#arma3_feedback_tracker
1 messages · Page 8 of 1
le bouton?
into the void
Has anyone had a closer look at https://feedback.bistudio.com/T171346? Would like to start taking a closer look at long frames but not having a working diag_captureSlowFrame on servers makes that difficult 😅
addMissionEventHandler's _thisArgs fails to hold an object after save and load the mission, will only return <NULL-OBJECT>. This is annoying...
Group also gets vanished
addMissionEventHandler ["Map",{
systemChat str (_thisArgs)
},[player,group player]] ;```Minimum repro
could you make a ticket and ping me?
Sometimes Uniform and Vest inside a container getting bugged after a restart. (Dosent happen to backpacks)
You cannot take it and if the container is LOCAL to you, then your Uniform/Vest on your character will duplicate. The duplication will drop into the container (Take&Put Eventhandler are not triggering). The duplicated Uniform/Vest has the same netID then the Uniform/vest you have on your character.
The duplicated Vest and Uniform are visible to every player on the server.
It only duplicates when the Container is local to the player.
Some commands I tried to get more information:
Client Sided:
getItemCargo CONTAINER // 2 Classnames getting listed
[["U_FRITH_RUIN_sdr_fabdpm_rs","U_FRITH_RUIN_sdr_ltrdrk"],[1,1]]
everyContainer CONTAINER // 1 Container is getting listed
[["U_FRITH_RUIN_sdr_ltrdrk",21a122d8100# 1067575: dummyweapon.p3d]]
Server Side:
everyContainer (ObjectFromnetromNetId NETID CONTAINER): // 2 Containesr are getting listed
[["U_FRITH_RUIN_sdr_fabdpm_rs",1c9607436f0# 1065412: dummyweapon.p3d],["U_FRITH_RUIN_sdr_ltrdrk",1c9738bb0b0# 1074841: dummyweapon.p3d REMOTE]]
SetOwner Command dosent help.
What is Intresting that there is no REMOTE behind that object on the bugged uniform/vest
Also in the inventory UI the bugged unifrom/vest dosent show his load bar.
I am pretty sure a ticket on the Feedback Tracker already exist, if not, please create one with these details and information
ok I will check
thanks!
would it be possible to add waypoint completion radius outline to 3D view? lines are already rendered both on map and in 3D so unless there is some limitation I can't think of this can be pretty useful
maybe to avoid screen info saturation
it's a design decision / forgotten thing, but why not for selected group(s)
I'd much rather have a native command to visualize areas in the same way Eden Editor does for trigger areas.
Visualizing waypoint areas of all kind (random, completion) is then just one script away.
Now I wonder. How is the Editable area in Zeus visualized?
ah yes, these transparent cylinders, I would love that
similar to this: https://feedback.bistudio.com/T172248 (ticket which I was asked to create)
After more investigation I found out it has something to do with the ownerShip of the vest/uniform. I'm 100% sure.
All the bugged uniform/vest having a ownership of 0 and they are stuck in there.
If a player takes a uniform and drops it back in, then the uniform has the ownership of the player.
But as soon as he disconnects the ownership needs to get back to the server and thats where the problem is getting created. It stays stuck at 0. But if its on 0 no player can interact with it and then you can abuse it duplicate items.
Also JIP dont see the items from uniform/vest which they putted into an container before reloging.
Repo (Multiplayer Enviroment):
1.Container on Server
2.Connect to server
3.Put Unifrom Inside the Container
4.Disconnect
5.Reconnect
6.Check Ownership
Execute on server: (owner (everyContainer (ObjectfromNetID NETID OBJECT) #0#1)) If its 0 , then its bugged.
If step 6 didnt work, repeat from step 4
ticket
Maybe it is related to this: https://feedback.bistudio.com/T126030
yea must be, the problem has something to do with ownership not getting changed back to the server and stuck at 0
id like to suggest getArtilleryComputerSettings return grid or position2D of the targeted position
then make a ticket
Grid is useless in this context definitely
agreed
You can get the grid of a position anyway
https://feedback.bistudio.com/T168411
I mean, please fix this
Ok this is just funny: https://feedback.bistudio.com/T174760
removed 😬
Another spam:https://feedback.bistudio.com/T174723
Any chance of getting a selectionOrientation command? No way to get orientation of selection ingame afaik but have selectionPosition
selectionVectorDirAndUp
WAHT
holy fuck i feel stupid now
will put in #community_wiki for see also of position
Not sure if someone already did a ticket for this here. If you drive over a bridge your 3rd person view gets very close. It sucks sometimes if you just use the 3rd person view to shoot at any target, or even to drive over the bridge. First i was thinking its a bug of my own made vehicles, but vanilla vehicles got the same problem.
I've tested it on Tanoa with vanilla tank, in these pics
I'd say vis LOD is tweaked for attached objects too would be great
Is mdmp wanted for crashes if we have 100% ways to reproduce them?
Negative.
how about the log slow frame command?
Yes. But I would say only if waypoint is selected?
Needs ticket, sounds neat haven't done such stuff in eden yet, a new learning opportunity
Yes. That often saves time.
Drag&drop file to instantly see where problem is, rather than setting up repro, starting game and waiting to crash
Thanks for the spam cleaning 🫂 I've noticed barely any
that one? diag_logSlowFrame
says it's not implemented in Arma 3 builds (https://community.bistudio.com/wiki/diag_logSlowFrame)
Gotchu -- it's not a 100% repro anyway (trying to find one for bluetooth connection bug, had first two attempts crash but none since then
)
most odd bug ever, can only get it to crash reliably if i open video options before audio options..? 
Cannot get it to 100% crash but I have a crap ton (9) of mdmps from this, some of which are freezes (always from hitting refresh button tho regardless of if device connected or not), think it's just random -- should I post all mdmps and just basic steps to reproduce (i.e disconnect and reconnect bluetooth) or is 1 mdmp enough?
In the VR training, for Portable UAV target designation:
- I place down the darter
- connect it
- take control
- rise a bit
- take uav turret control
That last step makes me stuck on a 1st person view of my character holding my gun angled
Err this is the release candidate branch of 2.14
Then, when you open the terminal and click the camera view boxes "control driver" "control gunner" they don't do anything except grey out alternately
AK12 GL Lush variants have a bogus baseweapon value. Is that the sort of thing that's worth reporting or not?
https://feedback.bistudio.com/T174831 Script error in Eden Editor when excuted fn_3DENInterface
This happened to me as well in the UAV showcase. Never had this happen before and I am not using any mod.
https://www.youtube.com/watch?v=flfWP036P6s
Ok, I can reproduce it 100% on vanilla
- Connect to a placed Darter
- Take control of the Darter
- Once you have to control, use the action menu entry "Take Turret Control"
Idk if you're running 2.12 or whatever the current non-RC version is, but I want to add I've played Western Sahara on non-RC 2-3 weeks ago and the drone in Extraction worked fine, even the turret
I am on RC 2.14
Two very stealthy spam comments on this: https://feedback.bistudio.com/T173832
first one is definitely generated by AI but the last one looks like manually written (or copied, but I couldn't find the original post with the file). I think I'm gonna use the bouncing mine troubleshoot procedure whenever someone posts in #arma3_troubleshooting without giving any details
https://feedback.bistudio.com/p/haileyryan101/
Spam account...
Why does Arma freeze 80% of the time when you tab out ?
It's super annoying 
I still wonder if it is a Windows' or Arma issue
there is no progress on this one: http://feedback.arma3.com/view.php?id=16822
Worth a ticket that composition thingy also detects stuff like zip files and adds an entry for it into the tree view?
yes
I completely missed that these were on RC 
I'll make one when I am back home.
might not need ticket, its now already on the list
unless you want to be linked in dev branch changelog
Sharing FT ticket in the hope fix in 2.14
An error in the config that does not allow turning on head's NVG using an M2 .50 iron sight
https://feedback.bistudio.com/T172106#2439788
Please if possible fix this issue too with 2.14 version
https://feedback.bistudio.com/T174906
So, if I do a setPos thingy in Eden Editor for whatever reason, it doesn't update Eden Attribute position (obviously) but once I do Preview, boom the object is now on the previously setPos'd position, and go back into the Editor, is in the position not the attribute position before the preview. Is this a bug that something we would want to see a fix?
- I place a Huron there
- I executed
vehicle player setPosWorld (getPosWorld vehicle player vectorAdd [0,0,50]). Note that the position that is defined in Eden remains the same - After preview, it starts from 50m above than Eden defined position
- And after the preview has ended, you're no longer on the ground
well you should use set3denAttribute
the object you see is just a "dummy" to show the effects of what you have changed. it doesn't get saved to the .sqm file
as mentioned on this page:
https://community.bistudio.com/wiki/Eden_Editor:_Setting_Attributes#Attributes
That is not a case in this point. Sometimes I wanted to have setPos without updating position actually
ah so you mean you want the 3den attribute pos to be re-read in the preview?
Yeah
In the first place, why it does read the actual (object's getPos) position not the attribute's position?
well if you save the mission and reopen it it will be reread 😅
but anyway, dunno 
I have a reason to do this actually. I have an attribute to setObjectScale like in EAS, what if I try that without changing its center position, boom this is the thing
Also speaking of: if we have a setter/getter of Eden rotation function from/to setVectorDirAndUp would be nice
2.14 is in RC, unless its a new issue that appeared in 2.14 it won't make it
Pretty sure that can be scripted?
