, make them run somewhere, enable FSM again?
What location?
Stratis, NE side of Agia Marina. Stdby for gridref
030060 , the road going east-west
I spawned 1x CSAT fireteam (or section?) and then made them walk towards the western end. Had a FIA group on the other side of the road. Firefight breaks out, I disableAI "FSM" on allUnits and order CSAT to run away. I enableAI "FSM" on allUnits and suddenly everybody is stuck.
Since I didn't go into this with the intent of reproducing the bug (I didn't know it existed before I started googeling) I'm not 100% sure if I've missed a detail or two
hm...I got them stuck right after disabling the FSM ๐ฎ
You were disabling the FSM after the group was already in combat behavior, right?
On every unit of that squad?
I can confirm that getting the achievement is still impossible, task always getting failed.
If a ticket is needed, here is one: https://feedback.bistudio.com/T170768
you're getting both their id's, not names.
This can actually happen if the owner is being transferred
looking at code, it should work.
There is only one place that starts recoil effect. And that one takes the muzzle coef into account.
If you can attach a packed pbo to the FT ticket, that i can test with, then I can test it.
I won't spend the time to pack my own mod
Possible, but low priority. And adding new shaders is quite annoying and I don't like it.
Okay I can try in the next 2 daysish, anything in particular u want in the packed pbo?
just simple example?
Afaik Polpox was also still waiting on some pp effect, that is stuck for the same reason
Actually I found polpox's ticket, we already have a internal one for that. I added the FT ticket to that one. So if I do any of them, I'll do all of them together
Please add if you find time ๐ ๐ ๐ ๐
Please link the feedback link here.
And, if not done yet, please attach a mission (.zip or .pbo) to the ticket that can be used to reproduce this issue, and make sure all the info you wrote here is also in the ticket (broken since october)
but maybe you could make new sqf file execution when client enters the lobby. like onLobbyOpen.sqf which would be called when lobby opens
A player only downloads the mission when they leave the lobby to go into the mission.
When you enter the lobby, mission scripts are not available. But mod scripts could be.
yes UI/Mod scripts can run in lobby
description.ext is parsed in the lobby, before the main download, right? Maybe it could have a string entry for code to execute when that's done.
I asked for this feature so that side selection could be limited in lobby according to the mode the game is running. like if resistance isnt used the res button could be hidden in the lobby
https://feedback.bistudio.com/T188597
I don't see a share button so I'll guess with this link, there I have left a mission which gives me error and also a screenshot of the error.
If you need anything you can talk to me privately without any problem, I hope you can fix it. ๐
I'd really love for https://feedback.bistudio.com/T177832 to get looked at again, its a pretty big deal in multiplayer and I think it can be fixed with little blood spilled
Fixed soon, thank you
I'd assume if it sees something is wrong, it holds off for a bit assuming its a temporary outage?
Thank you ๐ซ
No. Make your own function or macro.
We could add a function, but its unlikely to happen. Definitely not a script command for that
If it doesn't, then its a deeper engine issue with all line intersects.
The script command seems to parse it correctly and pass it along.
If you make a repro I could take a closer look at what the exact issue is. For now I'm assuming user error
Indeed, that feature only comes after 2.18. There is even specific version number check in code
My changelog message even says so
It was enabled on profiling and dev branch since august, to see if it causes any problems. It didn't so far. So it'll be in 2.20
Was listed as 2.18 on Wiki; fixed now.
I actually believed that was pushed to the stable so... yeah indeed
I will take a look again
I saw Sa-Matra replied, what he said is probably correct
btw is there an reason why lightpoints which are created via "#lightpoint" createvehiclelocal [0,0,0]; instead of creating an light point via config vehicle class lags , it feels like it only getting updated every second frame
It was in changelog 2.18, incorrectly
I realized the same issue there, seems simulation step is too slow
for static lights its completely fine, but as soon as you attach it to an moveable object its horrible
The update frequency of objects, depends on their simulation type.
For lightpoint, it is 15 times per second.
15 per second is actually the default for anything that doesn't set a custom one. Like #smoke has 5 per second
If I create an object via a CfgVehicles class that contains a light point, it will update more frequently than 15 times per second. So is there a way to set a custom update rate for the light point while still being able to use light commands on it? Its not really important to me since I can just create light points via configs, but Iโm just curious
Yeah the vehicle will have a different simulation entry then.
No there's not a way to set it.
I guess it would be useful to be able to set a vehicles simulation rate ๐ค
Hmm. Being able to lower the simulation rate of parked cars might be a much less janky solution than disabling it entirely.
Is there a way to view old feedback tracker issues? Where the links ended in view.php?id=22223...
To expand on why I'm asking : What I'm looking for is why the Huron Ammo Container doesn't allow players to open it's inventory. It was mentioned in SPOTREP 38 but the link no longer works, so I can't find out the details...
Try
https://feedback.bistudio.com/Txxx
Where xxx is the id
Yeah, that's the first thing I tried... Didn't work. Tried adding 0 in front of it too, to make it the same amount of digits. No dice.
Yeah I just tried those myself ๐
The current feedback tracker milestones only goes back to 1.58 too... And this was from 1.38 changelog
Well when I search for Huron I see tickets as far as 2014
Not sure when 1.38 was released
There is a ticket about no water interaction when the slingloaded ammo container touches water
I doubt that's related to inventory tho ๐
There's another one about AI not being able to sling load them. Maybe that's more related 
Yeah, it's an odd one. Checking with maxLoad it has a capacity of 2000. And you can add items into it via script, editor or zeus. But there is no way to actually open it.
There was a similar issue with the AL-6 cargo case props (has cargo space and config'd contents, but can't be accessed) and I believe it was fixed in a recent update. So the Huron container may be fixable as well.
Since it was mentioned in a changelog it seems like it's a conscious choice by devs, rather than a "bug"... Though it seems rather inconsistent since the smaller vehicle ammo boxes has accessible inventory.
Box_NATO_AmmoVeh_F has a maxLoad of 2000 and
VirtualReammoBox_F even has infinite space (1e+09).
Anyway, I feel like I'm getting off topic for this channel so I'll end it there.
Looking forward to. Adding some niche effects to make scene dramatic is very interesting to me (yes, me specifically)
Oh don't take my words seriously, but hear me out. I really just thought of a fisheye ppeffect
Was there a bug with cargo inventories not being cleaned up when the container was deleted? I thought I remember seeing something along those lines being fixed in 1.52 or 1.54.
Using logEntities though I saw something like 1000 varying "Suppy" containers at [0,0,0]. All with simulation enabled and when I tried to get its position with objectFromNetID it returned [0,0,-300]
Hello, I have uploaded a PBO to the issue (I can uplaod here if neeeded too). Sorry for delay was busy with work ๐
anything else I should do?
seems to be working.
recoil is 1.0, then the muzzlecoef is applied and its 0.25..
Ah found it
For players it was ignored, not for AI
Fixed! Thanks 
It seems some Ai related FPS issues were being looked at recently but this ticket detailing a significant FPS reduction with Ai units on a player's group based on their Behavior hasn't been looked at -----> https://feedback.bistudio.com/T167185
Here is a video with repo too: https://www.youtube.com/watch?v=IXDaHrGk2ak&t=96s
no thank you ๐ซก
now to decide if suppressors should cause more or less recoil in my unit 
So muzzlecoef actually works now?
I had actually written a script to simulate it using the playerRecoilCoef commands
Ah, with 2.20 it will reach stable
Idk if thats easy to add, but it would be kinda dope to be able to give the player an blurry distance effect. there is alrdy one for the camera which you can adjust, so it might be easy to allow it for player pov as well
by any chance is this something live on the development branch?
I fiexed it one day ago, no
Workthought: setUnitPos [boolean, boolean, boolean] where you can set AI's preferred state [up, middle, down] so like my usecase be likesqf unit setUnitPos [true, true, false]; // aggressive soldier unit setUnitPos [false, true, true]; // sneaky bastard
Allowed states rather than preferred state?
[false, false, false] - the unit folds into a black hole
a way to disable prone would be handy. so they can stand/crouch but not prone, if set in script
@exotic hearth I'm continuing here because its offtopic for scripting channel, ref #arma3_scripting message and #arma3_scripting message
Thanku yu.
Btw your mission triggers a warning, you cannot spawn AI units as local only objects.
I can instantly reproduce.
The issue is that you call findDisplay too early.
The display is only created, when it is first rendered, which only happens if its visible on screen.
I can see you call findDisplay, while no displays are registered at all. And it fails ofc.
And shortly after that, the display with exactly that name that failed, is being created.
The first 3 computers work just fine. The last two, don't update the percentage screen, but the other two screens are fine?
And I can see there are 12 displays registered.
console 0,1,2,3,4,5's Bar displays are present
and 0,1,2,3,4,5's Count displays are present.
For some reason, the 4 and 5 ones don't update their bar display, but I can see that they are registered, and in debug console findDisplay finds it properly.
I can also see you are calling displayUpdate on the 4 and 5 displays.
BUT, only the 3 bar is rendered. The 4 and 5 bar displays are not rendering..
console_5Bar says, the texture was never visible on screen, even though I'm looking right at it.
Same for 4Bar.
1Bar wasn't visible for quite a while, but I can see that it was visible once before.
So the engine just thinks the 4/5's are not visible, and that's why its not rendering them.
The engine does get told, but while looking up the console_5Bar texture, it instead finds the console_5Count texture, and set's that as visible on screen.
It finds the wrong texture!
That will also be the issue they had with heli displays! @tropic jewel
My sorting code for the displays is just messed up! if 4 values are in a certain order, this can happen.
I mixed up a && where a || should've been ๐คฆโโ๏ธ
Thank you! Finally we got this beast of a bug!
Thanks for finding it! What AI unit is created locally? Unless the desks (Land_PortableDesk_01_black_F) count as AI?
I don't know I didn't check that ๐คฃ
Also, to clarify, the am I updating the displays too soon / before it's rendered or was that just a wrong turn in the process of trying to find it?
there is multiple
Only thing local is the desk, so I guess they're AI??
Wrong turn, or your fault.. Both?
You said findDisplay doesn't find the display, I couldn't reproduce that, it always found them for me, when it was supposed to find them.
It only did not find them, when the display was not created yet. Maybe thats the "doesn't find it" that you saw and that you thought was the issue
I was getting display much later, tens of seconds to minutes later
Probably something I'm doing, thanks
fix is on prof tomorrow
similar to new syntax for "lineIntersectsSurfaces" to allow multiple, possible for something similar with drawLine3D? much of the use-case for drawLine3D is for drawing many lines, like a bounding box (12 calls). would there be any gain to allowing an alt syntax for multiple concurrent lines?
drawLine3D [element1, element2, ...]
afaik it won't make much difference
No. Because drawLine3D doesn't do any drawing
It just pushes lines into a queue to be rendered some time later.
That "some time later" could get MT though, but don't think its worth it
Mh someone else apparently thought it is worth it
optimization would be nice
drawIcon3D is such a slow command , maybe drawLine3D too
drawIcon3D is slow?
