#arma3_feedback_tracker
1 messages Β· Page 41 of 1
dunno, try it?
Well what is written on the FT ticket for it?
wasn't the ticket private? (like all other "exploit" tickets)
nope, iirc someone posted it publicly and then it was hidden. but not 100% sure if it was this bug. (it was a duplication bug tho, that's for sure)
Please make a new ticket :sadcat:
you deleted it?! now I have to type all that text! 
I copied it before, in case you needed it
B_GMG_01_A_F is autonomus variant and you shouldn't be able to get inside it
it is? 
oh, you're right... 
I always spawn all vehicles using Alt+Click. didn't even notice it doesn't have any gunner...
btw somebody plz close this: https://feedback.bistudio.com/T163479 
https://feedback.bistudio.com/T163481 @sinful kettle some steps are missing? cannot repro
I don't think so 
let me try again
no it's correct 
let me try without mods...
I can also repro in vanilla. let me upload a video there
added the repro video
what am I looking at? When you add that backpack it is filled with config values
INCLUDE INHERITED ENTRIES: false
SHOW CLASSES ONLY: false
UNLOCALIZED TEXT: true
CONFIG PATH: bin\config.bin/CfgVehicles/B_AssaultPack_mcamo_Ammo
SOURCE ADD-ON(S): A3_Weapons_F_Ammoboxes
*/
class B_AssaultPack_mcamo_Ammo: B_Carryall_mcamo
{
author = $STR_A3_BOHEMIA_INTERACTIVE;
_generalMacro = "B_AssaultPack_mcamo_Ammo";
scope = 1;
class TransportMagazines
{
class _xx_30Rnd_65x39_caseless_mag
{
magazine = "30Rnd_65x39_caseless_mag";
count = 6;
};
class _xx_100Rnd_65x39_caseless_mag
{
magazine = "100Rnd_65x39_caseless_mag";
count = 1;
};
class _xx_NLAW_F
{
magazine = "NLAW_F";
count = 1;
};
class _xx_HandGrenade
{
magazine = "HandGrenade";
count = 2;
};
class _xx_MiniGrenade
{
magazine = "MiniGrenade";
count = 2;
};
class _xx_1Rnd_HE_Grenade_shell
{
magazine = "1Rnd_HE_Grenade_shell";
count = 3;
};
class _xx_3Rnd_HE_Grenade_shell
{
magazine = "3Rnd_HE_Grenade_shell";
count = 1;
};
class _xx_10Rnd_338_Mag
{
magazine = "10Rnd_338_Mag";
count = 2;
};
class _xx_20Rnd_762x51_Mag
{
magazine = "20Rnd_762x51_Mag";
count = 2;
};
};
class TransportItems
{
class _xx_FirstAidKit
{
name = "FirstAidKit";
count = 4;
};
};
};
oh 
I thought when you add a new backpack using that command it is supposed to be empty
tbh. I was not aware of that myself
It's not supposed to be empty.
Get base variant if you want empty one. These are for AI/created units.
I added a note on the biki
Dunno, it was long time ago when I could just confirm immediately if you needed to empty added backpack if you needed empty backpack. I expect there are empty variants
yeah thanks. I see now 
double facepalm in a row 
please close this ticket too 
Leopard having a good day
the first 3 tickets on FT rn are spams
not anymore
nuked, thx
I would really like to see a filter for: any message with an embedded URL from a user < 1 day old bans the person (direct URLs might be OK)
it would cover 75% of the cases
There would have to be a clear warning about that so people don't get banned for trying to link videos of bugs or Steam screenshots or whatever
yes of course
I mean embedded URL as "link shown as normal text" not an https://β¦ thing
since the support is dropped, maybe I can fiddle something at one point who knows
or it will be replaced by something else perhaps
Spam comment: https://feedback.bistudio.com/T126974#2283830
oooh, a sneaky one, thanks! dealt with π
first 2 tickets on FT are spam
no takers?
done
removed, thx
Hey, just found out a bug/feature that CfgAISkill >> aimingSpeed[] changes the speed of turrets movement on a player.
For example:
class CfgAISkill { aimingSpeed[] = {0,0,1,0.1}; };
Turrets turn very slowly and aimingSpeed[] = {0,0,1,0}; completely halts the turrets and locks the camera too after moving once or twice. Static weapons and turrets on vehicles.