For the other thing, ticket. But I'd say low chance of fix
Indeed it can be scripted, but official function would be handy, I mean
is there any chance of angular velocity commands this late in armas development
there is angular add force physics thing
um, what's the successor to phabricator?
i know its phorge, but im asking in the scope of BI
not known
meanwhile, phab could be upgraded to phorge which has active development
something's brewing
no getter for that though and is an awkward format to work with
change of vectordirandup per second would be nice 🤷
(that and as far as i can work out theres no engine interpolation currently for vectordirandup and such)
👍
Question about say3D maxDistance param
in the cfgSounds the value of the maximum distance at which the sound is heard is entered, but this parameter is useless if you execute
OBJECT say3D "MY_SOUND"//i.e. w/o params inside script command
because the command will drop it to 100 meters. I think say3D should take maxDistance param from config
For example, scripts from the DLC Jets:
in the** config** devs set the audibility range of the ejection sound to 1400 meters, but in the script they did not doubling the parameters:
_ejectionSeat say3D _ejectionSoundExt;
as result the sound range is reset to the default's 100 meters
How about remove from say3D the auto-reset maxDistance to default(100m) if it is not set when calling the script
There are many examples where say3D used in vanilla scripts without specifying parameters and as a result maxDistance is wrong
And I think that the scripts authors expected the command to use params from config, this seems like a logical expected result, why it overwriting?
@trim acorn i feel your pain :)
@untold sky Here is the repro for the height offset with compositions https://feedback.bistudio.com/T174949 (Needs Spearhead 1944 though)
Video with repro + demo mission + sqe file
Also, to have it all in one place, here's the link to the issue with compositions not showing up in Zeus https://feedback.bistudio.com/T174947
BIS_fnc_customGPS doesn't seem to be working even thou I have the required sqf onMinimapScript[] = {"BIS_fnc_customGPS_Spawn"}; on my descrition.ext file. (https://community.bistudio.com/wiki/BIS_fnc_customGPS)
Or is it that the "ca\missions_pmc\data\ui_gps_ca.paa" file doesn't exists in Arma 3?
Indeed. This is pretty much for A2PMC, and none of such features should carried on into A3 in the first place
Indeed, it has the Arma 3 icon so that is confusing
Script is there after all. There should be A3 icon
It doesn't mean it is a working function
It is just this 😅
/*
Author: Karel Moricky
Description:
Replaces GPS frame.
Following line have to be defined in mission Description.ext:
onMinimapScript[] = {"BIS_fnc_customGPS_Spawn"};
Parameter(s):
_this select 0: STRING - Image path
_this select 1 (Optional): NUMBER - Delta X
_this select 2 (Optional): NUMBER - Delta Y
Returns:
True
*/
BIS_fnc_customGPS_Params = _this;
true
D:\Arma 3 - Unpacked Data\functions_f\gui\fn_customgpsvideo.sqf
D:\Arma 3 - Unpacked Data\functions_f\gui\fn_customgps.sqf
D:\Arma 3 - Unpacked Data\functions_f\gui\fn_customgps_spawn.sqf
These file use BIS_fnc_customGPS_Params
Hello; getting a CTD while loading some maps / entities on the latest MacOS branch. Here's the eon.txt complaining about some problems with the shader
please create a ticket, reports get lost in here 🙂
👍 thank you
One of the weirdest bugs I've encountered: https://feedback.bistudio.com/T174962
iirc there are multiple minimap displays
I think I've also encountered this problem with other vehicles too
mostly air vehicles iirc
like the Blackhawk or Little Bird (well their Arma 3 versions; was it Ghosthawk and Hummingbird?)
and iirc I found a workaround. but don't remember anymore 
But placing a vehicle shouldn't break scripts anyway 😦
well I think it's because the vehicles create some other display that breaks it
So if vehicles had their own unique display would probably fix it?
dunno what the problem was exactly. maybe
I guess I should keep the ticket so devs can see it anyway, I almost went crazy trying to figure out why a script that worked perfectly fine on some missions broke in others
this is late to ask but I have to ask if it's possible that we get more sides than the east, west, resistance?
I think this was answered at some point with "No" as the sides are set in engine and changing that is too risky.
ok
pretty sure it would need massive rewrite of half of the game xD
including some 20 y/o code
Class Missions (Uppercase C - instead of class) is apparently making the A3 server not able to launch
worth a ticket? or just a BIKI note
Does it throw an error in the RPT?
it doesnt launch the server at all
class SPE_M01_Liberation_of_Saint_Gilles{
template = "SPE_M01_Liberation_of_Saint_Gilles.SPE_Normandy";
difficulty = "Custom";
};```
any time a capital 'C' is used the server throws an M class error and exits
https://community.bistudio.com/wiki/Description.ext has a warning
there we go
desc.ext != server.cfg
i got a launch error with it captialised
it wont push through
references to class needs to be lower case in my user trial
basically yes: use class and nothing else
i guess same parser
As long as it throws an error then it's fine.
It's a ripped-off C parser, right. Can't expect it to suddenly ignore keyword casing.
I've just thought that, why is wounded rvmats are also defined (still) in uniform vehicle's damage >> mat[]?
Also is class Wounds still a thing? Does it do anything?
it's terrible what the bots did to the ft, well at least our tickets don't get lost with all the spam🫣
There is no logical reason to keep it limited.
Question is, is "Class" supported in config.cpp ?
Do you mean upper case C?
yes, like I wrote it
At least pboProject hates it. But let me check if mergeConfigFile hates it
No, diag_mergeConfigFile fails to compile
Class test
{
};```Failed
```cpp
class test
{
};```Merged properly
no surprise
Ok so then we don't need to document or change anything.
server config, description.ext, config.cpp they are all config files.
And if you don't follow config syntax you're doomed
This is the error BTW
Here is a another tickets related to Custom Compositions defined in config.cpp. They seem to break the asset browser search https://feedback.bistudio.com/T175089 (at)dedmen.
Seems that the Asset Browser Search only searches for the EditorCategories and EditorSubcategories in CfgGroups 
I've found out that we cannot have an AI unit in empty and/or empty object (house?) into a faction CfgGroups recently. Let's say if I wanted to have a predefined USS and its weapons into one CfgGroups, cannot make it work properly
I don't understand that question or statement
I think he's talking about compositions?
I even forgot that I've made this post 😅 let me make an example real quick
something like concrete bunker (side Empty) + ship turret (side Blue) in a single CfgGroups entry?
Bottom: source
Top: exported compositions BIS_fnc_exportCfgGroups
Left: class Empty one
Right: class West one
I can understand they're empty if I do it as an Empty comp, but if was a side (BLUFOR etc) comp you can spawn them properly yet it fails to assign into Eden
delete3DENEntities (all3DENEntities#0) results this
still sounds like a Composition job 🙃
@trim acorn i would prefere this
http://feedback.arma3.com/view.php?id=22000
or this
http://feedback.arma3.com/view.php?id=23328
but you know 1 step at a time :P
22000 (configured weapon in inventory) is a great reason why scripting systems based on individual script commands are always lacking something.
one little detail is still annoying me most of the weapon stuff is modular except the magazine which is still defined by the model...
Does the engine not allow you to define arbitrary new attachments? I.e. you could make a gun that takes magazines as attachments. Would have to do some scripting to keep things synced but might be doable.
maybe but the engine should be able to do this
i mean they advertised it so much "oh look the weapons are fully modular" but than this xD
Is there any possibility of revisiting this: https://feedback.bistudio.com/T73775 [Exact time is visible in the map even if you don't have a watch or GPS]
Even KK wanted it to be fixed, back in the day. It would be a nice bit of verisimilitude, and the reason why it wasn't fixed in 2013 is..........stupid.
Spam account
https://feedback.bistudio.com/p/joystailpro123/
all recent spams were taken care of (as far as I saw...)
You missed this one and from the message after.
I didn't look that far 
Yea that spam was aging for a little while 😅
I love the new update 😄
Literally downloading more ram
Just get a bigger screen 😄
new crash update xD
Anyone know if the scripted helicopter damage will be fixed on the new update?
read the changelog
Got a link off hand?
it was never able to go below 200 though… did this update change that?
yes
You can go to 5 now
same with obj. VD, but the command returns the correct value.
the command returns the correct value
you mean obj view dist command?
yes
I guess view distance is still limited to 200, but the options menu suggests otherwise.
need to check with Cheetah and some poor helicopter volunteer 🤣
after setObjectViewDistance [5,5]; Cheetah turns invisible as soon as i stop touching it 
Yeah that's what I thought. I thought there was engine limit.
But didn't bother to test it and assumed it'll be alright 
Guess there is after all. I assume we won't fix that
I thought you've fixed the issue but you've only fixed the cap model? 🙁
What is the ticket for it?
I see reyhard already replied to you there.
But since the ticket is about to fix the entire custom rvmat issue, not the specific model, so I thought it was fixed
And please check the timestamp of my picture and reyhard's reply there, reyhard's reply is only 40 minutes ago
anyone did some fps comparsion yet?
we seem to get considerable worse fps with same settings in 2.14
Is there a legacy build available?