But it doesn't do anything, it parses your parameter, and pushes it into a queue
Ah well during the parameter parsing, it loads the texture that you asked for, maybe that is slow sometimes. But theres no fixing that
It probably is, but just in case: the texture is cached, right? So if it's slow it's only slow once, and not every frame?
yeah, but finding the texture in cache, don't know how long that takes, but not that long
What I noticed is that drawPolygon is much faster when drawing multiple lines at once compared to drawing each line with drawLine.
The only problem I had was that drawPolygon always connects the last point to the first one, closing the path.
I was wondering if we could have a parameter to tell drawPolygon to treat the arguments as an open path.
Wouldn't be much of a polygon then would it? ๐
polygon't
poly gone
yeah I noticed some performance drops when using drawIcon3D or was that drawIcon on the map? ๐ค anyway maybe the command isnt slow it self but the rendering that it causes probably is
MapDrawLine != drawLine3D
map draw line/polygon/icon actually does drawing immediately, and could be multithreaded
Interesting
wait, there is multithreading in Arma 3ย ?
Only since a few profiling builds ago or so I heard ๐
@wind light Good morning, any thoughts/progress on these issues? been a long time people been requesting this function -> http://feedback.arma3.com/view.php?id=12782
Related -> http://feedback.arma3.com/view.php?id=18174
Hey @deft marsh i'm no longer taking care of feedback tracker. I will forward it to @trim acorn
thanks for bringing those issues to our attention.
ahh ok thanks for the tip ๐
I could add more thought on those issues but its already covered in FT, and you guys can see it will be a super useful feature
Hey @deft marsh , I have checked the ticket and we will review your suggestion on the closest occasion
thanks!
chance of a sister command to "fileExists", something like "folderFiles <folder>" which can read the mission file only? I am thinking of it for image files so if I create a mod which can put images into a menu, that users wouldnt have to edit code, they could just drop files into the folder and have them appear in script... maybe if security is an issue, just dont allow sqf or hpp?
for "_i" from 0 to 99 do
{
fileExists format ["image_%1", _i];
}
```eventually ๐
lol
It's bit weird it does not support absolute paths in mission dir.
As you should be able to loadFile getMissionPath "' != "" anyways
http://community.bistudio.com/wiki/addonFiles
A mission pbo is also an "addon" Though finding its prefix would be a bit... complicated ๐
honestly idk why fileExists is a security risk. I made my own extension to get better file exists which I need with my tools ๐
extensions are a security risk! ๐คฃ
@thick hare Any news on AI getting stuck due to disableAI? ๐
Updates:
- I reported it to nvidia but haven't received a response so far that confirms the issue or that they've registered it as a bug
- Variable Refresh Rate / Gsync doesn't work either in borderless windowed mode
- Both work fine in Reforger
Not sure if it's related to the issues above, but I also found 100% reproducible steps for a bug that incorrectly sets the refreshrate for monitors above 60hz https://feedback.bistudio.com/T189233
Thought: alternative side that is not civilian after dead
doesnt sound too backwards compatible but could have few more sides for CTI mission, where they battle each other ๐
That is why I say alternative
you mean side command?
@untold sky were rabbits able to open doors too?
#community_wiki message
https://feedback.bistudio.com/T155081
it turns out a ticket has already been created for this problem, I'll remind you about it once again
https://feedback.bistudio.com/T175464
skiplobby it's also broken
it won't cost you anything to fix it ...
"broken" You lie, its intended purpose works.
"it won't cost you anything to fix it" You lie, it costs the company money to have us fix bugs.
I can make a video that skiplobby doesn't work
Yes
ah thanks, I was not crazy ๐
so maybe the setting should be agents or environment agents to not open doors perhapsโฆ?
No, I have already explained it when added, it doesnโt stop agents from entering, so would you prefer rabbit to magically appear through closed door or open it first?
I am not sure to understand
I mean about the https://community.bistudio.com/wiki/setMissionOptions page, SnakesCanOpenDoors is kinda """inaccurate""" (not only snakes) ^^
I would rather have agents not enter ๐
Yes only snakes
All snakes
Previous explanation near this message: #perf_prof_branch message
At that time KK said he would add a config thing for it, so the behaviour could be changed for different types of animals by mods, rather than specifically hardcoding for snakes in the engine. If that config thing was added, it might be worth reflecting that in the wiki page.
maybe the problem is that the slots are virtual
16:56:07 Unrecognized PlayerRole target side: 7, index: 3
16:56:07 Unrecognized PlayerRole target side: 7, index: 4
16:56:07 Unrecognized PlayerRole target side: 7, index: 5
who else can confirm that the fall animation makes fences break?
my goal is to help, I don't understand your hate
Just the fall animation, or fall animation + high velocity?
I know that high velocity does damage to the environment
I don't know, I know it's mostly related to takistan-type maps.
isn't it ragdolls that do it
which is typical of physx
where the lightest object can apply infinite force
It's definitely ragdolls (source: Ragdoll-on-Command user). It's not just fences and walls either; a fun way to tell when you've killed someone is seeing their ragdoll knock over the bush they were hiding behind.
I imagine it depends on the object in question and how much damage it's configured to resist. "Takistan-type maps" probably frequently share a particular fence object that's not very sturdy.
the point is that on takistan maps you will probably have to restore fences every 5 minutes - because it has no synchronization - and you are simply killed through the fence that you have and the bot does not have it****
If the damage state isn't synced then that's a whole different problem
what's the difference - it's all the essence of the same problem
it breaks the gameplay and the game
Can you really not see how "ragdolls knock down walls" and "the damage state of walls isn't synchronised" are different problems?
They might interact to cause trouble for the player, but they are clearly not the same issue. Fixing one doesn't fix the other.
Also, are you sure the problem is that the damage state isn't synced? It could just be that the AI decides it knows where you are, and clips its barrel through the wall and shoots you. They do that with perfectly intact walls sometimes.
no, some players don't have a fence, but you have this fence
If the damage state isn't syncing for walls, you should make an FT ticket with details (exact objects in question, video if possible), because that's a much more important problem than just ragdolls knocking over walls.
It's not funny - houses are being shot at, I even made a video about it - no one cares
I'm lying
everything is perfect for everyone
I didn't say it was funny, I said make an FT ticket with details.
there are no complaints against you - the developer - you make such a face - there is no problem, but the game is broken
No. You can simply stand on an object and do player setVelocity [0,0,-100] (disable player damage)
If it doesn't work, then its not the ticket you linked. Because the ticket you linked doesn't have anything "broken"
I see ragdolls killing fences alot, I've never seen fences being desynced in MP, even on takistan
Never saw this happening tbh, sounds like it's an issue for certain maps only?
There were common destruction-desync issues previously but I think those were fixed in 2.18?
Not sure if they applied to fences either.
Did you for real, just copy paste the text of the original ticket, that I told you is not a bug, into a new duplicated ticket?
Is this actually about not seeing the download progress (as you have just written in the ticket title and description) or is it about something else??
I added a critically important argument - the virtual viewer slot, with it skiplobby does not work
you pointed this out in your remark
But your ticket is NOT about no visible download progress.
Because for that its irrelevant if you have a spectator slot or not.
on all other slots - skiplobby works, I checked everything yesterday
If you want to report a bug, then please explain it properly.
Here I'll do it for you:
"When building a mission that contains ONLY virtual spectator slots, skipLobby does not work and players get stuck in loading screen, and an error is printed to the server log"
Here, that is what you're trying to report.
Instead of writing that, you wrote "No visible Mission Download Progress"
I copied the work ticket because I don't know English - nothing more
Theres many free translators.
Feel free to edit your ticket to have correct title and description, otherwise I'll just close it as duplicate, because right now with its description, that's what it is
ok, next time I'll do that
I wanted to be more understood
Hard to understand the problem if you don't explain what the actual problem is
"skipLobby is broken" is not an explanation, when skipLobby itself, works perfectly fine except one special case
I wrote an RPT log - you can see the reason there
Virtual Zeus slots had the same issue.
Because we didn't consider that we have slots of side "logic", that code is still from Arma 2 or so when there were only 4 playable sides.
Fixed next prof build
Thank you!
Its probably still not good.
The code distributes players along all the available sides in the mission.
But, if there is one player in Blufor and one in opfor, you don't want to put the third player into a virtual spectator slot
I'll make it only use virtual slots, if the 4 main sides are full
I didn't really understand the problem... I create a virtual slot, then I create a unit with a script and transfer a player to it, already with the side I need
addMissionEventHandler ["PlayerConnected", {
params ["_id", "_uid", "_name", "_jip", "_owner", "_idstr"];
[_owner,_uid] spawn {
params ["_ow","_uid"];
private _var = missionProfileNamespace getVariable "TAG_MyMissionSidesave";
if (count (_var select {(_x select 0) == _uid}) > 0) then
{
_sd = ((_var select {(_x select 0) == (_uid)}) select 0) select 1;
_grp = createGroup (_sd call BIS_fnc_sideType);
_type = call {
if (_sd == 0) exitwith {"CUP_O_RUS_M_Soldier_VKPO_Desert"};
if (_sd == 1) exitwith {"rhsusf_army_ocp_rifleman_m16"};
"C_man_p_beggar_F"
};
_ap = _grp createUnit [_type, markerpos ffa_markspawn, [], 0, "NONE"];
[_ap] joinSilent _grp;
waituntil {!isnull _ap};
[_ap] remoteExecCall ["selectPlayer",_ow];
virtual slot
https://imgur.com/xLlnEgt
Was the Linux missionProfileNamespace not writing bug found & fixed yet?
Afaik no
civilian's have a terrible color in the statistics table
for comparison with virtual
It's civilian's color, not independent's.
I don't have it on my radar, did it have a ticket?
It did, main issue is that it is a bit erratic.
This fix, for muzzlecoef for players
Is it on perf/prof branch or only on dev branch?
Also, if it is in perf/prof, do you have the build number of the version
I'm trying to set up preprocessor commands for the sqf workaround I had written, but I don't want the sqf to apply when the fix is present in the game and duplicate the recoil change effect
Nvm, found the changelogs
Was marked as fixed for 2.18.152659 new PROFILING branch with PERFORMANCE binaries, v23
So my question is now:
Do the build numbers always increase for later versions of prof branch?
No, but they're supposed to :P
there is such a problem: you are flying in a helicopter, they shoot at you, you jump out of the helicopter with a parachute and the helicopter is considered -1 plane and -1 car
the problem is that it is not clear how to detect this and return the minuses in the statistics...
This is old bug/feature.
I don't mind - the point is how to track it and return the statistics to the player - constant whining)
build numbers always increase
there is a key factor - you must be shot at from the ground - infantry or vehicle - there is a problem
if a missile flew at you or a helicopter or plane shot, then there is no problem
CH-67 Huron rotors seems work incorrect, if 1st rotor destroyed then heli engines will off as if both rotors dead although the 2nd one is fine. Seems due incorrect config hitpoint value for 2nd rotor, ticket https://feedback.bistudio.com/T189559
Will it be possible to teach the AI to change magazines to attack its target? For example - Offroad (SPG-9), when the gunner selects an infantry as a target, it will aim at him, but will not shoot until the HE shells are loaded (via script by author or AT shells out)
If he got it target is man then why doesn't he switch to anti-infantry magazine?
I don't think there's any good/fast way to do that.