Don't know if it's intended but I think it's unknown.
how are you testing it? if you're using remote control from zeus the AI subskills will apply to the unit still
Sitting fresh on a static weapon, just my addon for changing skill.
does it do it if you use keyboard controls for the turret turning, believe its arrow keys by default
reason i ask is that the keyboard controls have always been faster to me so wonder if it might just be mouse thats affected
Tested, both get slower.
Does only the config value have this effect, or can it be caused (or un-caused) by setSkill?
nuked, thx
it seems displayAddEventHandler remains persistent if you use the restart button (at least while in Eden)
this seems not correct, or is it?
can someone confirm that?
another strange problem
Tried to create group while group limit is reached on side Game Logic
β₯ Context: [] L31 (A3\functions_f\Modules\fn_initModules.sqf)
[] L19 (A3\functions_f\Modules\fn_initModules.sqf)
[] L22 (A3\functions_f\Modules\fn_initModules.sqf)
[] L23 (A3\functions_f\Modules\fn_initModules.sqf)```
p:\a3\functions_f\initfunctions.sqf
588 [] call bis_fnc_initModules;
p:\a3\functions_f\modules\fn_modulehandle.fsm
217 init = /*%FSM<STATEINIT""">*/"_modulesActivate spawn bis_fnc_initmodules;" \n```
these are the only two instances i can find the function called
initFunction.sqf calls it only when
//--- Mission only
if (_recompile in [3,5]) then {```
as the problem happens mid mission, no idea if recompile can happen at that point
initModules has this:
//--- Create groups for module categories
_sideLogic = sidelogic call bis_fnc_sideID;
{
_category = configname _x;
_side = getnumber (_x >> "side");
if (_side == _sideLogic) then {
_groupVar = "bis_fnc_initModules_" + _category;
_group = grpnull;
if (isserver) then {
_group = creategroup sidelogic;
missionnamespace setvariable [_groupVar,_group];
publicvariable _groupVar;
} else {
_group = missionnamespace getvariable [_groupVar,grpnull]
};
_group setgroupid [gettext (configfile >> "cfgfactionclasses" >> _category >> "displayname")];
};
} foreach ((configfile >> "cfgfactionclasses") call bis_fnc_returnchildren);```
so each call should create new logics for all cfgfactionclasses with sideLogic - no reuse or caching
this looks like a design problem/weakness
nuked, thx
it does and it wonβt be fixed because of backward compatibility as the issue is wider than restart button in 3den
you should always use a guard for event handlers anyway, to make sure you never dupe them
or do setVariable vars get reset but EHs don't? 
Depends on what display.
Display eventhandlers go away when the display goes away
if some handler stays even though the display gets closed thats a huge issue, but I don't think thats happening?
Are you adding handlers to the eden display while in eden preview? (The eden display is still open in background during preview)
(findDisplay 46) displayAddEventHandler ["KeyDown",
executed from mission's init.sqf with waitUntil {!isNull (findDisplay 46)};
i will confirm tomorrow and post on A3 FT
I expect that even if you didn't have the problem after restart, you'd still have the problem after loading a save 
so you should just do this
sure. easy to do - just wasnt aware of that behavior
[BUG] displayAddEventHandler [KeyDown?] remains available after mission restart
https://feedback.bistudio.com/T163625
displayAddEventHandler seems not stored in savegames - is that intentional too?
so one has to use Loaded EH to re-init them?
ref:
https://community.bistudio.com/wiki/Arma_3:_Mission_Event_Handlers#Loaded
https://feedback.bistudio.com/T124597
https://forums.bohemia.net/forums/topic/203935-addmissioneventhandler-loaded-event-code-not-happening-after-loading/
Hi guys, the old story - the joys of surviving a savegame. There are some tricks with diag_ticktime, but it takes more time to understand... cannot afford. Devbranch here. Ive got the below code in an init.sqf, very much at start of a mods code. I have checked the variable handle1, it is existing...
has also a funny quote at the end π
On 12/8/2019 at 10:16 AM, killzone_kid said:
I think there is a bug with UI event handlers that do not get removed from displayDedmen:
Would make me happy if there was a ticket with repro :3
Displays never get serialized (which is also why you get serialization warnings when you store a ref to them in missionNamespace)
I suppose this is an easy (and very useful) feature to add:
https://feedback.bistudio.com/T163473
just add two other surf properties, which I'm pretty sure are already known when you know the surface
maybe even return the whole surface (3 pos) 
selection probably maybe, surface material probably, verticies no πΎ
otherwise model hacking made easy 101
since when do the GEOM/VIEW/FIRE/etc. lods count as "model hacking"?
nobody rips models from those. they are simulation lods and do not represent the model
they have no texture, and they're modified and low-poly...