Arma3Legacy212seemingly yes
Cheers
whats the most meaningful/establish benchmark mission these days?
There are also two others, "Altis bench" and "stratis bench" , both on the workshop by the same person
I feel these two are more representative of general single player performance
I'll make a ticket later, but you can hold the charge on the T-100X even after leaving the gunner seat, as long as you hold the fire button. It'll then fire when you release
Not important but just pointing out that the "!" looks like an "I" when the achievement is displayed on steam. 🔍 "workI"
Started Arma 3 with at least 10 mods loaded - thanks to all modders for their splendid work!```Definitely a !. It is your display/Steam "issue"
Yes that's exactly what I'm saying the steam font makes the "!" look like an "I"
Just something random I noticed lol
And there's nothing we can do
16:55:42 8.0: SteamQuery: class SteamUGCDownloadPreviewFile - failed - Code: 21 - The user is not logged on.
16:55:46 11.1: SteamQuery: class SteamUGCFileDownload - failed - Code: 16 - Operation timed out.
trying to launch a mission from workshop via ingame scenarios menu
says first:
Downloading Steam file (0KB/0KB)
after a few seconds:
This user is not logged on.
if i click the scenario (workshop) button at the bottom, it opens steam overlay with me logged in
is Steam having problems atm?
Steam is down.
Steam appears to be having issues at the moment
Simply tells you if Steam is up or down.
Displays status of Steam client, Steam store, Steam community, Dota 2, TF2 and Counter-Strike
Like I said, not important, just pointing it out 👍
I have an issue where all the helos in the game are acting like they are sling loading something from the nose ever since the update. It’s weird bc it only does it on joystick… turned off all stabilizers, unbound all trims, checked center on stick w no luck…
what about vanilla assets?
Those are cup assets. Can you confirm that it still affects you with every mod unloaded and in a vanilla helo?
did the new update change something with config-class count & select command behaving? because I have problem with inherited classes
configProperties I think was broken for a bit in 2.12
need example code to see what u mean, and are you on profiling branch?
yes im on prof
class A
{
var = 777;
};
class B : A
{
};
str ( count (missionconfigfile >> "b") ); // Prints zero
should print 1
ok well its driving me crazy ;D
I shall remember to test that monday
cool thx 👍
did you close and reopen the mission?
yup
Though```sqf
str ( getnumber (missionconfigfile >> "b" >> "var") );
prints 777
does it w no mods or dlcs loaded too
can you revert to legacy v2.12 to make sure it's caused by the update?
use the Arma3Legacy212 code in Steam->Arma 3->Properties->Betas tab
My "small" proposal for new hash map objects: https://feedback.bistudio.com/T175336
Very neat stuff, too much to implement but maybe we can select a few
More spam:
https://feedback.bistudio.com/T175341
Will addWeaponWithAttachmentsCargoGlobal allow negative values to remove weapons? We can't manage cargo in a easy way if there is no solution for weapons, at the end we still need to use old hack solutions.
behaves the same in legacy 2.12, though?
private _cfg = configFile >> "CfgVehicles" >> "B_MBT_01_cannon_F";
[
count configProperties [_cfg, "true", true],
count configProperties [_cfg, "true", false],
count _cfg
]
// returns [473,11,11]```
and for "_i" from 0 to (count _cfg - 1) do {... _cfg select _i is used in bis_fnc_overviewauthor, which is called by BIS_fnc_arsenal to display stuff author. And i remember having reports of Arsenal not displaying inherited author some time ago (and all vanilla (and all properly working mod) content explicitly includes author property, even if it matches the parent's one) 
like, here's conversation from almost a year ago: #arma3_config message
at least #key/#newkey and type checking will be great
@rose cloak the problem is count doesn't count inherited variables
so the question is should count include the inherited properties/variables
typeName HASHMAP
Will break for all pre-existing code
Additional having isEqualTo check the nested object type would slow down isEqualTo
And it didn't since forever 
yeah I was actually wondering if it ever did
Don't make me download A2 again 🤣
really, I wonder what's the point with that...
yeah, it works
but I'm sure that it's already slower than most comparisons against other types
Does it rely on the binning being deterministic given the same hashmap data?
I guess it's more likely just cross-comparing by key
doesn't seem to. At least similar loop in the Arma: Gold Edition does show "scope"/"side" as a properties for "SoldierWB" class (rifleman), but not for "SoldierWG" (grenadier)
well, i'd argue that "the command behaved the same way 3 games ago in 20067" does answer the first half of initial question of "did the new update change something with config-class count & select command behaving?"
And for the second part of that question, " because I have problem with inherited classes" - we thankfully have configProperties with inherit set to true since A3 1.36
We were going to feature the T-100X in our session tomorrow but it turns out it doesn't work properly in multiplayer, which has put a damper on it
https://feedback.bistudio.com/T175344
yeah I just have to use configProperties then, thanks for that. The behaviour with count just surprised me and I dont get why it ignores inherited properties
CfgRemoteExec will block it in MP
Maybe that's your issue.
class BIS_fnc_setCustomSoundController { allowedTargets= 1; };
class bis_fnc_fire { allowedTargets= 0;};
CfgRemoteExec is set to blanket mode 2 (all allowed). Also, the core functioning of a vanilla vehicle shouldn't be broken by CfgRemoteExec, in principle.
Only if someone write this code from last update
And it can be avoided by introducing copy of command with slightly different name
It doesn't needed
there is no need to check for the nested type, because it is determined by the value in the hashmap, which means it will already be checked when comparing hashmaps by its values like before this update
And even more - this can increase performance, because it can determine if two hashmaps are different early just by checking nested type first, before check every key in hashmap
Like typeNameOOP or something
The only slowdown here - it is isEqualType, but for cases when you compare with hashmap type on right side
And I think it will be negligible
And this also can be solved by introducing copy of isEqualType* with slightly different name like for typeName
PS Speaking about performance - it's should be already slowdown because now every command that modifies hashmap must check sealed flag and finalization by compileFinal
Recently I've found out that IR seekers are basically looking for “0 or not 0” vehicle heat signatures. I've found out this is somewhat strange to have, because I've expected to wait a bit to have proper signature to be detected?
Maybe possible to add some sort of IR intensity parameter, 0 to 1? And then different IR seekers/sensors can have different thresholds. Might be something that can be generalised to "signal intensity", and applied to all types seekers/sensors.
Almost-helpful spam comment: https://feedback.bistudio.com/T175338#2497454
it mentioned pizza!
https://feedback.bistudio.com/T175392 vanilla fn_animalBehaviour.sqf script tweak request, adding the dog and sheep sound reactions
I noticed we neither have allLogics nor allLocations commands. Would it be worth requesting that? Would complement allGroups, allPlayer etc.
I'm not as such against allLocations, but worth mentioning that since 2.14 you can effectively do it with nearestLocations by using an empty type array and a worldSize radius
Good point. I actually didn't know that.
The Put event does not seem to be firing on containers (Take does). I'll make a ticket in the morning.
Just found out that ctrlText does not return a line break from a structured text control. I don't think a line break should not be detected
Basically this is what I've tried and failed because it can't detect br. I mean, this formatting should be done by engine side in the first place, but anyways...
Are you fixing the broken scenario menu formatting?
And failed miserably
Well, they haven't done this for a decade after all, we can expect no hope here! 
There is always hope 🫂
I thought with the introduction of regex we could match the patterns of Steam formatting and replace it with Structured Text.
But I guess it's not that easy, is it?
I mean, it already uses regex to mass replace and patterns, but since the ctrlText only returns it as a one-line text, we have no place to line break
I don't usually like to pester about tickets (...more than once) but I'm very eager to be able to properly use 2.14's flagship vehicle in MP.
https://feedback.bistudio.com/T175344
Why was T171566 hidden?
most likely risks of abuse
I gave you access back to it
Thanks
I thought I got a ping 
I pinged you but then figured you probably didn't touch the ticket (turns out you did)
I did yes. well I think I shouldn't have used that access
append for strings?
I know 😄
https://community.bistudio.com/wiki/joinString for hipsters 
but instead of ```sqf
_testString + _testString = "Testing!";
// Maybe
_testString append "Testing!";
i think that does not modify the original string?
i don't think any string operators do modify in-place now, though?
well lets have one ;D
like, even "Hello" insert [" World!", -1] is the only insert variant that returns the result, non-string variants modify in-place
@untold sky Do you have access to our Spearhead Gitlab issue tracker? An issue that was reported by QA seems to be 100% reproduceable in vanilla as well. Just wondering if I need to duplicated all the info on BI tracker as well
https://feedback.bistudio.com/T175457 Here is the official ticket.
I have access. I need it duplicated in our internal tracker (which is not FT). So no
they're still at it, https://feedback.bistudio.com/T175471 added 🤣
what we're looking at? Other than big draw distance? 
myb color scheme ?
low object draw distance?
ya’ll all blind
clean your monitor, it's dirty
no haze, no?
I'm just wondering why your landing gear is still down.
malfunction
ah, I've been using no haze mod for last 5 years so it's normal look for me 😅
oh, you
so what actually was done? Some weather influence on haze or commands to control it? So far we had only terrain config.