You could poll getAttackTarget if you weren't concerned about performance.
not yet, we were discussing the functionality and as it is now the FSM is disabled, not reset. So an AI in a cover doesn't know how to get out of the cover, because cover system is governed by the FSM.
But there should be cover component inbounce, so it should be possible to disable cover via the same method as FSM and disabling them both wont make the AI get stuck in a cover.
Some textureSources classes have an incorrectly written factions[]={}; parameter in the configuration, ticket https://feedback.bistudio.com/T186947
because of this textures in the virtual garage is not correct shown (for example, civilian textures will be available to independent)
I wonder, is it easier to make faction[]={}; in the game engine to be the analogue of factions[]={}; than to add 75 letter ยซsยป to the config?
i requested an event handler for โAttackTargetChangedโ or similar but request denied
Hello ๐
About a year ago I inquired as to whether someone might be able to take a look at how the horizon object on maps is influenced by the sky lighting.
I know in the ticket it was discussed that it could be a result of the way the shader is set up, but I figured it couldn't hurt to ask again if someone might have a moment to look. I'm sure there are many people who would be thrilled to have the scenic panoramas return to some of the classic maps.
@thick hare Thanks for the info! It's always great to learn more about the AI ๐ Really looking forward to whatever you come up with ๐
flyInHeight and flyInHeightASL are ignored and they are respecting the plane's initial (spawn) Z altitude, not the given number
["Arma 3","Arma3",219,152675,"Development",true,"Windows","x64","development"]
Or... can't tell, it's just weird to me
Are they supposed to perform this? This doesn't seem to make sense to me
one takes precedence over the other, the highest wins
And... none of numbers I gave is 100 or 110
have you tried applying it to the pilot?
I can't imagine this is intended behavior but wanted to drop here to confirm before opening a ticket on feedback tracker for you guys ๐
Essentially a players currentWeapon will be "" for a good time after they land in a parachute. This behavior was exhibited by another player on stable and reproduced by me on the most recent profiling build. Going to attach a video and some weird RPT messages I got earlier when testing with a plane in the event they are correlated.
10:23:44 Array tex in bin\config.bin/CfgVehicles/Plane_Civil_01_base_F/Damage/ not even
10:23:44 C_Plane_Civil_01_racing_F: mfd_center_needle_oil_psi - unknown animation source hitengine
10:23:44 C_Plane_Civil_01_racing_F: hull_lights_coll_on_white_1_blinking - unknown animation source collisionlightwhite1t_source
10:23:44 C_Plane_Civil_01_racing_F: hull_lights_coll_on_white_2_blinking - unknown animation source collisionlightwhite2r_source
10:23:44 In Vehicle: a3\air_f_exp\plane_civil_01\plane_civil_01_basic_f.p3d missing cargo 0 get in direction point
10:23:44 In Vehicle: a3\air_f_exp\plane_civil_01\plane_civil_01_basic_f.p3d missing cargo 0 get in direction point
10:23:44 a3\air_f\data\plane_flag_medium_inv_f.p3d: No geometry and no visual shape while trying to check property cratercolor
10:25:05 Duplicate weapon Throw detected for C_man_polo_2_F
10:25:05 Duplicate weapon Put detected for C_man_polo_2_F
uh, what's up with the update rate
Haha no idea
Something might just be overloaded.
Yeah I was testing with a dedicated server and two loopback clients so wouldn't be shocked if there was something there contributing to the update rate. A player in our live server also had this issue though which caused them to put in a bug report with us so definitely a bug. This hasn't been reproduced outside of our mission though so that's the only thing I would say super worthwhile taking with a grain of salt I guess.
Something happening slowly in Arma is not necessarily a bug. The guaranteed queue can easily have a bubble that long with a bit of packet loss, for example.
I can confirm that when getting kicked out of a parachute, the currentWeapon will ALWAYS return an empty string on a non-local player immediately after until the non-local player changes weapons/firemode manually or after a seemingly random amount of time.
(_unit weaponsInfo [""]) select {_x select 1} will also return an empty string, where its supposed to return the weapon that is currently selected
Just tested on a blank mission file on dedicated server with 2 separate clients; Tested both current profiling and stable branch.
To fix it via script, can just watch if a player gets out of a parachute, then cycle firemode, which iirc forces a weaponState sync cross network.
If the firemode cycle updates immediately then the parachute case probably is a bug.
Added: Ability to 'force load' Eden Editor scenarios with missing addons (use carefully at your own risk) - FT-T157981 https://feedback.bistudio.com/T157981
Can same be added for this message, campaign menu for example?
Sound shaders wrongly was added into sound sets list for 8 classes https://feedback.bistudio.com/T189943
11:11:10 Trying to modify add-only entry bin\config.bin/CfgFirstAid.Update entry:HealSpeedSoldier
Getting a lot of these lately in RPT. Is that a new message?
Showerthoughts:
Improve Launcher's Mod search/sort functionality
- by name
- by author
- by type (eg show only terrain Mods, sort by Workshop tags)
- by collection
- by recently updated
- by publish date
I might write a ticket later
Also plox this:
https://feedback.bistudio.com/T175589
Fix mod initialisation taking ages and preventing any player input.
In Launcher?
Yes
Fired event handler reporting wrong projectile just bit me in the ass, I totally forgot about it. I wonder if this will ever get fixed.
wdym wrong projectile?
Previous shot entity
I remember encountering this bug before, but it happened again and broke my script quite a bit
Is shotCM global?
Here is how I reproed it on a live server: Cancel manual fire, fire shotBullet, then fire shotCM, remote clients will have shotBullet projectile for shotCM fire
// Fired EH
params ["_unit", "_weapon", "_muzzle", "_mode", "_ammo", "_magazine", "_projectile", "_gunner"];
diag_log [_ammo, typeOf _projectile]; // ["SmokeLauncherAmmo","B_127x99_Ball_Tracer_Yellow"]
Ah, that makes some sense for a bug.
shotBullet is kinda designed for sending less data over the network. I guess they missed a check.
if(_ammo != typeOf _projectile) exitWith {};
```fix for now
Workthought and sanity check: there is no setter for map angle righto?
Nope, hardcoded for GPS only
No
Yeah we know. Fixed that error not being reported, but there is/was bug in vanilla data.
Someone ages ago apparently intended to put that config elsewhere, but never noticed it never worked.
Only MP?
Can you give me a repro I can use, actual vehicles to use and what to press?
Only MP, seen on remote vehicle so its a complicated setup
Also needs latency involved
Even with latency its somewhat random
My guess is that some variable doesn't clear when there is nothing to set so it keeps feeding old value into EH
Wish there was a way to introduce artificial latency in the client through a cheat or something
I've been testing Arma networking with this thing, works pretty well:
https://github.com/jagt/clumsy
Works fine on localhost. Better than the docs say it does :P
(local client -> local DS)
Thanks, gonna check it out!
I'd take a guess that shotBullet creations are non-guaranteed but potentially reliant on guaranteed data.
In that case a bit of packet loss would probably introduce some screwups.
Okay, narrowed it all down, its very easily reproducible, no complicated setup needed
What interesting is that it is indeed random with latency involved yet I just managed to have it happen constantly in a closed environment ๐ค
Repro done: https://feedback.bistudio.com/T190141
lol why am I surprised that it's the smoke launcher
those things always gave me such jank vibes
Wow
That could cause impressive issues in some scripts
Its any global projectile, I originally seen it happen with helicopters firing missiles in no-firing zone where they're deleted
The key to this bug is deleting global projectile by firer right away, otherwise its fine
By global projectile you mean non-shotBullet?
I guess so
Not sure if any others work the same way.
Having a wiki page with projectile network behaviour would've been useful
They are jank, creating local smoke grenades on each client is the worst idea ever
Wasn't shotCM local some time ago? I remember dealing with vanilla smoke script not producing anything if camera is too far from remote vehicle and then you move camera close and it doesn't have any smoke coverage
Or is it global yet only reports Fired event close to the camera?
/*ArmA 2 smokscreen, by Maddmatt
Uses code from VBS2 Smoke launcher by Philipp Pilhofer (raedor) & Andrew Barron
*/

its the fired event yeah
they're scripted but not via fired event it's the config based 'fired' event
Armed variants of MRAP Ifrits have lost their armored glass Fire Geometry
https://feedback.bistudio.com/T190093
Speaking of repro, I'd love to open a ticket on assignTeam unassigning (or assigning them back to MAIN really) group members for seemingly no reason, but I'm not really sure how to begin reproducing it
Anecdotally it seems to happen most when lots of players are using it all at the same time, and then just randomly throughout a session to random sub-groups of a group
Any ideas?
I have never seen that happen (except for respawn or server leave/rejoin where I expect that)
Are you sure its not some mod or script doing it?
When the smoke launcher shot message arrives, the message passes a null projectile.
Usually, the message passes the projectile the weapon shot, and then stores it as "last shot" coming from the weapon.
Code later on then grabs the last shot projectile, and passes that to the EH.
So the bug is, when shooting smoke launcher, the projectile reference is not transmitted...
It not being transmitted for shotBullet is normal, becase the receiver of the message, will fire and spawn their own bullet, and store that.
It looks like for countermeasures, that is not happening. Countermeasures are global projectiles. Not local per client.
And the sender forgot to transmit that in the message
The messages are also arriving in correct order, first the CM is spawned, then the fired eh is triggered.
It happens across both our vanilla/I&A (through the command menu) and private zeus server (via ACE) so I'm hesistant to say a mod causes it
I guess null entity being recieved from deleted CM doesn't replace the last shot?
yeah.
But there is something very weird going on.
The sender does send the projectile.
The receiver first receives and creates the projectile, then receives the weapon fired message that contains the projectile.
But then when it looks up projectile by entity ID, it doesn't find it, and it becomes null.
If client sends projectile that is set to deletion, maybe projectile deletion arrives before Fired?
Create projectile -> Delete projectile -> Fired (no projectile found by ID)
Anyway, just trying to highlight that the key to the bug is instant projectile deletion
๐ค Maybe
Still I wonder if this can also happen in non-scripted scenario where projectile creates, hits something and deletes on the same frame
So fixing this could also fix some random jank somewhere
But if scripts struggle with wrong projectile, wouldn't they also struggle with getting a unexpected null projectile?
I can detect if the sender sent us a projectile, but we failed to find it, and in that case do store the null
Storing null sounds right, the firing event happened but the entity is no longer there
As firing authority instantly deleted it
Oh no, I just had one case where the fire message arrived before the countermeasure shot 
But in that case we'd also make it null

Bah indeed as you said.
The create message, then the delete, then the weapon fired.
๐ค
What actually deletes the projectile? is it script? init script?
Bah I see how that happens.
So its not lag or packet loss.
newProjectile = {
Object created - Network message sent to everyone about it
Object init script runs
Script deletes object - Network message sent to everyone about it, object marked for deletion next frame
return object
};
send network message informing everyone that weapon just shot newProjectile
The network messages, all arrive in correct order there. But the receiver immediately unregisters the network ID, even though the projectile does still exist, until next frame, until after the fired network message arrived.
I guess fired message gets processed next frame or something?
Still, I think reporting null projectile in Fired event handler in this case is correct behaviour
๐ซ
dedmen (2)
the cloning process finally succeeded
guess that being able to disassemble nearby drones when controlling another one isn't a desired behavior?