Dedmen, while you are working in the ammunition event handling area, is there hope for https://feedback.bistudio.com/T157204 ?
And also on the ammunition event handling area... I was told a long time ago that the fact that ammunitions (bullets) hitting units on limbs also damage other adjacent parts, for example a chest shot would also damage the head a little bit or the other way arround? was due to a "design decision" so that limb damage can also be eventually deadly?... I was wondering if it would perhaps be a better and more performance friendly method to keep that mechanic only when said "limb" or "first hit" part is 100% damaged and not all the time like it currently is... what do you think about that?
yes thats on my list now
nukeded, thx π
boom
https://feedback.bistudio.com/T148575 last comments are spam too. Maybe add an extra captcha upon registration or email verification?
cleared up
regarding the captcha, I wish
I don't know much about Phabricator besides that the official support is ded since last year, but I bet there must be a community plugin or anything to make the additional step
Hi, after the 2.08 Update on my mission the BIS_fnc_taskCreate doesn't work anymore.
it still works, go to #arma3_troubleshooting or #arma3_scripting and post your whole script
Very happy to finally have disableBrakes to complete some projects. It does work fine as-is, but there are a couple of peculiarities about it that would be cool to smooth out: https://feedback.bistudio.com/T163652
https://feedback.bistudio.com/T126282
Is there any chance this could get looked into, seeing as its been years and not much dev attention put into it? At least to know if this is something difficult to solve.
...and this already has a spam comment on it \o/
"If the autobrakes are disabled while a vehicle is currently autobraked, the autobrakes don't actually release until a force is applied. This is noticeable when parked on steep hills - the vehicle doesn't start to roll down the hill unless you give it a push or tap the accelerator"
I tried to solve that for towing, twice actually and failed both times.
"The mission also has a tow system included that lets you disable brakes on vehicles you're towing."
just use the new towing system if you want to tow
It force turns the wheels for you the overcome that stickyness
Unfortunate but not a huge deal I suppose.
My system does use setTowParent, and it works without disableBrakes as well. However, practical testing indicates that autobrakes at low speed or when stopped are still active on the towed vehicle, meaning sometimes it can be excessively hard to get moving, and if you're towing at low speed the brakes can sometimes engage. So when disableBrakes arrived, I just added it to my system as a QoL thing since it doesn't have much overhead.
Also because it will be hilarious when people hard brake at 70kph without re-engaging the towed vic's brakes and get sent to space
I hope here is the right place to put my bug reports for some bugs that annoy me for a long time. Maybe there redundant but maybe it is the good time to address that now again π
(1) Weapon-Optic transfer from dead body's suit freeze Inventory
- start VR
- create OPFOR Soldier with Katiba (player)
- create Viper
- Start Scenario
- kill Viper and open his Inventory per key "I"
- take your own ACO and put it in your suit
- switch your suit by right klick on Viper suit
- open your old suit (witch is laying on the ground) by double klick
- try to mount your old ACO by right klick or Drag and Drop on the Slot of your weapon -> doesn't work (and all following Inventory transfers to your soldier. Only close dead soldier inventory and reopen and then drag and drop to your inventory first is working).
it's not, but close enough π π
you can create Arma 3 Feedback Tracker (FT) tickets (if they do not exist already) following this link:
https://feedback.bistudio.com/maniphest/task/edit/form/3/
it is regularly checked so no worries π
if it is close enough, i continue positng here, right?π
n o π¨ π π
inventory issues are fun
nnnukeded, thx
https://feedback.bistudio.com/T163630 I never before saw such a well made test script that literally lists all the bugged things.
Thats great
don't praise the good bug reports or there will be even more!!1! 
Now I copy-paste this script in every of mine bug reports to become good feedback...
lol I take it back that script doesn't work
it checks if projectile is still alive one frame after it was triggered (btw you can use diag_frameNo instead) but, most projectiles trigger only on next cycle.