Nothing has been done yet, just assessing options for now
taking a pic of your monitor vs using printscreen :p
all nuked, thanks
Shadow distance have strange behavior - it reduces gpu load with increasing distance above old max of 200m in forest area (tested in southern region of Livonia) if objects quality set to any level except the worst... not sure if something else, just quick test
(Dropped from capped 60fps with 200m shadows (80-100% load, full maxed out, 1440p 3km objects) to just 10-15fps with 1km shadows (15-20% load))
sounds like a CPU bottleneck
It's 13700K with just 3% load
yes that's arma 😄
like when i say CPU bottleneck i don't mean your CPU is inadequate, just that the engine is doing something really funky
Looks like too many synchronization inside engine
Or someone just place "sleep" in render queue for shadows 🤣
@north gull try https://game.intel.com/story/intel-presentmon/ and see the difference between frametimes and "GPU busy" metric
I will check tomorrow
Almost everything runs in the same thread in the A3 client.
@north gull What shadow quality setting is that btw?
Any, doesn't matter
Even if you turn off them, but set 2km in shadow distance
oh, that's fun
The only setting that affect this behavior (except shadow distance itself) is object quality - the worst preset doesn't have this behavior, other - have
is addForce having no interpolation at further distances intentional or a bug
finally got around to testing it in isolation and it has none
? It's a force?
force has no "distance" metric involve in it
force = mass * acceleration...
I think the command actually adds an impulse.
no impulse... but I haven't checked to source
I think he is talking about the second param for the command that takes a position of the force relative to the object
wait, that command is dumb
No shit :D
all i know is theres no interpolation on its rotation
that's a torque
and it gets worse the further you are
WHY
physics
Depends if the position & direction are in line with the center of mass or not.
You could use it for torqueless acceleration if you really wanted to.
use rv4 as a physics simulation to engineer time travel so you can go back in time to kill isaac newton
I DON'T WANT TO LOOK AT THE ARMA 3 SOURCE
God knows if it works for non-physX objects.
just addforce has no interpolation across network
when its turning
same for addtorque
John there is no position command for that function
oh yeah :P
So I guess it breaks the input into physX addForce + addTorque.
I'm going to ignore that
shrug i just used addforce
this is a dumb command
i gave the object no gravity and rammed it with myself
using addforce
there was no interpolation of its orientation
it was local to server
yeah, it also very quickly breaks if the sim type is not a physx type
it was a physx type
thingx iirc
this is the same issue i had with my capital ships stuff its just reducing it down to making sure its not my fault (its not)
it will need to me applied each frame
it was
you are doing that right?
yep
with a time step?
i did both with and without
yes
i did both
it had the issue in all instances
ive tested extensively when doing my capital ships
around a month of debugging just this issue
always ran into the issue of no interpolation
I assume that the "velocity change accumulator" is getting messy and breaking or something
who knows at this point
although with the sizes physx handles... could it be floating point precision?
dont know enough about backend to give an informed guess
For orientation interpolation? No.
i have no clue why it doesnt do it then
Wait, you said there was no interpolation of orientation even on localhost/SP?
did i
its only ever been for non local objects in testing
never managed to test localhost with non local but would imagine its the same
sp obviously everything is local
its updating on network updates but theres no interpolation between those updates
I'm guessing that it just doesn't send angular velocity at all.
that'd be my assumption
Although second option is that it's bugged :P
inb4 its just an unused param because someone forgot to copy paste a line
Like in most cases you can't tell. Might have slipped through given that Arma's physics is expected to be a goddamn mess anyway.
who knows, shall wait for one of the green beans to advise
having three different physics engines running will do that
dedmen said his assumption was vectordirandup would override angular velocity
which i would imagine would extend even if youre not setting vectordirandup yourself but who knows
but even with free-floating objects + addTorque it didn't interpolate?
at this point i would even take a local param to addforce tbh
i have never got interpolation of orientation
in any combination of commands for manipulating orientation
and i spent over a month trying to resolve this nearly all day every day
also by this i mean the further you are away the fewer network updates you get for the object
which makes sense theres just never interpolation for it
whereas there is for velocity
Any chance you want to profile that on the profiler branch?
diag_captureFrame
added diag_captureFrame results in comments to ticket
was replacing Low graphics with Very Low in texture, objects & terrain an oversight or intentional
You mean there's no Low now?
Only Very Low?
Yea I miss low setting.
yep
Hey , it would be great if weapon shooting would be simulated the same distance as the sound goes. For some Rifles it only goes 400 meters far.
Even tho the sound would go 3000 meters far but you cant hear it since they only get simulated when you are close to it
It is different from weapon to weapon. Snipers for example are working 1200 meters away but the sound would normally go 4000 meters far
https://feedback.bistudio.com/T175562 this has been on my wishlist for a long time and was surprised to not see an existing ticket (or at least I couldnt find one )
https://feedback.bistudio.com/T169148
Can be closed
Slightly bumped, but here're features I always miss in Arma 3 Launcher:
- Add a tag into an item (Local/Workshop/maybe CDLC) and activate/deactivate them all in one click
- Add a CDLC dependency notification if one loads a Mod that depends on some
PS. Let's say I add “RHS” tag all RHS Mods and click the tag (I don't really know where to put the UI but) will activate the whole tagged Mods/CDLCs
too complex at this point imo
having an option to append mod list to currently selected instead of replace would be enough.
ye, BI open source the launcher so we can add features ourselves as I know nobody wants to touch this old code 🙏
It's .NET so we can peek inside anyway
Nobody wants to touch two decades old engine after all 
Anyways your idea sounds way better than my idea
has nobody made a third party launcher yet
WithSIX
one guy from my unit has made this one for linux:
https://github.com/muttleyxd/arma3-unix-launcher
One thing I realized is we have no function/command to get/set/whatever related to Quaternion. Should be handy to have as an official functions at least
Since we have a plenty of commands that can get dir and up, it would be a good addition to have a way to rotate them easily other than vectorCrossProduct
BIS_fnc_rotateVector3D?
yaaay, 3d maths
Oh I see, so more like what vectorModelToWorld does but without the object?
Yyyyeah, I guess?
inb4 https://community.bistudio.com/wiki/BIS_fnc_transformVectorDirAndUp in modelspace 
a Leopard specialty =D
Yeah Mr cat, please do
evenshually-bait
Is there a reason tan 3 returns 0.0524078 unlike any other language where it returns -0.1425465?
I'd like to ask before I make a ticket
What is your Architecture? Arm x86?
Follow up... what is your cpu
oh... lol
radians
AMD64 - Ryzen 5800X3D
Ah, makes a bit more sense then - Thank you
Made a ticket for add command. Sorry if this has been asked before. https://feedback.bistudio.com/T175587
IMO an append overload would be better, both for dev and for the name itself (add is waaay too generic) (edit: ah for Numbers too - mieh?)
anyway, marked "not possible" just now 😄
Dedmeeeeeen…!! ((✊)) 😄
https://feedback.bistudio.com/T175589
Stole veteran's idea, but here is my wish 🙏
closed too
oh i thought everything was passed as reference so it might work. or is there other reason?
just curious 🙂
I don't know
Dedmen closed the ticket
yeah saw that...
I don't remember "fully", but that was because even "at all" was more than what Arma 2 or VBS2/3 supported... though the devs have proven (via the underbarrel and more recently the UGL round slots) that the amount of slots accessible to the player IS subject to change.
@tulip wasp Chalk it up to magazines not working like attachments do? Though this is probably one of the things that's changing in DayZ/Enfusion since they've been quite in flux under-the-hood.
there seems to be quite some spam in CDLC projects
is anyone checking and cleaning these up?
or anything else can be done?
cleanup happens, and we might have a special weapon soon™

update: Dedmen brought up the big guns and we can now nuke spammers more easily
😮
What about adding "nofollow" to comment links so it's not worth to spam the tracker as the links won't be indexed:
https://www.conductor.com/academy/nofollow/
It should be a single template edit
1/ they will most likely still spam (either bots or human farms, doesn't matter)
2/ nobody will touch FT's code
3/ we should be only some months away from a significant FT change
on the one hand, yay new FT
on the other hand, oh no my tickets
I wish we would get that voting feature back that we had in the beginning
I think one could expect that magazine model switching is coming
@rapid badge I'm thinking make a model where the magazine is invisible. Put in a new memory point for a new "attachment slot". Config that slot up but don't expose it in the inventory UI. Then add tons of EH's to handle switching the attachment dependening on the currently loaded magazine.
You might loose magazine movement during reload animations, though, unless you can animate weapon attachments (animate the memory point they're proxied to?)
I'm not saying this is feasible or a good idea, just that it might be possible to hack something like that together
AI team members ignoring AI Squad Leader order to rearm https://feedback.bistudio.com/T175631
was there a ticket with request to have a height option for Hide Terrain Objects module or is it just a dream? I'd bump it up rather than making a duplicate
Fixed: Server not closing files in some cases (like server.cfg not being editable)
Wonder if similar issue can be fixed with mission PBOs. Sometimes when we restart the server and update the mission players are unable to rejoin, mission PBO remains locked and downloading new file doesn't work and you just get thrown off the server.