(to clarify, if you're controlling a drone and are close enough to another one that can be disassembled, the action menu option pops up, so your little drone can disassemble the other drone with its little... rotors, I guess?)
has this happened to anyone else in stable?
Ever. Seen once in about a month
ok wasnt first time to me neither but it happens only sometimes
A square cloud casts its shadow on everyone
it is known, but ugh.
dont have engine actions have conditions like radio commands? sth like (1 - inVehicle)
With the new change to currentZeroing to make it return laser zeroing, just wanted to flag the opposite issue as well - setWeaponZeroing not overriding laser zeroing (https://feedback.bistudio.com/T162056)
keep breaking Dedmen's motivation like that and there will be no more patches ever ๐ฟ
I didn't mean to pile on more work, but more that the work on the currentZeroing fix might shed some light on this and make it easier to fix, since it involves the same parts of the same system
I am only joking, genuinely zero problem (on my end ๐) about it ๐
https://feedback.bistudio.com/T67521
Fixing this bug years back made large smoke particles fairly ineffective and unfair. I wonder if this change can be reviewed to make it all better
IIRC the issue was that all particles were cut off if they were closer than SCALE meters to the camera
So tiny casings had 1 scale and didn't display 1 meters in front of the camera (invisible casings in first person)
I don't know/remember how this was fixed but this made large particles disappear way too easily if they're close to the camera or zoomed in too much
Result is large smoke particles are very ineffective in creating visual cover, you just have to look at the smoke with binoculars/optics and you'll right through them
Also if you stay inside such smoke source, can easily see everything around you, but unless zoomed nobody can see you inside, which makes no sense and is unfair
Visualized:
- Thick smoke covering the vehicle 100%
- Same thick smoke being 50% transparent because half of cloudlets disappear due to zoom
Can something be done about it? I don't know the current formula but if it hides particles by model size X scale formula, it far from optimal for large particles
boundingBox of "\A3\data_f\ParticleEffects\Universal\Universal.p3d" = [[-0.5,-2.8871e-08,-0.5],[0.5,2.8871e-08,0.5],0.707107]

@BIMINIONS Any progress on this issue? http://feedback.arma3.com/view.php?id=24971
heh, FT is being passed around within BI like a hot potato
@waxen tundra, maybe razazel should be made @BIMINION here, in this case
at any rate, it's not so useful on continuity
stuff that @icy turret had assigned in the past was relicted (probably because it was not best practice to create an internal ticket), hope you got this covered better now
right, @wind light ? ๐
https://community.bistudio.com/wiki/say
https://community.bistudio.com/wiki/say2D
maxTitlesDistance seems ignored, sound can be heard by any player all over the world
https://community.bistudio.com/wiki/say seems ignore for Spledid Camera or Zeus, for player ok
Pretty sure I already noticed that and noted it. But I'm not going to fix it soon.
It is the exact same issue there too
@untold sky May I ask you what's current particles cull distance to the camera?
Yes. No
@untold sky Finally reproduced the bug with RE calls being dropped after mission restart: https://feedback.bistudio.com/T190364
Repro mission, server setup provided
Also has two more bugs reproducible: being stuck in lobby unable to select any role and having debriefing screen after new mission already started
Versions dont matter, been an issue for years
Couldn't reproduce setTerrainHeight data not arriving, but at least its something
It happened under stand setup on real servers, could be some other trigger somewhere
Kinda similar issue I've seen is remoteExecs from the previous mission session arriving at the server and being processed after the restart.
Just told admins never to use restarts in the end.
Seen this happen too, not that often though
"being stuck in lobby unable to select any role" That happens when the server is busy, for example loading the mission file or terrain, or frozen on some heavy scripts. In that case that is not a bug
Did it get you unstuck with that repro?
For me you were in that stuck state forever, you can only leave the server
Were you able to try having the admin force drag the stuck player to a slot? There's been a long-running occasional issue where one player might be unable to select a slot themselves (before mission launch) but the admin can force slot them. I dunno if that's the same thing
Nope, stuck completely, nothing you can do but leave the server
Admin commands don't do anything either
https://feedback.bistudio.com/T184143
can we get this nice addition ๐
so we dont have to use per frame event handlers for airburst munitions ๐
Hello @untold sky I finally had some time to test this out https://feedback.bistudio.com/T167366
sorry for the late response I was just busy and then kept forgetting ๐ฅน
I made a suppressor, change some of the coefs around. dispersionCoef, recoilCoef, MuzzleCoef initSpeed, and AmmoCoef typicalSpeed work fine. However in AmmoCoef the hit and airFriction and recoilProneCoef dont seem to be effected.
maybe I could have formatted this better lol
Just a quick question does this happen to you guys as well:
https://feedback.bistudio.com/T190081
Actually. I recall Ded has some Launcher feature updates, but it is holding due to the localization is not yet done. Or maybe I misrecalled some, but still localization is the biggest issue to add one feature is the point Ded ever made.
The question is, is the holdout worth of anything? If we really have to, we can have some volunteered translators? Like me ever done to the base game (accidentally)?
That looks like the bug I already fixed on profiling branch.
Where server is waiting for all players to get mission file, but if someone didn't ready up, they never get the file and it gets stuck.
I think server indeed might not allow slotting when in that state.
If thats not it, then something else froze the server there ๐คท
Probably yes, its on todo, but 2.22 :i
But, I fixed it, so is it okey now or is it still broken? There is nothing new on the ticket so going by the ticket i expect that is closed and done
understandable ๐ซก
oh I can put on ticket, I didnt know if it was best to bring it up here first or on ticket.
Original output filename: Arma3RetailProfile_DX11_x64
Exe timestamp: 2025/03/20 13:23:39
Current time: 2025/03/27 23:25:43
Type: Public
Build: Profile
Version: 2.18.152746
Tried with this version for server and client
All repro'ed bugs are there
Server didn't freeze btw in that case, something broke on client side, other players should be fine

Haven't seen this one before
How to repro, only seems to happen on latest perf build:
- Same dedi + client setup
- Select "debriefing bug"
- Wait until you see a seagull in the background
- Join and see your copy
So yeah its just bugged in a different way in perf
Confimed that server isn't stuck in that second bug, only you as client is stuck
Commands don't work, lobby doesn't work, all you can do is leave
So lotsa stuff to fix 
Okay I have responded on the ticket
@trim acorn since the team is looking into improving shaders with TANOA, i guess it's the right time for this? http://feedback.arma3.com/view.php?id=27319
Is there a reason why getItemCargo returns empty array, but everyContainer, magazinesAmmoCargoand weaponsItemsCargo returns filled arrays when used on an exploded vehicle?
I found an old ticket https://feedback.bistudio.com/T146514 where KK said it won't get fixed. Is this still the case in 2025?
Apologies but any update on this #arma3_feedback_tracker message? Although I haven't been able to confirm this behavior anywhere else, I feel like I have been seeing an increase of issues where it may be related. We rely on currentWeapon a lot for our handleDamage event handler 
Has there been any progress on perhaps adding the ability to get (and maybe even set) the timer bool and timer duration for explosives?
#arma3_feedback_tracker message
This would be really useful as it would allow less convoluted canceling of bomb timers. Furthermore, being able to get the value of the bomb timer with a command, be it new or existing such as through a new return value from getShotInfo (or make return value #2 timeToExplosion return the bomb timer (if set) for applicable explosives), would also be a lot more convenient than what's possible at the moment.
Hm. Could you elaborate? I'm unable to get either a bool value of whether or not a timer has been triggered, nor an integer of the time remaining.
getShotInfo on an explosive charge before setting a timer:
After setting a timer:
Huh? I thought it would be timeToExplosion
Yes, me too. Although this doesn't work on explosives that are placed on the ground by the player
It's only working for grenades (including smoke grenades etc. where timeToExplosion will be the time until the smoke effect is triggered)
At least this is the result of my preliminary testing
yeah its kinda annoying theres no way to see if an explosive has been 'armed' with a timer like that
no 'sane' way anyway, you can probably do it with the useraction or w/e theyre called eventhandlers
Add action
Unfortunately, arming/setting timers on explosives isn't a userAction, so you can't detect it by EH. UserAction EHs don't return any details about the action anyway, just whether it was started or stopped or its analogue value.
yeah not useraction I meant:
https://community.bistudio.com/wiki/inGameUISetEventHandler
and like I said it isn't 'sane' its plain awful
I recently came across your content and thought our latest article on MDMA-assisted therapy might be a great fit for your audience. It explores the fascinating history of MDMAโfrom its early days as a therapeutic tool to its recent resurgence in treating PTSD, anxiety, and emotional trauma.
Spam bots seem to know how much PTSD we get from dealing with Arma bullshit over the years
removed, but still :kek:
At the moment, only dead units have sounds of falling from a height, alive will be completle soundless. Is it possible to improve the sounds of falling for alive units? Maybe connect them with gertting height damage, or with the animation of falling?
Means sounds from this config: CfgVehicles โ SoundEnvironExt โ surfaceType โ "bodyfall"
Can you make a ticket with the files attached, I don't know what you mean by selecting debriefing bug
ticket to add that to getShotInfo
Its the same ticket, just behaved a bit differently on prof build
"Debriefing bug" is an action in repro mission
Question, i have a problem and i have no idea where to address it:
Whenever i alt tab from arma, i cannot click on anything on the screen, and i need to CTRL+ALT+DEL then cancel it, in order to be able to click properly.
Where can i get help about this? (Please ping me so i see it asap)
I think that's been confirmed. You can also see this in the video, no?
#arma3_questions Try using full screen window, changing resolution and other related settings.
?
So far zero success on that front.
Was on borderless, change it to window, nothing. Fullscreen, nothing.
Resolution was always native, changed, seemed to not care.
The worst thing is that i remember having that issue before and fixing it, but for the life of me dont remember how.
In the configuration for 32 surfaces of Tanoะฐ character movement sounds (sit down, lie down, prone, stancesโฆ 70 psc more) were not added
https://feedback.bistudio.com/T190719
just want to ask again regarding this "issue" and if its worth to open a ticket so it gets changed
reposting will not get you very far dear sir ๐
Depends where they want to go ๐
~~spam https://feedback.bistudio.com/T190896 ~~
I wish controlling a squad leader with AI subordinates through Zeus let you use the squad command interface
I actually really want this to make a Delta-D system for tanks.
in case there is something attached to a rope the ropeUnwound command will incorrectly return true too early when the rope isnt properly reeled in (ropeUnwind). will this thing be looked if I create a ticket?
most likely
Context: #arma3_editor message
- Put an empty HEMTT Transport
- Put a Rifle Squad
- Put an unit who is not a leader driver
- Put the entire group
- Put a waypoint somewhere
- Play
- The group issues a move order, but HEMTT doesn't move
- If the player is a group member, get out, wait until Get In order, get in, the HEMTT finally starts to move
["Arma 3","Arma3",218,152405,"Stable",false,"Windows","x64"]
wish we had command similar to compileScript but it would take string as argument. because the compile command cant handle comments and probably not preprocessor either EDIT: Meant comments
Is the GUI control page just... destroyed for anyone else?
https://community.bistudio.com/wiki/Arma:_GUI_Configuration
https://i.imgur.com/tzL4Bhi.png
Not for me
I'd guess your internet or browser is causing it to default to minimal mode
Looks kinda like the raw html pages that sometimes load before all the css and scripts load
(There is also the #community_wiki channel, for wiki stuff)
Try Ctrl + Shift + R.