And some projectiles just stay alive after getting triggered, like smoke grenades or flares
Even with mines it doesn't work. A simple explosive charges does a "beep beep beep.. boom" when triggered
that takes like half a second
player addEventHandler ["FiredMan", {
params ["_unit", "_weapon", "_muzzle", "_mode", "_ammo", "_magazine", "_projectile", "_gunner"];
if (isNull _projectile) exitWith {};
systemChat str [_thisEvent, _weapon, _projectile];
diag_log [_thisEvent, _weapon, _projectile];
_projectile spawn {
sleep 0.2;
systemChat "trigger!";
triggerAmmo _this;
}
}];
My test is more fun, when UGL grenades, RPG rockets and hand grenades and mines just explode infront of your face
inputAction doesn't seem to work with user actions (modded keybinding). it always returns 0
is that a bug?
I tested that
I'm also using it in RHS in many places and so far no one reported that. Are you using some specific key combination?
well it's Right Ctrl + \ and Left Ctrl + \
let me see if it's related to the keys...
nope 
I test it like this btw:
addUserActionEventHandler ["DBUG_OpenConsole", "Activate", {
systemChat str inputAction "DBUG_OpenConsole";
}];
works for vanilla actions tho 
also this opens a new display. I wonder if that's why
well anyway instead I'm gonna make a feature request: bool = [DIK_CODE, shift, ctrl, alt] canActivateAction "ActionName"
to override actions using display keyDown 
or even better:
getKeyString [DIK_CODE, shift, ctrl, alt]
then using a hashmap to match it against actionKeys ["action"] for fast key overriding
ah, I'm using in on frame loops
I checked in debug console watch fields
It might be that specifically inside that handler its not yet set π€
I think this might be why
I'll test it
btw: I was wondering if it would be possible to expand setVehicleCargo and add optional param to set cargoAlignment
what alignment, example values?
left/center/right?
Can't, these need to stay the same for every object in cargo due to how cargo thing was made
weird... I just tested another action and it works 
I'll keep looking
ah no I create a dialog 
it doesn't work in dialogs it seems
well repro is simple:
addUserActionEventHandler ["ReloadMagazine", "Activate", {
createDialog "RscDisplayEmpty";
systemChat str inputAction "ReloadMagazine";
}];
shows 0. if you remove the createDialog line or replace it with findDisplay 46 createDisplay "RscDisplayEmpty" it works 
should I make a ticket so you can take a look?
btw may I suggest an optimization for nearestTerrainObjects?
when radius is large, I think the command has to do many reallocations because it doesn't know beforehand how many objects it'll return right?
so this means the command is too slow for people who want all terrain objects.
if that's the case this simple check will significantly speed up the command for those people, since I think the game already knows how many terrain objects there are and you can just allocate once and return them all:
if (is2D && radius >= worldSize / sqrt(2) && center.x == worldSize/2 && center.y == worldSize/2) {
//return all objects
}
this check can also be made to work with all centers too, as long as the circle encompasses the square (terrain), but that's a bit slower (tho this one is not 100% accurate either since it only considers the upper bound, but it's much faster than the actual equation)
if (is2D && radius >= worldSize / sqrt(2) && distance2d(center, {worldSize/2, worldSize/2}) <= radius - worldSize / sqrt(2)) {
//return all objects
}
what do you think? is my assumption even correct?
High Command is broken on Profiling v3, was working around prior 2.08.
hcAllGroups player report the HC groups of player at start but after clicking HC button or hcShowBar true, it returns empty array and bar never shows up.
reallocations are rarely a problem
The main problem is more likely that its just single threaded.
Iterating through all objects, doing type checks and sitance checks for each of them, a somewhat badly written check if the mapType matches.. mapType matching even being checked if you didn't provide any and don't even want to check for it..
And the sorting by distance at the end
Ok but anyway my point is what if someone wants all objects (types is empty array)? Can it be optimized for that case? So that it does no type checking or anything, just returns all terrain objects directly...
(you may prefer a (getA|a)llTerrainObjects?)
Well many existing scripts already use this command so they can benefit too... 
tho I don't really mind as long as it works...
but I think making a few changes to an existing command is easier than implemeting a new one too
oh no, legacy code support π
yeah we can... But I'm not sure, can you spawn objects that are returned by that command, can you spawn them outside of terrain?
I believe a TerrainObject in that case is something baked with the terrain, so always within the world's boundaries and non-spawnable?
it should be a quite "static" array ("static" not in dev term hey π)
Is there really a common use case where you want ALL terrain objects?