Been an issue since like Apex
Happens seemingly randomly, only to some players
@granite mason It might be one of those "not worth it" things for BI since it'd require going back into every single last gun model.
"config that slot up" - there's the problem. you can't define custom attachement slots.
@unborn flame Argh, that's annoying. Does that mean you can't define custom memory points for it? Because you certainly can define new config-level attachment slots, after all that's how ASDG JR works
they still only use the underlying (hardcoded) system for the available attachements...
@unborn flame What's the hardcoded bit though? It seems like the slots are just proxies in the model, so you should be able to define new ones, right? https://community.bistudio.com/wiki/Arma_3_Weapon_Config_Guidelines#New_config_parameters_.7Bconfig.cpp.7D
Mh, the swapping bit might be hardcoded, you're right. However if by "link to the UI" you mean the inventory screen, that's all just living in data, not engine, iirc.
Has anybody actually tried this? Make a new proxy object, config it up, see what happens?
if you put item xyz in slot abc the UI has to tell the engine that slot abc is now filled with object xyz and swap the model
thats the link
@unborn flame Thanks! Makes sense
AnimDone - EH fires constanly for dead units
https://feedback.bistudio.com/T175658
It stops once you disable simulation of the body, so maybe it's related to bodies draining quite a bit of perf.
But if you disable simulation on the body players are no longer able to take items from its inventory 😦
Looks like T-100X doesn't work in MP at all anymore
No matter what I do, just doesn't shoot
server-created, client-created, locality combos, AIs on other seat, never shoots
That's weird. I've literally just (in the last hour) been testing two vanilla missions with it where it works fine 🤔
I just tried it on server, local with AI, local with ppl, server with AI, all worked.
I did notice when AI destroyed cannon on hostile T-100X, with AI still inside, they refused to fire at it to fully destroy it. Commander faced disabled tank but gunner completely ignored it.
PIP view distance on a low amount is showing as extreme, will ticket in a few
extreme doesn't mean "too high". it means too low or too high
oh guh
no need for ticket. it's intended
roger doger then
Hmm, does it use any remote exec'ing?
Didn't find anything yet
Afaik cannon is scripted, and i saw it mentioned...somewhere. i figure it does
fired = "_this call BIS_fnc_RailGun_01_chargingStart";
```oh yeah
Expected it to be engine-driven
I see these function do BIS_fnc_setCustomSoundController but that's whitelisted 🤔
Found the issue:
[_vehicle, "cannon_railgun"] call BIS_fnc_fire;
which has
if (!isServer) exitWith
{
_this remoteExecCall ["BIS_fnc_fire", 2];
nil
};
https://feedback.bistudio.com/T120107
I assume HitPart always passes null for global projectiles like missiles and passes proper entity for local projectiles like bullets, can this ever be fixed in the engine?
it returns something for me. it's not null
What were you firing?
RPG32
17:28:29 ---------------------------------------------------------
17:28:29 "Frame 3505975: HitPart: O_APC_Tracked_02_cannon_F"
17:28:29 ["189ee046040# 1780600: apc_tracked_02_cannon_f.p3d (O_APC_Tracked_02_cannon_F)","B Alpha 1-2:1 (Sa-Matra) (B_Soldier_F)","<NULL-object> ()",[23625.2,18543.5,4.86392],[173.227,53.1108,-0.322573],[],[150,28,3,1,"R_PG32V_F"],[-0.70793,-0.184872,0.681657],2.48759,"a3\data_f\penetration\armour_plate_heavy.bisurf",true]
17:28:29
17:28:29 ---------------------------------------------------------
17:28:29 "Frame 3505975: HitPart: O_APC_Tracked_02_cannon_F"
17:28:29 ["189ee046040# 1780600: apc_tracked_02_cannon_f.p3d (O_APC_Tracked_02_cannon_F)","B Alpha 1-2:1 (Sa-Matra) (B_Soldier_F)","1780602: empty.p3d (ammo_Penetrator_RPG32V)",[23625.2,18543.5,4.8639],[902.997,276.785,7.13465],[],[480,0,0,0,"ammo_Penetrator_RPG32V"],[-0.70793,-0.184872,0.681657],2.48759,"a3\data_f\penetration\armour_plate_heavy.bisurf",true]
17:28:29
The missile itself is null, penetrator isn't
getText(configFile >> "CfgAmmo" >> "R_PG32V_F" >> "simulation") => shotRocket
getText(configFile >> "CfgAmmo" >> "ammo_Penetrator_RPG32V" >> "simulation") => shotShell
Its not an issue of submunitions though, I've been testing it with AA missile recently:
getText(configFile >> "CfgAmmo" >> "M_Titan_AA_long" >> "simulation") => shotMissile
16:12:09 ---------------------------------------------------------
16:12:09 "Frame 2853577: HitPart: O_Heli_Attack_02_F"
16:12:09 ["O Alpha 2-1:1 (O_Heli_Attack_02_F)","B Alpha 1-2:1 (Sa-Matra) (B_APC_Tracked_01_AA_F)","<NULL-object> ()",[24079,17625.8,58.1367],[0,0,-0.0196132],[],[80,60,6,1,"M_Titan_AA_long"],[0.345211,-0.0911817,-0.934086],1.81265,"a3\data_f\penetration\armour_plate_20mm.bisurf",true]
```Null once again
Interestingly shotMine appears properly in the EH despite also being slow projectile 🤔
Try checking the warhead type for consistency. This smells like a difference between HE (no penetration simulation, deleted instantly) and AP/HEAT (projectile remains active for penetration simulation)
well in the case of the rpg32 the missle projectile would cease to exist
as it gets deleted when the penetrator spawns
hitpart probably runs too late
Testing DAR:
getText(configFile >> "CfgAmmo" >> "M_AT" >> "warheadName") => HE
getText(configFile >> "CfgAmmo" >> "M_AT" >> "simulation") => shotRocket
getText(configFile >> "CfgAmmo" >> "M_AT" >> "submunitionAmmo") => ""
Null in both direct and splash, same as other missiles and rockets:
18:03:50 ---------------------------------------------------------
18:03:50 "Frame 3810754: HitPart: O_APC_Tracked_02_cannon_F"
18:03:50 ["189ee046040# 1780600: apc_tracked_02_cannon_f.p3d (O_APC_Tracked_02_cannon_F)","B Alpha 1-2:1 (Sa-Matra) (B_Heli_Light_01_armed_F)","<NULL-object> ()",[23626.8,18547,4.84325],[194.642,-256.054,-34.758],[],[300,50,8,1,"M_AT"],[0.0116035,0.155283,0.987802],1.46166,"a3\data_f\penetration\armour_plate_heavy.bisurf",true]
18:03:50
18:03:50 ---------------------------------------------------------
18:03:50 "Frame 3810754: HitPart: O_APC_Tracked_02_cannon_F"
18:03:50 ["189ee046040# 1780600: apc_tracked_02_cannon_f.p3d (O_APC_Tracked_02_cannon_F)","B Alpha 1-2:1 (Sa-Matra) (B_Heli_Light_01_armed_F)","<NULL-object> ()",[23626.8,18542.6,4.5291],[0,0,0],["damagehide"],[300,50,8,1,"M_AT"],[-0.786542,-0.617537,-0.000196199],0.266976,"a3\data_f\penetration\plastic.bisurf",false]
18:03:50 ["189ee046040# 1780600: apc_tracked_02_cannon_f.p3d (O_APC_Tracked_02_cannon_F)","B Alpha 1-2:1 (Sa-Matra) (B_Heli_Light_01_armed_F)","<NULL-object> ()",[23626.4,18544.2,5.84574],[0,0,0],["otocvez"],[300,50,8,1,"M_AT"],[-0.786567,-0.617505,-0.00015249],0.710202,"a3\data_f\penetration\metal_plate.bisurf",false]
...
Well, so far it looks like shotRocket and shotMissile are always missing in HitPart
Yeah, but other missiles and rockets without penetrator are also missing from the EH
what about like - HE shells or grenades
Its there.
18:34:21 ---------------------------------------------------------
18:34:21 "Frame 4072195: HitPart: C_Hatchback_01_F"
18:34:21 ["187d6bfc080# 1780685: hatchback_01_f.p3d (C_Hatchback_01_F)","B Alpha 1-2:1 (Sa-Matra) (B_Soldier_F)","1780686: handgrenade_throw.p3d (GrenadeHand)",[23621.3,18566.9,4.43768],[0,0,0],["glass3"],[8,8,6,1,"GrenadeHand"],[-0.786147,-0.61804,-0.000111496],0.678808,"a3\data_f\penetration\glass_plate.bisurf",false]
Looks like its not about entity being slow or global 🤔
Can't test if shotSubmunitions is affected by this too, but it feels like all projectiles that can turn into submunition, even if no submunition is defined in config never appear in HitPart EH
Tested A3 Alpha (0.5), same picture there, HitPart doesn't have missiles in it:
[5a632040# 160080: galkin.p3d,B Alpha 1-1:1 (GhostBear),<NULL-object>,[1578.47,5200.06,6.10836],[0,0,0],["wheel_1_1_hide"],[800,13,10,1,"M_AT"],[-0.291083,-0.956697,-0.00132837],0.674498,"A3\data_f\Penetration\tyre.bisurf",false]
So looks like its an old issue
well the reason is that the EH gets called in the next frame
Why is it different from other simulations?