I had tried that, along with 2 different browsers (Firefox and Chrome) on 2 different computers and asked a few other people if they had the same issue, which they did.
Cant check right now, but if it's working for everyone else, just ignore me lmao
I had the same issue a moment ago but it fixed itself after reloading.
Just wanted to add I was also having this issue with command pages as well yesterday. Pages were also really slow to load.
Me too on some other BI command reference pages yesterday.
Is it by design that only the side chat channel show message in quotes?
So this issue seems to happen every time when I enter the biki without getting verified by cloudflare first. @gray wharf not sure if IT is aware or can do something about it.
I was about to ask you to try fully clearing tbe browser file cache (files only, not cookies or anything else), but I suspect it is something entirely different - I will report, thanks!
I tried all that and it also happens on different devices. All using Edge
Hey guys!
I just tried join a dedicated server on the most recent profiling build and got a bunch of script errors on what I am presuming to be every frame. I was messing around with launcher parameters earlier and somehow BattleEye did get toggled off so I imagine that is likely the reason for this and it is not a large issue. I tried to see if I could find the actual origin myself but the only file I was able to find was functions_f\Feedback\fn_incapacitatedEffect.sqf which seemed to be entirely commented out. I've got all of those PBO's unpacked on a drive that is kind of slow so perhaps I just hadn't actually indexed everything.
12:17:57 Error in expression <er != "DEAD") && (abs((damage player) - BIS_oldDMG) > 0.0001) && BIS_applyPP3 &&>
12:17:57 Error position: <BIS_oldDMG) > 0.0001) && BIS_applyPP3 &&>
12:17:57 Error Undefined variable in expression: bis_olddmg
12:17:57 Error in expression <((BIS_oldLifeState == "HEALTHY") || (BIS_o>
12:17:57 Error position: <BIS_oldLifeState == "HEALTHY") || (BIS_o>
12:17:57 Error Undefined variable in expression: bis_oldlifestate
12:17:57 Error in expression <able ["BIS_fatMod_EHAdded", false]) && !BIS_fnc_feedback_fatiguePP && ((getFatig>
12:17:57 Error position: <BIS_fnc_feedback_fatiguePP && ((getFatig>
12:17:57 Error Undefined variable in expression: bis_fnc_feedback_fatiguepp
12:17:57 Error in expression <
(GetOxygenRemaining player >= 0.8) && (BIS_oldOxygen < 0.8) && BIS_fnc_feedback>
12:17:57 Error position: <BIS_oldOxygen < 0.8) && BIS_fnc_feedback>
12:17:57 Error Undefined variable in expression: bis_oldoxygen
12:17:57 Error in expression <BIS_PP_burning && !(isBurning player)>
12:17:57 Error position: <BIS_PP_burning && !(isBurning player)>
12:17:57 Error Undefined variable in expression: bis_pp_burning
Anyways again likely a very isolated non-issue but figured I'd still pass it on though just in-case ๐
Ahh okay yep VSC wasn't indexing everything. Appears to be sourced from functions_f\Feedback\fn_feedbackMain.fsm 
BIS_PP_burning
You should probably get that checked out 
Just got this error after loading into the Paradise Found mission of The East Wind campaign. Stable, no mods
Would it be possible to expand BIS_fnc_drawBoundingBox with a new parameter for line thickness which was introduced in 2.18?
It'd be reaaally reaaally nice to have Arma to create an automated backup of your profile vars file. Another power outage. Another long SP mission progress lost... (and all the addon settings I forgor to back up myself)... fml
If you have issues with power outages you might wanna invest in an USP
while I appreciate the suggestion, that doesn't solve the problem of the vars file getting nuked after Arma closing unexpectedly
Yeah, normal games make a backup of their data if it's prone to getting corrupted. Which Arma's profile certainly is.
Never has happend to me. ๐คทโโ๏ธ
Never happened to me either, but I've had to deal with an awful lot of players who got their profile wiped.
Combination of it being quite a large file and frequently written, I guess.
Would be nice if missionProfileNamespace worked at least, because that seems a bit less fragile.
Well, while I might have given the impression that power outages are a normal occurrence in my particular case, it is not. But after all these decades of Armaing it's happened enough for me to notice that it resets the vars file. This time it was me trying some of my stuff while some quite big power outage going on (maybe you've read the news). Incidentally, I had a game of "Deadzone: Vietnam that I would totally finish this time"... so, yeah
ctrl type CT_OBJECT (80) CT_OBJECT_ZOOM (81) | CT_OBJECT_CONTAINER (82) in dialog make spam "No skeleton for this shape" in rtp while control alive....
<- From models in cfgWeapons cfgvehicles cfgAmmo.
Also game crash if call
_control ctrlSetModel ".p3d"
idk, shutting down arma 3 on startup somehow clears rpts and other things, it once reset my profile cfg completely for some bizarre reason
You are no longer the master of your PC. Arma takes away all your user rights. 
?????
welcome to arma ๐
...
bruh
ยฏ_(ใ)_/ยฏ
Hasn't happened to me a whole lot but it has happened a few times over the years. Those times though were with my profile I use specifically for development so it wasn't much of a loss. I've found where it tends to happen is when I'm booting two clients at once. Maybe something to do with writing at the same time.
A tip I got off a friend which I thought was pretty genius was just symlinking the profile folder to a personal GitHub repository. Allows one to have previous versions and backups of their profile with a simple git push ๐
That's what I do. Also have armake2 setup so I can git diff them, not that much useful due to amount of changes usually though.
# .gitconfig
[diff "armaconfig"]
# quite likely you can use "hemtt utils config derapify" instead
textconv = armake2 derapify
# .gitattributes
*.vars.Arma3Profile diff=armaconfig
Still that's pretty sick! Honestly had just never even thought to do something like that. Cheers @uncut briar!
Game crashes when using ctrlSetModel 
not in my mission. maybe you give it a bad model something?
are you on stable ?
yes
will check if it works there
if you are on prof there is the prof channel and you can sen crash dumps to dedmen
funny bug if you create a unit in another group side that was originally in the config for a side that doesn't match the group side.
Side unit != Side group unit
hmm yes , join fixes that
That's why you almost alway wanna check side group unit;
I remember a long time ago. More than a year ago. There was a bug that players could find themselves on the "ITEM" side after respawning. They were not seen by the AI. Not by players via DrawIcon.
The curator's interface showed that the player was a normal item and could be controlled without access to edit players.
this really needs to be fixed/adressed after all the years with 0 reaction...
Tank AI commander with 0 ammo, makes player or AI as gunner reloading 7.62 mm coaxial machinegun after each shot, if/after enemy AI gets spotted
https://feedback.bistudio.com/T154433
List of vehicles with commander turret not returned by AI to direction vehicle/armor is facing/driving, when player switches to another seat (unlike all the other vehicles/armor, where it works that way, as intended)
https://feedback.bistudio.com/T148411
List of helicopters/jets where it's impossible to get out when under specific armor fire, due to get out animation restarting non stop, because of specific damage per second
https://feedback.bistudio.com/T148491
Mi-48 Kajman APDS/HE ammo selection order is inversed when ordering to switch via CTRL+F
https://feedback.bistudio.com/T143981
BTR-K Kamysh APFSDS-T/MP-T ammo selection order is inversed when ordering to switch via CTRL+F
https://feedback.bistudio.com/T143984
FV-720 Mora APFSDS-T/MP-T ammo selection order is inversed when ordering to switch via CTRL+F
https://feedback.bistudio.com/T143989
AMV-7 Marshall APFSDS-T/GPR-T ammo selection order is inversed when ordering to switch via CTRL+F
https://feedback.bistudio.com/T143987
AFV-4 Gorgon APFSDS-T/MP-T ammo selection order is inversed when ordering to switch via CTRL+F
https://feedback.bistudio.com/T143988
AMV-7 Marshall, MSE-3 Marid, AFV-4 Gorgon and Strider wheels spinning in different directions on water, at different speed or not at all and not synchronized in MP
https://feedback.bistudio.com/T143986
BTR-K Kamysh has no possibility for player as commander to cycle through different armament types (Ctrl+F), when turned out, unlike AFV-4 Gorgon with same turret/armament (where it works just fine)
https://feedback.bistudio.com/T154445
Left Mouse Button and Ctrl+LMB don't work to manually fire or order the AI gunner to fire from 120/125 mm cannon, when player is turned out as commander (using Arma 3 Apex keyboard preset) in T-100 Varsuk and Rhino MGS (UP) (unlike on all the other tanks, where it works just fine)
https://feedback.bistudio.com/T154446
More configuration fixes please

https://pastelink.net/cvjhmvof
Pastelink.net - Anonymously publish text with hyperlinks enabled.
Door states of V3S Wreck (class Land_V3S_wreck_F) can not be edited in Eden due missing numberOfDoors=6; parameter in configuration https://feedback.bistudio.com/T191571
I am wondering why that wreck has doors in the first place, you cannot enter it
it walkable inside, player can open door via useraction and go enter it
oh I wasn't aware. ๐
https://feedback.bistudio.com/T191578 Maybe KK or dedmen can have a look at this? Grouping via add3DENConnection doesn't work and get3DENConnections returns garbage for group connections.
@void lava Feel free to add something to the ticket, otherwise yfi
Lol what the...
Arma 2 in headers for "event scripts"...
I cant find this page if seek
Init.sqf arma 3 Init arma 3 Initserver Arma 3 InitplayerLocal Arma 3 onPlayerRespawn.sqf arma 3
But work only event scripts
It took me a long time to remember the name of the title.
#community_wiki is there for BIKI discussion, also you can use BIKI search for more better result
Has any change been made to the tools UI recently?
For some reason if I use BankRev the config.bin no longer gets converted to a text file as well.
The hell is going on with those icons? 
I saw that in earlier versions indeed
A riot!
windows 11 issue, or hidpi, it's fine on 10 and 7 iirc
Windows 10 @ 125% scaling
scaling might be the issue after all - @daring wagon do you happen to be at 150%+?
yes
I am on my TV in the living room, got it cranked up there. Thats why I've never noticed it before.
It would be nice to add native support for rounded corners (like borderRadius) in UI controls such as RscText or RscButton. This would allow us to create more modern-looking interfaces without relying on image-based workarounds 
It's to save us from the horrible ""modern"" rounded UI!
Windows XP/Vista vs Windows 8/10! ๐
it goes by wave, round, then sharp square, then round, etc
Will this happen anytime soon?๐ซฃ
https://feedback.bistudio.com/T79777
Is there any possibility of getting more scripting control of rope parameters? This was originally "resolved" by adding custom rope CfgVehicles classes, but while that's neat, it's still a bit restrictive because you can only do it through a mod. Being able to adjust elasticity and strength dynamically (or at least when creating the rope) would be much more convenient.
Possible alternative: allowing rope classes to be defined in mission description.ext as with sound sources. Still not as flexible as would be ideal, but a bit better.