Thanks for having a look at it!
should be fixed in the next dev version (once it's approved by Dedmen and KK)
Compass move on 2D map is not reset when close the map with holding LMB
https://feedback.bistudio.com/T175707
Workthought: isTouchingGround, can we have an extended return of it? Like how many wheels they have and how many is touching ground etc. It's possible to have selectionPosition for contact LOD, but not reliable to get which is touching or not
Also for heli and planes there is no generic way to get how many the wheels are and where are they
(AFAIK)
One thing I also wonder: what is the lag when the in-game time reaches 0:00 AM (midnight)? Sometimes it bothers me
https://feedback.bistudio.com/T175720
I've made multiple suggestions into one but I think 1 is nice to have at least
I'm trying to improve colourblindness support in a custom radio channel system. Unfortunately, not being able to customise the VON speaker indicator is limiting it a bit, since I can't display the channel name and can only distinguish by colour.
https://feedback.bistudio.com/T175722
Is the change on dev branch yet?
(I guess not, usually build is mentioned in the ticket)
I just noticed that when searching for assets in eden editor (in the right sidebar) you can't look for any assets starting with "mod" because game expects you to provide a pbo classname. So when you want to find a Modular Bunker HQ, you have to search for bunker because modular will provide 0 results (game thinks my query is mod ular or something). Was it always like that?
Can confirm, ["Arma 3","Arma3",215,151027,"Diag",false,"Windows","x64"]
I'm on latest stable and I don't remember having this issue in 2.12 (although I don't remember searching for any mod* stuff in Editor either back then)
remove spam pls https://feedback.bistudio.com/T175707#2504305
still curious why this is not possible: https://feedback.bistudio.com/T175587 ? just asking because i want to better understand what is doable so I wont make impossible feature requests, thx 🙂
Maybe the word add is too common
could be something else
inb4 "it's better to nag Ded to hear what Ded thinks"
pinged @untold sky for you
alright thx 🙂
Can I know why it was an issue and what kinds of shots are effected so I can script an exception until its fixed in the engine?
Is my assumption about shotMissile and shotRocket correct?
I don't think it's approved yet
I didn't check all but all I know is that those two were affected yes
do submunitions count as separate projectiles or a part of a parent projectile, though? 
They are, I dont think they have any connection to the parent
strings and scalars are treated as immutable in SQF
afaik
so they will (and should) never be modified by ref like hashmaps and arrays
Making them changeable by ref is a really bad idea
Also think about the amount of work require to handle every type.
cant you make exception 😄
that's Dedmen's territory. if he says no then no
What about numeric_variable++?
asked before, no can do
Actually struggling to think of any SQF language features that assume strings and scalars are immutable, although my interest is only curiosity.
Darn
then they had to change the name to SQF++ 😄
I guess if _var = 2; _arr = [_var]; is using the reference to the original var rather than making a copy then making scalars mutable would be a shitshow.
Once upon a time I figured that scalars would be a special case with in-line storage but it turns out SQF is using something like 48 bytes per scalar.
ARRAY add INDEX 
that's actually why placing arrays into arrays is slow(er)
it will check for recursion
it gets even worse with nested arrays
ticket?
While you dig into the deep hell of Eden Editor search code. Perhaps you find a switch you can flip to enable wildcards? 😄
https://feedback.bistudio.com/T171218
sighs
Attempting to put together a FT report for the upper-floor pathing bug, find a case where the guy just stops in the doorway and recalculates a complex path twice a second forever.
maybe it's just bugs all the way down.

@ Dedmen: almost every time you try to link the crash dump guide page on the FT, it's actually about:blank. If you're using a macro or something for that, might be worth double-checking it
no.
Size must be equal
Then it does a Get lookup for every key, to check if the second one also contains the key.
If it contains the key, then it does isEqualTo on the value
some config things were touched, but nothing should have changed.
There was already one bug in.. configProperties I think where something was accidentally changed
I don't quite understand that text.
Probably needs ticket with repro. At best a script only repro that can just be ran in editor
Is this just the same as https://feedback.bistudio.com/T175612 ?
Then do that and send a patch 😄
Launcher additions are annoying as they need localization
Can I unpack it to GitHub first? 
It is passed by reference. Which is also a reason why it doesn't work.
I tried that.
_var = 5;
_var += 2;
At first run it returns 7, at second its 9, at third its 11
its the same thing yes.
My fix is not a full fix.
I talked about how the Arma client flushes out files when you tab out. Which some people very vocally complained about that it makes the filecache useless.
Well server doesn't do that, that also causes the files to stay constantly locked.
I could make it force-flush every frame or every few seconds, but that destroys the filecache. Maybe just doing it on mission unload or something, but thats also meh
Not really. Unless you want to pay the performance cost for it which I don't.
Engine just has one boolean
Grab profiling and check 😄
I have no idea
separate. I think I had some trouble with some recent changes because they don't know which projectile they came from
That would be smart.
But in SQF every value and every script instruction is a pointer to a class instance. That's part of why its so slow. That's extremely cache unfriendly.
Do you know of anything else in game that has wildcards?
I just open the page, then go to address bar and copy it. I have no idea how it happens. I never notice it
Negative =/
Yeah I think we don't have any wildcard expression kinda thing code anywhere
We have regex now but.. eh
Is possible and would be neat, but I won't have the time to implement that
Too bad.
yes, I also mention in my ticket that disabling shadows don't return FPS, but it isn't primary idea of my ticket (i focus on low gpu utilization and low performance especcially in forest area)
in the middle of additional information
Not good to mix it in with the disabling shadows bug.
One is just how the game works and is expected and will not be fixed, the other is a clear bug
not good, but this was not the main idea of my ticket, and is mentioned only to show that something is really wrong
i can remove this mention, but it will not change anything because it is just one picture of 10
One is just how the game works and is expected
but I think if you fix this bug, you also fix my ticket, because I think they are both have same source
i understand you being a bit touchy on what seems like a report about newly introduced bug in recently changed code, but wow, that's some epic necropost by standards of this server 🙃
I just managed to clear my months backlog and already have 5 pings again 
Ping-on-reply being on by default remains one of Discord's all-time worst changes
script error massage in "CoF: Green" firing drill mission https://feedback.bistudio.com/T175788
Are (Arma 3) Biki bug reports done on the Arma 3 project on the Feedback Tracker? (i,e, https://feedback.bistudio.com/project/view/1/)
Just making sure I'm reporting it to the right place
yes
wiki? no
#community_wiki
oh i missed biki
Clouds problem on extreme graphics value https://feedback.bistudio.com/T175798
Can be closed as resolved: https://feedback.bistudio.com/T167317
I found a bug which lets dead bodies fall through the ground on my server
Can someone try to repro:
_Unit setdamage 1;
uisleep 5;
_Unit enableSimulationGlobal false;
uisleep 5;
_Unit enablesimulation true;
uisleep 5;
hint str (getposATL _Unit);```
When you move dead bodies with setPos commands, they desync
If you wake their PhysX up with awake before moving the corpse, it will not desync
I'm not using setPos
Look at the script above
It happens when deactivating the simulationglobal and then activating it again client sided
?
nvm, I can't read

@fast lodge Can't replicate here, body stays where it's supposed to.
Spawn this building via "Land_Radar_01_antenna_base_F" createvehicle position player and then try to do the code above. I think it has something to do if the surface is flat or not. Because inside that building the bodie dissapears 10/10 times for me and on terrain it dosent happen all the time
Are you spawning the building or using one that's already in the terrain?
I'm using one which is already on the terrain
ok, give me a map ref then.
Livonia should have it
Specific one. Terrain alt might matter.
I tried it on my terrain, the one on livonia is on the south site
You can also try to spawn the Unit inside any building and see what happens
Yea I tried 10 different buildings and it worked inside every building
Just on terrain it dosent sometimes
hmm, got a slightly weirder effect.
body's still there but it's being clipped for rendering as if it's somewhere else.
not 100% replication, more like 50/50
Did a quick test on altis and same outcome.
Video: https://www.youtube.com/watch?v=06_RHOWhy8g
Code executed (server exec).
0 spawn {
_Unit = createVehicle ["B_soldier_AR_F", getPosATL (objectFromNetId "2:1001"), [], 0, "CAN_COLLIDE"];
_Unit setdamage 1;
uisleep 5;
_Unit enableSimulationGlobal false;
uisleep 5;
_Unit enablesimulationGlobal true;
uisleep 5;
};
Original version is client exec.
PhysX bodies and localities are a pure nightmare
How come there is still no scripting command to disable physx on a dead body is beyond me (so it can be posed like in OA for example)
don’t think you can just change type of simulation at will. besides dead body physx simulation goes to sleep, there is a command to temporary wake it up, otherwise it takes minimal resources, why do you need to disable what is essentially not running?
it seems that dead bodies take a non-negligeable performance hit and disableSimulationing them saves that (but disables inventory ofc)
The issue is that when you do anything with them, they wake up - setPos*, switchMove
Disabling simulation sort of works but you won't be able to interact with inventory for example
In my case I had to start doing switchMove each frame just to stop body from doing its physx thing
Which is pretty unreliable anyway
Some kind of hard rest state
Being able to make body to sleep earlier (or later) would've helped too
There was an attempt to add body dragging to ACE but it's buggy due to setPos* or switchMove being broken on bodies.
I managed to do it in KotH, it works most of the time, sometimes first drag fails but second one succeeds
Had to come up with tons of crutches, not pretty at all
Would've been SO much easier if we could just stop the body from going back into PhysX state
ENTITY forceRest BOOL or something
In case of ACE the problem was that the body did disappear sometimes
anything further on this?
Did you change locality?
This was my original approach and it happened too (all bones teleported to [0,0,0])
I haven't tried with anything else moving near them, but in general this doesn't seem to be true. The simulation cost is high during the ragdoll process but after that it's negligible. Render cost on the other hand is just as high as a living unit.
I don't remember who complained about having dead bodies during the mission (maybe some CTI) where they had to disable simulation (until a player got near) to save some perf
I complained but that was hideObject rather than simulation.
Now parked cars on the other hand have unreasonably high simulation cost.
Old versions of Antistasi used to disable simulation on them until someone got into the car, but 2.06 broke that by removing actions from objects with disabled simulation.
So if someone could convince physX to put parked/engine-off vehicles into the sleep state then that'd be a good boost.
ah mb then, that was most likely that
Other people mentioned disabling simulation but as far as I can tell, they're wrong :P
maybe if you make a pile of corpses they never sleep.
why would hideObject save the (dedi) server though :p
like, no network traffic at all?
Oh, this is client.
You know Arma client perf is terrible, right :P
The slightly wacky part there is that simulation of parked vehicles is also expensive on the client, even when the vehicles are server-local.
I can only guess that's it's calculating all that stupid suspension animation stuff on every client, regardless of distance.
for all/any or just some?
when profiling we saw frequently 1-3ms sim cost for some carX (each)..
Didn't check. My tests have been with vanilla civ cars, which I assume are all carx
Maybe there's some older stuff in CUP. Not sure.
If this worked with both units and vehicles, it would've been perfect
what i mean is - did you get for all the same (more or less) or some low footprint (~0.1 ms) and only one/few outliers (1+ ms)
I mostly did global perf checks. Spawning a lot of stuff with vsync off and comparing frame rates with various simulation & hideobject setups.
When I've done profiling before I haven't seen any real outliers. Some vehicles were certainly more expensive than others but factors were more like 2-3 rather than 10+.
When merged this pull request will:
Changes related to corpse carrying from #9269.
Continuation of #7939.
Requires 2.14 because of this: https://feedback.bistudio.com/T168091.
IMPORTANT
If the ...
Since I can't check, are mines visible on thermal cameras (at least in the day)? If they are not, can they be added? Because it turns out that this is realistic. Especially because Arma mines cannot be buried.
I do think mines are visible from thermal
Oh, interesting approach, creating a clone of a body during drag
Having these ragdoll bugs fixed would benefit both ACE's and mine approach to the feature
Can anyone confirm that findDisplay 313 createDisplay "RscTestGrids"; freezes the game when executed in Eden? It has worked before 2.14
Seems to also haven on display 46
gotta ask about command addItemCargoGlobal for parameter count, where bug happened in 2.14 negative value when item is throwable, it was loaded to grenade weapon but no longer have it in inventory, the problem can be reproduce by this code
player addItemToUniform "MiniGrenade";
player addItemToUniform "MiniGrenade";
(uniformContainer player) addItemCargoGlobal ["MiniGrenade", -2];
_result = currentThrowable player;
// where _result should be [] but the actual is ["MiniGrenade","MiniGrenadeMuzzle",[1.00002e+007,0]]
// and Arma 3 interface still showing RGN Grenade x0
Interesting, you can throw non-existent grenade too. Same issue when you add grenades. Is there a command to select throwable grenade?
i dont think so there is a way to switch/cycleThrownItems but i might be wrong using this, this code never work for me
player selectWeapon ["Throw", "MiniGrenadeMuzzle","MiniGrenadeMuzzle"];
Yeah I tried doing selectWeapon or even action "SwitchWeapon" and neither worked
Dedmen is it intended that this ...
private _a = [["#type", "A"]];
A = createHashMapObject [_a];
B = createHashMapObject [[["#type", "B"], ["#base", A]]];
```... throws `Error #type: Type Array, expected String`? The Wiki says that `#base` can be Array or HashMap.
Creating `B` with `["#base", _a]` works.
Works for me (A3 2.14 Stable), didn't notice anything freezing. Tested with Debug Console in Eden and Eden Preview. In the preview, display 46 also worked.
Huh? Interesting.
B = createHashMapObject [["#type", "B"], ["#base", A]];
That's no valid createHashMapObject syntax (and also throws an error).
ah, yeah i see it now
you can use reload's alt syntax (Cc @gaunt depot)
tried 3 ways reload after addItemCargoGlobal, all return false
player reload [];
player reload ["",""];
player reload ["MiniGrenadeMuzzle", "MiniGrenade"];
reload ["Throw"] perhaps?
that's for changing loaded magazine in the muzzle, not for selecting muzzle, though
just tested with reload ["Throw"], it return false
like setCurrentThrowable to complement currentThrowable? I think is doable, make a ticket and tag me I’ll have a look
works fine
Yes, I tried it again and it works now.
Super weird.
Maybe switching between branches fixed it
which branch?
It froze my game like 6 times
Profiling and stable
tried both, had the freezes on both.
Always verify local files when experience weirdness
Wise words
bis_fnc_timeline_play calls bis_fnc_timeline_unpause, which doesn't exist
Steam verify for Arma seems to be partially broken, at least in the latest version. Whose problem is that?
Example: If you delete/rename addons/missions_f_beta.pbo then verify doesn't detect a problem or replace the file.
That would indicate the file is missing in SteamPipe manifest/app depot
And that would result in broken fresh installs.
Additionally steam verify completes very quickly and only reports ~200 files validated.
Addons/missions_f_beta.pbo
something might be broken for you locally or your closest CDN has some issues 
try chaning steam region.
Anywho not BI issue, the manifest look good.
that would be handy but could get somewhat complicated to do fuel usage sim
like should fuel usage be the same when idling vs when in high RPM driving uphill?
if there is fuel sim, at the minimum it should be lower usage at idle and higher usage at high rpm. then there are questions of towing/slinging/cargo/mass. a hemtt flatbed with a hunter on the back maybe uses more fuel than an unencumbered hemtt
there is room for a good, no-dependencies fuel sim mod
far easier to mod that with this
iirc fuel consumption doesnt change depending on rpm but i could be wrong
either way just flat coef would be super useful
yea i agree
It does change with RPM, I think
knowing arma it probably depends on the vehicle sim if it does or not
Yeah fuel consumption is one of my oldest wish
@solid marten enable environment windy coef appears not to be working 2.14
cant turn down the wind Tanoa
Just tried it, doesn't work. What it does is makes selected muzzle\mag reload but not select it (goes red for a bit)
Btw, are vehicle CMs also throwable? Can you select them with script? 🤔
Select as in... select single/burst?
Yeah, or different mag?
you can fix that by putting an “unlock” action on sim disabled assets to manually toggle sim
i used to toggle sim when player hit the “get in” action, but as you say they removed those actions
it does
works for me
You can only throw something if you have a muzzle for it. Theoretically adding CM muzzle to player would allow him throwing CMs but I dont think you can add that muzzle as it is vehicle weapon?
19:13:41 Error select: Type Number, expected Array,String,Config entry
19:13:41 File A3\Modules_F\Environment\HideTerrainObjects\init.sqf..., line 182
19:13:55 Error in expression <t = (_module getVariable ["#counts",0]) select _i;
Worth reporting that script error? Can be reproduced 100%
Let me know and I make a ticket with repro.
what are the repro steps?
- Place Hide Terrain Objects Module. Make sure that the area covers all object types represented by the modules attribute UI
- Place a Cover Map module (Any module other then Hide Terrain Objects module should work)
- Select both modules in the Entity List (this is important), right click and select "Attributes..."
It doesn't break if one selects them in 3D space and edits the attribute from there. 🤔
@robust quiver Is that still valid after yesterdays dev branch build?
Fixed: Scripting command deleteVehicle didn’t remove its subcontainers (stored containers like vests, uniforms, etc..) properly
Just to piggyback off the otherone made a bit ago, would it be possible to also put this one in the pipe to be added, as well?
could you make a ticket plz
@dusky wigeon yes, i see that, i will try soon :)
Made a ticket about bullshit HandleDamage fires on units, check it out: https://feedback.bistudio.com/T176124
tl;dr; Damaging units calls HandleDamage twice on shooter client, one fire (hit_index=-1) calculates getBleedingRemaining, another fire (HitHead) does seemingly nothing but kills the unit if returned value is >= 1.01
Also bleeding is completely local and units bleed only for shooter client side, not for target client side
Been like this since very first Alpha 0.50 build
👍
Subbed, gonna double-check the locality thing myself, but yeah, there are two handledamage events that are not like the others and are very obstructive.