Current sounds from soundFakeFall parameter cannot been played correctly for tank shells due to long audio samples; although there is possibly a sound in the game that should have been used instead
https://feedback.bistudio.com/T191737
I am sure this has already been reported but the discord overlay can cause Arma 3 to loose focus/pause itself or just hang (doing alt tab or clicking can resume a normal state of the game)
Probably a Discord issue thou
Windows 10 latest Discord app
so you want a "preprocess" command is what you're asking. Unlikely
Ticket please 
I have never seen that, that happens in vanilla too? A mod could be doing that
Can you make a repro mission and a ticket for it? It would take too much time to learn how to set up CT_OBJECT
It happens in vanilla but I think it's only for the sideChat command, not player-submitted messages
I fixed the ctrlSetModel one.
But I cannot reproduce the "No skeleton for this shape" with my test model (soda can from CfgVehicles)
Or make your profile folder a git repository itself.
I do that with my AutoMissionBackup mod, creates a local git repo in profile, and pushes every time a mission is saved in editor
it crashes if the model doesn't exist
I don't think so, we did a .NET update for some tools, but not that one
You could check in Arma 3 tools folder the last changed date, it shouldn't have changed
never did for me, BankRev just unpacks, doesn't convert
if I get a FT ticket with repro, it might be fixed later this year
Ropes are vehicles, which can't just be changed to look up in description.ext.
Script control unlikely. These parameters are pulled directly from the type, for script control we would have to duplicate them and I don't like that.
Maybe we could add a couple rope types to vanilla, but data changes are ugh
๐ discord ๐
yes or alternate syntax to compileScript where it would take string instead of a file
Weird. I got an Excel tool that used bankrev.exe to unpack all pbos in parallel and that does the conversion. Maybe because I start it via command line. Can't check the parameters I pass to it right now.
Anyway. If it has never done it before then there is obviously no issue.
Yeah its because of my new TV which runs as 4k @ 225% scaling. Never had these settings before, that's why I never noticed.
Is the Ref to nonnetwork object on every unit creation known?
Might be specific to running with ACE, but it looks like you can't do basic stuff like damage _unit during entityCreated?
You are right, I played on vanilla KOTH and didn't know that chat can be modified on mission side. Checked on the official warlords and it works without quotes, sorry
Its a fake side channel made out of custom channels
Entity created EH is very early, often before the entity is properly initalized
https://community.bistudio.com/wiki/listRemoteTargets
listRemoteTargets opfor
return targets with -lastDetected
The last detection may be below zero after some time.
And how remove RemoteTargets from center?
ps. Detected by CSAT AR-2
I'm sorry what is the question/issue?
-How to remove RemoteTargets from the center?
-lastDetected: Number The negative number is a bug or a feature.
-lastDetected: Number The negative number is a bug or a feature.
if all numbers are negative, then "-50" most likely means "now - 50 = 50s ago"
if not, Dedmen may know
-How to remove RemoteTargets from the center?
from the center of the screen? use distance from [0.5, 0.5]
otherwise I do not know what you mean -forgetTargetperhaps?
RemoteTargets for sensors on radar
lastDetected = lastDetectedTime - now.
so it should be positive ๐ right
lastDetectedTime = spotted at 1500 seconds into the game
now = 1600 seconds into the game
1500-1600 = -100
its probably not in seconds, just example
I AM SMART I SWEAR
brainfart ๐ but okay yes ๐ thanks
Is in seconds
UAV Ar-2 make 10 sec detect time. After 10 sec next time -1sec.. -2sec.. -3sec...... -infinity sec
Why can't array lose targets on the side (center) if lastDetectTime < 0
๐ค
There is not even a command to delete this target from the sideโs memory
As a result, the array with targets will always exist until the object is deleted. And the time will forever decrease simply because someone did not make the condition for deleting targets that have not been detected for a long time. 
@dusky tendon tried forgetTarget yet?
group forgetTarget target
group: Group or Object - group itself or unit in a group
side reportRemoteTarget [target, time]
listRemoteTargets side
side: Side
How forgetTarget must work.....
allGroups apply {if (_x == DesiredSide) then {_x forgetTarget _target};};
if not working, you probably need to ask for it to be expanded to side forgetTarget _target; (like reveal)
Worth remembering that groups <SIDE> exists
Don't forget about Farming Simulator 26!
Why isn't the word "Gummies" already banned from being used there?
You can probably ban 100 words that these bots always use, for example "Facebook" and clean these spams a lot faster
Like how all of the Youtube spam bots all use the same 5 PFPs, but take weeks to be banned.
google "phabricator" and you will know
How's that replacement project coming along? :U
going on
IT accidentally replaced forums with new feedback tracker and now everything is on fire
and here I thought we had a warm month of May, though that didn't explain the smoke
I hope the new FT system turns out to be real and not a phabrication
Phab is open source and in PHP so quite easy to customize/extend actually.
But yeah, it's not maintained
in PHP
PHP is great for what it does
Sounds like you are volunteered to do it
I know, and I now have access
I need some time to figure out its entrypoint and where goes what however
Happy to help, I have around 6 years of experience with making Symfony web apps.
Pretty sure phab is bit worse/more legacy but nothing too hard for sure.
Famous last words๐
so far I tried editing controllers, entrypoint etc with even a simple ```php
die('Hello there');
IIRC Phabricator recommends OPcache so its prob enabled
that's out of my waters
Has anyone noticed that when another player uses night vision goggles, we will see animation of his goggles too late after few sec? I noticed this on a dedicated server with two of my clients
If use script command there no delay in animation: https://community.bistudio.com/wiki/Arma_3:_Actions#NVGoggles
here is ticket https://feedback.bistudio.com/T191859
The command says when the rope was reeled in, to minimum length.
like 50cm.
Your sample mission, has the role fully wound in, but, the rope is stretched to 10x its length.
The rope is pulled in to 50cm, but its stretched to 5m. And according to its elasticity it slowly pulls together
Network sync is not instant. Not a bug
oh so not a bug? i just thought it was weird that it "works" without anything attached to it (The rope should still be same length)
It works without anything attached, because it doesn't have a weight that adds inertia and makes the rope stretch
didnt notice any stretching but ok
Well its hard to see
ropeLength returning 50cm, while you can clearly see the rope is still >5m long == its stretched
ok
If only there was a way to control rope elasticity to stop it doing that 
And if you turn off the detach.
You can see how after stopping the reeling in, and after the rope length stays constant, the attached dude is still pulled up further
There is, ropeCreate lets you specify the ropeType, and you can give it one thats less elastic
thx for looking into this dedmen, i have to give this thing another go sometime
I was just being jokingly shitty about this, don't worry about it #arma3_feedback_tracker message
It might be nice to be able to see how stretched a rope is.
But its somewhat easily scriptable
ropeEndPosition _rope params ["_rStart", "_rEnd"];
_stretchFactor = (_rStart distance _rEnd) / ropeLength _rope;
you know in real life ropes arent made of rubberband ;D
well depends what rope
imagining guy using rope to get out of heli.. like in the blackhawk down movie. their ropes look solid ๐
hmm no phorge then?
bungee jump maybe
all ropes stretch ๐ even those
Those are heavy ropes so they dont go back to propelers becouse you know it would be a bad thing.
Again it depends what kind of rope you are using and in what situation.
Also with enough force it would stretch as well ๐
i guess
ropeSetDuctility needed
#arma3_feedback_tracker message
they are vehicles defined in config
If you actually log it
Rope starts out as 12m with stretch factor of 1.
At the end its 0.5m with stretch of up to 21
counter-offer: ropeSetConductivity for proper simulation of downed powerlines
springFactor/dampingFactor are adjustible in config, but not sure if that's enough to deal with this
Edited: This issue does not appear to be happening, but the original issue in the FT remains the same
Idk if it's too spammy to add this to the FT but I just noticed this command is broken in the context menu in 3den too. You can right click a unit, hit Connect -> Group To and the cursor works, but when you click to do the grouping, it's not working either. so its not just an issue encountered in scripting but regular usage
soundBullet[]={}; this is casing drop sounds, game play this sounds randomly if shoot not fast, if fast or burst game will no randomize sounds more until longer pause between shots or burst end no ammo
https://feedback.bistudio.com/T191902
video: https://www.youtube.com/watch?v=iwsG_vfR4uw
Good news, I was able to tweak the Feedback Tracker to support custom spam filters for tickets (comment filters are not supported for now). We should see less of them, and if it happens we can always add more custom-tailored filters. Cheers!
I would like to be able to set different fog values โโwhen exec on a specific client. Can you add a boolean to setFog to disable sync with server value?
https://community.bistudio.com/wiki/setFog
I fear this is intertwined with weather, rain and eventually day timeโฆ ideally a setDayTimeWeatherSync false to allow everything to desync, but I don't know how together it is
new option in setMissionOptions?
I would love to!
Could alternatively set their view distance down
in the last 32 hours, only one spam ticket (made of one line of text and one link) made it through 
Has anyone had issues with clients getting stuck on receiving data after role selection? We've had the issue for years and always assumed it to be an Arma issue but never really thought to reach out about it to confirm. If this is something else other people are experiencing I can try to get a memory dump of this next time it happens if that would be helpful ๐
I believe this is due to missing mods or something alike, try searching in #arma3_troubleshooting you should find quite a lot of matches
Cheers thank you. Very interesting- we do allow a few different client side mods but they aren't required for players to join. I do feel like I recall experiencing this without having any kind of mods loaded although I honestly can't say for certain as it has been months since I've ran without them.
Gonna drop some of my findings here too just in-case anyone wishes to reference:
#arma3_troubleshooting message
#arma3_troubleshooting message
#arma3_troubleshooting message
I'll try running without them for a bit in the meantime to see if I can reproduce but you're probably right about it being a mods thing.
@ripe crystal you could check rpt log for potentially more info. in addition you could try perf branch (compatible with main branch) to see if that helps
it would be nice to have constant variables similar to what compileFinal does now. or would that downgrade the performance?
There's a fairly common server state where it does this to all new clients and you can't solve it without restarting. But if it's only some clients then the issue is something else.
What's the point?
well you know, if you have variables that must not be accidentally changed
How do you accidentally change a variable?
forEach missionNamespace ๐
Then compileFinal
bad code/not paying attention while writing it ig
but
#define __SETCONST__(var1,var2) var1 = compileFinal (if (var2 isEqualType "") then {var2} else {str(var2)})
__SETCONST__(variableName,value);
like in if condition for example if(constVar = 0) then {}; not that i know if that even compiles ๐
doesnt that require call to use the variable?
yes
ah, well then not quite what i wanted ๐
Just name them _doNotChangeVariableName
and if you don't trust yourself to not overwrite it just have a loop that runs once every minute or so checking that ur const is still the right value, if not? yell in rpt to check ur code (or for anticheat reasons)
I would've assumed it was more about not trusting other people not to overwrite it
#arma3_feedback_tracker message
well he gave this example
I wonder. What if you use a hashmap and compile that final. You can then store all your stuff in an still have fast access time, no?
constant access time, but still gonna be slower than just a var no? so kinda depends on the situation ig
that would work, techically i guess
Sure. But there's going to be some slow down either way
Hash Map is actually directly returned by compileFinal, so no calling needed
For the record calling a compileFinal'd variable is negligible for performance, is there a specific reason you want that exact format? -
specifically
0.00002 ms increase with compileFinal'd hashmap vs normal hashmap
0.00012 increase with call over no call
Even with the command as you're asking, it'll still have some drawback, no matter how minute
those are just some syntax suggestions
so far we had three spam tickets (all in the same format and not in English, so another filter is definitely doable!) and one spam comment (I don't filter). I'd say it works pretty wellโฆ for now
(only) one spam a day keeps the burnout away ๐
So all it took was
if (_arrBadWords findIf {_x in _text} > -1) then ban;
Not sure if I should laugh, cry or just be happy.
IF (Facebook Link) then Ban; ๐
the hardest part was writing an SQF parser in PHP, otherwise it was fine
Regarding Z-fighting
https://cdn.discordapp.com/attachments/1045849933858734080/1373789825534005289/image.png?ex=682bb0fa&is=682a5f7a&hm=a6d9b26ce0f3fe3c9a9c545f8e55781a9cf3315d84daa7e2937e1db49a21a685&
https://cdn.discordapp.com/attachments/1045849933858734080/1373789892403789864/image.png?ex=682bb10a&is=682a5f8a&hm=5088acedea8192798215c032fe062b01c150efa4c25aec747ee46d5a9c154f1a&
Two pictures, same position, the one with less Z-fighting is with the unit standing upright.
For some reason lying down causes this severe Z-fighting.
Maybe something that can be tweaked.
Not necessarily. With a standard Z buffer the spread far from the camera is strongly dependent on the distance of the near plane. When you're lying down the near plane naturally needs to be closer.
I would generally go with "don't use a standard Z buffer" but that might take some work.
Back when I was writing this shit, I think it worked out as a one-liner in the vertex shader, but that was a while back :P
please add such notes to the ticket too otherwise I'll forget when I get to it
you can compileFinal a hashmap, to have a entire namespace of constant stuff
yeah we discussed that already
doesnt that mean you have to use get though?
compileFinal'ed hashmap has no difference to a normal hashmap, thats just measurement error
Send me RPT please
yes. The most constant way would be to use macros, but that only works if you know the value at build time
and defines needs to be included everywhere
-cpuCount=12 -exThreads=7 -enableHT -malloc=jemalloc_bi_x64 -maxmem=13024
remove these parameters. Though probably not your issue
12:19:13 Inventory item with given name: [hgun_P07_F] not found
Something is f'ed
verify your game files in steam.
Something is wrong with battleye, and it also seems like something is wrong with your game files, it shouldn't say all these base game weapons are not found ๐ค
I'd say reinstall your BattlEye, but I don't know how to do that
helloh question before I make a ticket, at this stage of development what are the chances of changes to the way the arma 3 server mission cycle works?
Because I've been running into an infinite loading screen recently when a mission cycle is defined but the actual mission path is empty, the server keeps trying to load the mission forever until you pull it out of it using #missions command
Is there any chance of a check for that being added and the server just being sent to the normal mission select screen rather than loading forever
Now you could say 'just delete the empty mission cycle that bugs it out and that is valid, but no
(aka I dont want to break other peoples workflow with my own because they keep using the mission cycle for some reason even when we're just playing custom missions on our own server)
I'd say chances are good that I rip out that whole thing and rewrite it.
When that's gonna be, uff
But that means theres a chance 
there is no technical reason against fixing it
not enough Dedmens is a technical reason
What is the legal/ethical stance/situation regarding cloning Dedmens?
done alreโ we don't support such unethical behaviour
We tried before and it didnt work out well
the second one came out fine and was sent into the code mines but it broke after the 3rd one
I told you, it worked perfectly
Feature request - http://feedback.arma3.com/view.php?id=27392 <-- surprised this isnt already implemented by now ๐
Is it intentional that killed event handlers (including EntityKilled) don't work on a building that was rebuilt with setDamage 0?
yeah usually when you do stuff like that if you try alive on it it will return false
last I checked the dead state is irreversible
I'm surprised you could even "rebuild" a building (except for anims, e.g windows)
With buildings it appears to work fine. Building pops back above ground in pristine condition. Takes damage as expected. Doors open. alive returns true. Only the event handlers don't work.
BuildingChanged EH fires correctly in both directions.
It simplifies accounting to the degree that I'll probably still use it and just workaround the one bug we had from expecting the killed EH to work.
im here
Hello
@fluid swift @round spade hiyo
what are the encountered issues please?
I open the thread that I'm involved in, I type out the reply and press submit, it goes gray for a ms and then back to blue like it refuses to send the reply
so when i try to create any bug report i get to the page that says "get lost, spam"
i cant send comments to any tickets, i press the button to send the comment but it doesnt do anything
and all my profile feed got deleted
My profile is still there though.
Recent activity and all is there
I was about to say yeah, nothing is lost
Also usually when trying to submit a reply to a thread you get a message preview
There's no message preview as I type it out
that is (quite an aggressive) antispam I did setup until the FT is replaced
can you DM me title + report please?
same
some characters / words are now banned that's why
I want to keep spam out but good users in obviously ๐
of course
many links? I restricted it to 5 for now
I can't think of any words in my reply that could be in the "banned list"
Did you mean title and report for the bug report or info to unban me?
even I cannot edit https://feedback.bistudio.com/T192058, so I assume I did a whoopsie just today, gimme 5
ok, thank you!)
I see, thanks for the clarification
@fluid swift @round spade should be good - I disabled a filter while I figure out what I did wrong
It is working again, thank you for help!
Thank you very much
Filter fixed, genuine thanks for the report gents! ๐ป
Just wondering... Can AI plant mines in the game without mods? I found that the FIA โโresistance engineer dropped 2 anti-personnel mines when he went into INCAPACITATED state. I don't have a single line of code in my mission where it would be written about mines for AI. I didn't even step on these mines, to my surprise.
Maybe after the fall the engineer chose "Put" weapon and fired them?
AI can't normally go into incapacitated so I don't know. Normally, they never place mines without player commands using the 6 menu.
mod
incapacitated AI + AI-planted mines
(otherwise I believe it can but only on player order)
as you see. It also happens that mines fall out of the backpack of bots without any mods. 
I was very surprised when I noticed 2 red triangles near me, especially when I had put the soldier down literally 5-10 seconds ago. 
Fall out of???
nope
How much did AI buy driver's licenses for? 
''fall out of''
I think the AI โโjust uses Put weapons. While Down.
In FSM/OrAiBrain when he aimed at me and started to execute the code "shoot at target" then at that moment I wounded him. And instead of Action "fire" for the main weapon, he executed it for the Put weapon.
I'll repeat. I don't have any AI mods.
What do you mean while down? There is no down for AI.
Sure there is. It's just not used in vanilla Arma. setUnconscious has plenty of defined behaviour though.
We've done plenty of setUnconscious on units with mines in their backpack though, and I don't recall them ever placing them.
What I'm saying is that some script is doing stuff and it's the script's problem.
shrugs
Can't really prove anything with Arma AI.
I gave up trying to replicate bugs with it, because all the other bugs get in the way and you can't isolate one of them reliably.
AI bugs probably won't get fixed anyway because it will break campaign mission somewhere or something
Could maybe add a setting to switch between fixed and old version?
Or detect if campaign and fallback etc.
setMissionOptions could be a great fit for it
inb4 some mod does it game-wide and stuff explodes
Just got an error pop-up due to:
Warning Message: Config: array does not have 2 entries
No additional context as to what array and where. (for clarity, in A3)
This is slightly vague considering how many arrays tend to be in configs so would it possible for some dev to go through these error conditions and give more details in rpt, at least when running the game with -debug ๐ค
May have probably something to do with something regarding cfgammo and sounds, mayhaps
getting a lot of log spam btw "No skeleton for this shape". it may have something to do with CT_OBJECT
just stable
We need to manually add context in alot of places where its missing, that is probably one.
I need to reproduce the problem to track it down
is there still a chance for Compartment5 to be added at some point? (see your posts below that its used also by A3 itself)
I don't remember that. Why can't you add compartment 5?
the value is not defined as macro in the engine. as such you cant set in configs/it drops at least a warning in rpt:
Warning: Cannot evaluate 'Compartment5'
I'd assume it uses then the default value
if the reply context doesnt work
@desert trench check dms pls ๐ฅบ
They are not macros, neither are they defined in engine. They are enums in config. You can put your own enum in config (Don't know if you can also patch existing enums).
Also they are just numbers, nothing prevents you from using the number directly, instead of the enum name
You could also make your own macro
You can also have a seat be in multiple compartments by combining the numbers. But I guess you do that with the enums by adding them together
compartment5 is used in a3 configs already but causes the error
due to not being defined as an enum
ok so binarize or the engine on parsing uses/applies the enums from bin\config.cpp. bit of a shame the unrapped config.cpp is no longer part of the BI tools to learn about the default enums
yes, yes
I don't know which config that is in ๐ค Mh actually it should be dta/bin.pbo
Which means I can change it
We even use Compartment 9
I'll add them up to 9, but you can use the rest by just using numbers yourself ๐ซ
I don't want to add more enums than needed
lol config can't handle trailing comma in arrays.
But in enum's its specifically handled
fix arrays too 
Huh I'm confused, the Compartment enum's are not in bin.pbo
if they aren't there, where are they then ๐ค
dta?
dta/bin.pbo yes, no
They are included in the cpp that gets packed into bin.pbo
But they are not in the config.bin ๐
armagic at it again?
Ah huh. enum's aren't even packed into binarized configs, they must be resolved before they get binarized.
But then why do we have strings with Compartment in our configs 
wouldn't that mean that they are defined in the engine code then?
Ah enum's ARE packed into the config, but not as enums.
They are serialized as script variables, at the end of the config.bin
rollercoastering
There they are, at the end of the config.
But why are the Compartment ones not there 
it sure sounds fun
anyway please never stop logging your findings and progress, I believe we all love to read that :)
(not sarcasm! ๐)
I missed the word "never" and was appalled
no I always loved these behind the scene moments, I could never diss that!
There are 3 enums.
Destruct types (DestructWreck is in the config.bin)
Stabilized types (StabilizedInAxesXYZ is in the bin)
Vehicle compartments (Compartment1 is there .. Ah I'm stupid
Roll back everyone, all is fine
Yea, please keep sharing these insights. I always love reading them.
so what was the end result for compartments then?
this
This was due to a silly mistake I made in a soundcurve array under CfgSoundCurves.
So something like
class CfgSoundCurves {
class WHATEVER_rangeCurve300 {
points[] = {
{0, 0},
{50, 0},
{100, 0.7},
{200, },
{300, 0}
};
};
};
ah, missed that
Cargo rope break and in-air detach parameters refers to wrong values. Mistake in all 10 vanilla classes: https://feedback.bistudio.com/T192204
spam accounts playing the long game
Do the accounts look like they were hacked and at one point legitimate or do they are just created then go dormant? Just wondering if 2FA could help you guys with that if not already required
we are getting rid of this Feedback Tracker soonโข to replace it by a newer, sexier and fancier one
is it still based on phabricator? if not, will the current tickets be transferred?
there will be an attempt ๐
I hope that the links won't be broken (e.g from changelogs) 
Hey people, I've recently encountered some weird behavior with script-created Zeus modules while locally testing rc compatibility of the Admin/Development-Tool I'm writing at the moment, and wanted to ask whether this might be known or I might have made some sort of mistake. From what I've searched online / on the feedback tracker I couldn't find much, although I might have just filtered incorrectly.