They are definitely local to shooter, you can easily filter them out by checking if target unit is remote
Filtering them out with local units is trickier though
It doesn't always fire them twice, rarely you get only of two per hit, seen it in local several times now
What doesn't quite makes sense to me is that ACE seems to rely on handledamage events being bracketed by head-selection events, which won't happen if that first hithead is purely shooter-local.
Inb4 they get mixed up with proper events when framerate is low enough
They would anyway, on occasion. Can always get hit from two different sources in one frame.
- Wrong info
Same with 2 spawned separate projectiles
What are the numbers there?
"sanity -10" enchantment levels
Wrong info, nevermind, mixed it up with 1 projectile and then another next frame
2 projectiles in the same frame into different legs: https://pastebin.com/raw/2HhYqAGk
1 shot going through 2 legs: https://pastebin.com/raw/a7csVH0G
Looks like my info wasn't wrong, it was about 1 shot through 2 legs, not 2 shots through 2 different legs
This artificial test with 2 spawned bullets had them go super slow btw (setVelocity [-0,-0,-1], https://pastebin.com/raw/2HhYqAGk)
Same test but with larger speed ([0,0,-100]): https://pastebin.com/raw/y2i9hc62
So despite overall damage getting 0.00100436 damage for super slow bullets, the fake event handler for bleeding calculation doesn't fire
Either way all these EH fires are bullshit and don't belong in HandleDamage
@sinful kettle The new syntax for steeringComponent. Is the argument global or local?
local
Alright, thanks!
Another question:
Also, currently this new AI driving works only for wheeled vehicles, otherwise known as Cars (from klamacz)
Yet the example is myTank useAISteeringComponent false; where track is usually a tracked vehicle.
Should we clarify this?
yeah it does work for tanks too
also previously the useAISteeringComponent false syntax would only cause tanks to not use that feature. I've changed it to affect cars too
interestingly, the AI drove better without that component (at least based on some limited tests) so I figured this command is worth adding.
ofc some newer game features rely on it so it could also break some things 
I remember when this was first showcased on dev branch. It was a huge improvement. Maybe something got broken again between 2016 and today 
Maybe something got broken again
yeah I guess so
it had some improvements (ie cars taking turns better), yet also made other things worse (ie crossings) and broke things (3+ tanks in a group)
BI later applied configs tweaks from CUP - not sure how much affect these had either way
unfortunately too often BI revised systems in A3, published them in WIP state wanting to tweak/revise them down the line, but then didnt happen
there is one for cars and another for tanks - that comment is probably from early on of the pathfinding revision
Yeah I was under the impression only cars got the new PID stuff
they showcased it in sterile VR terrain, it never worked in MP
Enginelimitations.jpg
esp since there is a "forced" moden ow
Ded / Leo / KK, can one of you look into that?
Make ticket
I don't see much use for it tho
You typically just want to set it to what you want. I can't think of any case where you need a getter
really just for when changing/modifying it. i use some UI for adjusting flyinheight and need to set a variable on the vehicle to manage it
if (_selected isNotEqualTo []) then {
private _entity = objNull;
private _currentFlyInHeight = 0;
private _step = [5,50] select (uiNamespace getVariable ['QS_uiaction_turbo',FALSE]);
private _altitude = 0;
private _min = 0;
private _max = 12000;
{
_entity = _x;
_currentFlyInHeight = _entity getVariable ['QS_air_flyInHeight',((getPos _entity) # 2)];
_altitude = round (_min max ((round (_currentFlyInHeight / _step) * _step) + _step) min _max);
_entity flyInHeightASL [1, 1, 1];
_entity flyInHeight [_altitude,TRUE];
_entity setVariable ['QS_air_flyInHeight',_altitude,FALSE];
50 cutText [format ['%1 m',_altitude],'PLAIN DOWN',0.2];
} forEach _selected;
};```
in a ui/key binding
any plan adding command such as
// Syntax
getUserMFDBrightness index;
index setUserMFDBrightness value;
// Example
getUserMFDBrightness 12; // Return value 0 - 1
12 setUserMFDBrightness 0.7; // Set brightness to 0.5
@sinful kettle If I was to ask for a new toy, it would be for an "attach/detach" object/mission event handler
and a simulation state object event handler, since the engine can toggle sim state
the attach/detach since many mods make use of the command and detection is useful for detecting the change
AI related toy, would be a toggle for agent/unit, or an "attackTargetChanged" unit/group event handler
toggle for agent/unit, being a way to convert a createAgent ["b_soldier_f" into a grp createUnit ["b_soldier_f" and back again. would allow for larger battles/warfare with better control of performance
Are agents faster than just disabling all the AI features?
I think what it meant is that it should be an actual hashmap, not hashmap object
anyway the hashmap one had an infinite loop bug too 😅 (fixed now)
the reason hashmap object doesn't work is that its type gets converted into an array, hence the error
e.g. this is what you get if you print B:
[["#type",["B","A"]]]
I can fix it to work with hashmap objects too but only if @untold sky is ok with it
You mean the value that the #type key maps to?
While you've got that code open, do you think we can get a magic variable _super that points to the base HashMapObject in HMO methods (similar to _self)?
Not necessarily. I tested something like this the other day ...
private _a = [["#type", "A"], ["myMethod", { systemChat "A"; }]];
A = createHashMapObject [_a];
B = createHashMapObject [[["#type", "B"], ["#base", _a], ["super", A], ["myMethod", { (_self get "super") call ["myMethod"]; systemChat "B"; }]]];
B call ["myMethod"]; //Two messages "A" and "B" in system chat.
```... and it works. But `_self get "super"` is inconvenient.
that's different from what you're asking
you say you want to use ["#base", A] and have a magic var _super
not ["super", A]
Aye, that. So that I don't have to manually set ["super", A] and call _self get "super" to access it.
^
when you use ["#base", A], A gets merged into B. it's not stored separetely
and game doesn't store "#base" itself into the HMO. you can print B to see what I mean
But that could be changed 😅
Additional limitation: _super actually has to be a HMO, not an array 
well that's up to Dedmen to decide 
How 'bout this request by some guy named Leopard20? 😂
And that doesn't even cover bones used in rtm but not animations, right?
https://feedback.bistudio.com/T150103
"but not animations"? wdym?
it doesn't ask for rtm anims if that's what you mean
they're different ofc
I mean I'd like to get skeletonBones[], instead of filtering from selectionNames "memory".
I wonder if there is a way to make explosionEffect have proper source and instigator by whoever killed\damaged the vehicle?
Example:
You shoot a heli with a tank:
- HandleDamage: source=tank, instigator=player
- Killed: source=tank, instigator=player
- explosionEffect (FuelExplosion): source=player, instigator=null
Have a repro ready but want to make sure anybody cares to look at this
More visual explanation in hint, heli explosion damages the heli itself and stuff around it with following parents
Made a ticket: https://feedback.bistudio.com/T176206
where does this explosionsEffect appear, engine or scripted?
Engine-driven it seems
so you want instigator to double the source?
No, have it the same was as Killed has it. Source - vehicle, instigator - unit
Or unit-unit if it was killed by a unit and not a vehicle
so basically for instigator to be the same as the source?
as in your second example
Yes, if no vehicle was involved
Basically mirror source and instigator of HandleDamage and Killed EHs
mirror as in flip?
No, as in same
When you do a vehicle kill, explosionEffect takes instigator of Killed event handler and uses it as source
Instead of having both instigator and source the same
I guess it is not difficult but the question is will it cause problems with the existing scripts?
Unlikely, I doubt there are special cases just for these explosions in scripts
HandleDamage will get proper source and instigator for vehicle explosions, doubt anything will break
so what is in the picture you posted?
the actual or desired?
Actual, repro result
but the source is instigator in it
Player in Tank shoots Heli:
Killed/HandleDamage:
source = Tank
instigator = Player
HandleDamage by explosions:
source = Player
instigator = none
No, check the end of hint, it has classname there
i’m confused, could you give me actual vs desired, concrete example
str vehicle returns commander unit string so it appears the same, but both point to different entities - vehicle and unit pulling the trigger
Bottom should have the same as the top
Issue is that when engine spawns vehicle explosion projectile it drops one of parents that caused the description
if anything the source should be the vehicle that exploded, and the instigator the one that caused the explosion
Well vehicle explosions were always considered to be result of whoever destroyed the vehicle
I think plenty of stuff relies on this logic
I think when you crash a helicopter units inside die because of that explosion and so pilot is considered a teamkiller for the crew
i see
so what helicopter crash looks like in actuality vs desired
what is the source and the instigator
Right now source will be pilot and instigator is null
let me double check
I actually remember it being somewhat random
and what it should be
Sometimes crew dies by itself (collision damage probably), sometimes pilot teamkills them
All during the crash
What happens right now during the crash:
- Heli
HandleDamage:
source = Pilot
instigator = none - Pilot
HandleDamage:
source = Pilot
instigator = none - Crew
HandleDamage:
source = Pilot
instigator = none - Explosion Parents (seen though HitPart and HandleDamage on a tank nearby):
source = none
instigator = none
I guess engine assigns explosion source from instigator so its null in this case?
imo crash should be source the helicopter, instigator the pilot