The Issue is the following: In SP / local MP environment script-created Zeus modules are incapable of remote controlling AI units and can't open the "edit window" when double clicking units. Editor placed Zeus modules work fine. Unexpected/unintended behavior comes along when having RCd a unit using an editor placed module previously. Subsequently using RC with a script-created module, it will default to the last RCd unit (if null / dead it displays the appropriate error message in the zeus menu - regardless of which unit you try to RC and their current status).
However: This is not the case in a dedicated MP environment. Script-created modules work perfectly fine and can use RC to control units just fine. This only happens in SP/local MP.
If someone is able to verify / this isn't already reported I'll make a ticket for it but wanted to make sure and ask first. I've made a quick video trying to showcase this (I recommend watching at 1.5x speed). First part is showing it working perfectly fine on dedi, second part is showing the behavior in sp/local mp. (Video: https://www.youtube.com/watch?v=ubV4AvNd9Tg)
For reference: while testing this, I was using this code to create modules:
[[_randomVarStr], {
_logicGrp = createGroup sideLogic;
_module = _logicGrp createUnit ["ModuleCurator_F", [0, 0, 0], [], 0, "NONE"];
{ _module setCuratorCoef [_x, 1]; } forEach ["Place","Edit","Delete","Destroy","Group","Synchronize"];
_module addCuratorEditableObjects[(allMissionObjects "All"), true];
[_module, (_this#0)] remoteExec ["setVehicleVarName", 0];
}] remoteExec ["call", 2];
I have a problem with HC. I don't know why, but HC has a RAM leak. In 5 h, the HC application took up as much as 3 GB. In 10 minutes, the value increased by 200 MB. The increase occurs every second. I have never seen a decrease.
I don't even know what to do with this...
Params.txt
-client
-malloc=tbbmalloc_x64
-mod="@extDB3;!mods\CBA_A3;!mods\CUP Units;!mods\CUP Weapons;"
-enableHT
-high
-limitFPS=700
-skipIntro
-noSplash
-hugePages
-world=empty
-connect= it doesn't matter to you
rpt
PhysMem: 16 GiB,
VirtMem : 131072 GiB,
AvailPhys : 14 GiB,
AvailVirt : 131068 GiB,
AvailPage : 36 GiB,
PageSize : 4.0 KiB/2.0 MiB/HasLockMemory,
CPUCount : 4
3GB for a headless client is normal usage though?
The more bits of map or objects you load, the more memory it'll use. Until it hits the limit, when it'll throw some currently-unused stuff out.
Our server stops working after 10 hours due to exceeding the RAM limit. You can't even connect in to the server in the game. Or connect to a remote desktop. The only option left is to restart the PC system itself.
What RAM limit?
RAM PC 16GB... Hc client app fill 16GB+++++... Windows systems crash with RAM out limit.
If you have 16GB physical then Arma processes should have a hard limit at 12GB, IIRC. You can set it lower if you like (and I would).
4GB limit is normally fine for CUP.
Shouldn't crash Windows even if it did somehow use 16GB (you have virtual memory), although it might make it somewhat unresponsive.
a3 generally doesn't respect the maxmem very well especially in client
memory in windows works differently than what you might expect, virtual memory when you let the page file automatically increase by system management becomes infinite and there's hardly any oom crashes
don't run a server with pagefile off
As if the fact that HC has a memory leak does not bother anyone at all. Let it fill all the memory from all over the world with incomprehensible information.
Restarted the server 3 hours ago. At the start, the consumption was 4GB 35MB.
After 3 hours, the memory consumption became 10GB.
After 1-2 hours, the server will stop responding to any actions because HC will want to write more and more to the RAM. But there will be no more space there.
I took memory measurements over a period of three hours.
98% usage...
12GB RAM eat HC...
In Paint screenshot in 4h 28min
Now time 8h 55min
After 30 min server say "bye bye i die"
@dusky tendon what game mode, or custom scenario? also what (major) mods in use?
Mods #arma3_feedback_tracker message
Game Mode - clear and capture zone\town
custom scenario - full custom
Mod functions no used in server (only server mods)
Mods on HC and D Server
There are no scripts that would work for HC (Only connected)
@dusky tendon Thanks for the info! One last question - if you dont run HC, the DS doesnt eat all available RAM (and crash eventually)?
And is this 2.20 RC, 2.18 stable, or perf branch?
== C:\Users\Server\Desktop\Sirius\Server\arma3server_x64.exe
== "..\arma3server_x64.exe" -par="C:\Users\Server\Desktop\Sirius\Server\SERVER\start_hc_main.txt" -exThreads=3
Original output filename: Arma3Retail_Server_x64
Exe timestamp: 2024/11/13 09:31:44
Current time: 2025/05/31 03:42:32
Type: Public
Build: Stable
Version: 2.18.152405
Allocator: C:\Users\Server\Desktop\Sirius\Server\Dll\tbb4malloc_bi_x64.dll [2017.0.0.0] [2017.0.0.0]
PhysMem: 16 GiB, VirtMem : 131072 GiB, AvailPhys : 13 GiB, AvailVirt : 131068 GiB, AvailPage : 36 GiB, PageSize : 4.0 KiB/2.0 MiB/HasLockMemory, CPUCount : 4
Forgot...
if you dont run HC, the DS doesnt eat all available RAM (and crash eventually)?
No such thing. DS RAM is within acceptable limits
There are no scripts that would work for HC (Only connected)
The HC isn't actually doing anything?
I'm running a test but apparently after 20min the HC decided that it didn't need RAM anymore.
Perf branch though.
VirtMem : 131072 GiB

VDS moment ? 
What's VDS? ๐
virtual dedicated server
class AnimationSources
{
class Door_L
{
source="user";
initPhase=1; //works
};
};
class AnimationSources
{
class Door_L
{
source="door";
initPhase=1; //doesn't work
};
};
Is it intended that in class AnimationSources the initPhase parameter does not work if the source is a door?
afaik yes?
vehicle doors can't be kept open iirc?
animateDoor works on them and they will stay open until something else tells them to close
fullCrew [cursorObject, "driver", true] will return the same thing regardless of whether a vehicle has a driver (config hasDriver=0 for static weapons). Intentional?
this is not right, empty positions "driver" returns 0 so this should return [] as well
video demonstration of https://feedback.bistudio.com/T192324
Please make sure we will still have access, there are many juicy unresolved tickets
for CfgVehicles non looped Sound Sets with loop=0; shaders plays not randomly, game play one sound always as loop=1;
but CfgEnvSpatialSounds with the same settings successfully randomize shaders every time object play sound
Before 2.20 delay for SoundSets in CfgVehicles caused the game to crash, maybe randomization of shaders for non-looped sounds can be done too?
Here is a better version of the carrier spawn script attached to the comment of the ticket:
https://feedback.bistudio.com/T124858
On a related note, which simulation type would be best for walking or driving on top of it?
Would it be better if it was simple/super simple object, or is "house" good enough?
both are good
well no point in further dwelling on that then
I propose an additional parameter called layer (integer) for switchAction, playAction, playActionNow and gestureState. This parameter would allow us to attribute a layer to the gestures. The reason why is because i'm still working on a close combat mod and i would like to apply a hand gestures for the weapon hold and a body gesture for the movement at the same time. I don't want to use defaut pistol slot because i don't want the character to be stuck in a pistol holding position. At this moment we cannot use 2 gestures on a character at the same time. The layer parameter woulkd allow this.
why 2 gestures? can't the "body gesture" be a full body anim?
No, even if i use a mask on each animation, the body animation will reset the hand animation when played
even if it resets it, you can replay it using switchGesture to the last point
then when i use switchgesture for the hand anim, the body animation dissapears as well. We cannot have both animations at the same time
no it won't
switchGesture only affects the gesture
not the anim
you're confusing it with switchAction
ok so what- do you mean ? I should play the hand nimation with switchAction and the body animation with switchGesture ?
i cannot use switchMove for body animation. I need the character to be able to move or run at the same time. It's more like a torso animation. With means both torso and hand animations are gestures. We can't play both at once
I'm unable to replicate your seam fix. The seams still seem to be in the same place. It appears that it might be a selection point problem on the base carrier object itself. This is just me doing it in post init, moving the parts after they were created, until a fix is uploaded to game. this is the seam between hull 8 and hull 9 (both 1 and 2 versions)
does the diag_log log correctly?
as in making sure it actually runs
pretty sure its running as I'm currently messing around with vectorAdd:
if (_dummyObj isKindOf "Land_Carrier_01_hull_09_1_F") then {
_dummyObj setPosWorld (_carrier modelToWorldWorld (_carrier selectionPosition _dummyPosName vectorAdd [0,0.1,0]))
} else {
to move them around in real time. but even with the pieces overlapping, I still fall through
so maybe its not a positioning thing at all
could be related to size I guess
this is with hull 9 moved over 0.1 and overlapping (i'm currently falling lol)
I once had a bug where I would fall thru Land_Wip_F near the edges
but when I moved it it would fix itself
are you spawning the carrier in 3den?
well actually I meant to say spawn it via script instead
I only had this issue with 3den objects
spawn it as a 3den entity or object while in 3den? or do you mean spawn it as an object in preview?
in preview yeah
it appears that when spawning it in preview, then changing base direction, then applying all the rotations for the parts, there is no seam. but this absolutely kills my workflow for modules and others looking to build their own missions with the mod as I can't use the 3den placed object
btw can you try doing it in preInit too?
I wonder if it's a timing issue
i take it back, there is seams
really hard to see in this one
and it does make me fall through
what if you swap the setVecDir and setPos lines?
it shouldn't matter tho since it's world pos
btw is this Stratis?
I wonder if it's the bounding radius issue again 
tho that was affecting animated stuff so maybe not
idk but I'm tempted to put concrete panels just above the carrier surface to prevent players from falling through
I wonder... what about simple/super simple?
nope didn't work
Can't you just overlap them very slightly
tried this with like 0.1m, and i still fell through
if you walk along the seam, do you fall thru everywhere?
curious
increases in chance as you approach perpendicular seam significantly
it is much less since creating it in post rather than 3den
it could be a bounding radius issue. can you try spawning a sphere centered around the part's [0,0,0] and same radius as its roadway LOD bounding radius?
there are helper models for that. Sphere_10cm etc (check the config) (spawn it as simple object and scale it)
btw if you don't rotate the carrier, do you still fall thru?
still working on bounding radius. I don't believe that you fall through at 0,90,180,270,360 but I'll have to test again
the way I see it this issue is due to one of these reasons:
- the quadtree doesn't find the object
- the object is skipped due to radius check
I'm pretty sure devs already added overlap into the roadway LOD so that's highly unlikely to be the issue (and you already confirmed it by overlapping the objects manually)
it's boundingBoxReal [obj, lodRes]#2 btw
this seem about right? multiply radius by 10 for a 10cm sphere for the scale factor?
just multiply by 2
just right about where you consistantly fall through
now try the model bounding radius
you want them as different colors overlapping?
(boundingBoxReal obj#2)