And even then not sure if we can preallocate without iterating through the objects twice
I don't know if we have a "number of objects" value
well, isn't the terrain objects array (that is parsed through in nearestTerrainObjects) that?
(β¦unless everything is mixed together)
its not an array its a QuadTree
spawn them? I mean baked objects into the terrain btw which afaik is what that command returns already...
afaik they're always within the terrain bounds, and their position is fixed (when you do setpos on them they don't move)
I don't know about common use cases but I've seen a few (not so good) scripts that do need all terrain objects (tho not sure if they wanted all or just certain types)
I personally want it for terrain caching and I need all... 
Ah Lou already said that... 
Anyway for now I just wanted to see if it's possible and if the perf improvement is worth it (like a few times faster) π
If you think my use case isn't good enough we'll discuss this again when I'm done with what I'm doing π
removing all distance and type checks will make it a few times faster yes
but optimizing for such a rare special case?
Might make more sense to use multithreading to optimize all cases
Hi Guys I have a dedicated server that is restarting just after restart with Exception code: C0000005 ACCESS_VIOLATION at 00000000
Exception code: C0000005 ACCESS_VIOLATION at 00000000
Exception code: C0000005 ACCESS_VIOLATION at 00000000
Exception code: C0000005 ACCESS_VIOLATION at 00A2EB62
Exception code: C0000005 ACCESS_VIOLATION at 00000000
Exception code: C0000005 ACCESS_VIOLATION at 00000000
Exception code: C0000005 ACCESS_VIOLATION at 00000000
cpu: Intel(R) Xeon(R) CPU X5687 @ 3.60GHz
graphics: No
resolution: 160x120x32 ANY IDEAS WHAT IS CAUSING IT
this doesn't say anything about the problem. Plz make a ticket on feedback tracker
http://feedback.arma3.com/
including the rpt and crash files (if any)
Arma generates a .rpt log file each time it's run, which contains a lot of information like the loaded mods, or any errors that appear, this log file can be very useful for troubleshooting problems.
To get to your RPT files press Windows+R and enter %localappdata%/Arma 3
Additionally see the wiki page for more info: https://community.bistudio.com/wiki/Crash_Files
To share an rpt log here, please use a website like https://sqfbin.com/ to upload the full log, that way the people helping you can take a look at it and try to figure out the problem you're having together with you.
Note: RPT logs can hold personal information relevant to your system, the game or others.
Are you running 32bit? if yes, don't
I've "surfed" a little bit through the FT and I am found some tickets that are open but abandoned for years. For my feeling it would be good to have some tidy place (The Mr. Monk in me want get rid of it π ). So would ask for someone that has the power to close some of tickets especially:
-
bug is well described but aren't reproducible anymore
example:
https://feedback.bistudio.com/T58985 -
are not well described and cannot reproduced (e.g. only text is "My game crashed")
example:
https://feedback.bistudio.com/T59282
-> For that I would to suggest:
(1) Try to reproduce the bug. If not reprocable, go to point (2)
(2) To write to the subscribers and ask if the bug is still present and ask for new information. If no one answer after a certain time with negative answer (bug is still present).
Then ticket can be closed.
I think there are some people who do that already, but there are, like. a lot of tickets, and a lot of them are...junk. it's taking a long time to go through it all.
Oh, found someone with that ATOC bug a couple of weeks ago.
Or possibly just something that looked a lot like it.
Yeah in general all game crashes older than 2020 can just be closed because even if they have proper data they are just not relevant anymore
But who wants to spent hours closing ticket that noone would ever look at
maybe someone who spill out endorphin from closing bug tickets
if you list them, I'll do it
not for the thrill, but for Monk-like cleanliness (ensuring a faster search for other users)
here?
sure, or in google doc if it's too much
Too bad you can't do a "Category:Game Crash" search in Phabricator
Mh I think this will be a long term endeavour. So here might be enough
Actually, doesn't this search just do the job?
https://feedback.bistudio.com/search/query/evG7S1BIOm3Z/#R
And basically anything below T155946 should be older than 2022, right? (I'm assuming the results are ordered by last comment date)
In case the query becomes invalid in the next five minutes: I literally just searched for "crash", tags ""Arma 3", status "Open"
yes, such as terrain object replacement via sqf
"common" not "two mods do it and run it once per mission"
well that "once per mission" actually takes 10+ seconds on some maps 
Even if I make it 10 times faster.
How much time do you save? a tenth of a second over a several hour long mission?
okey then a second over a several hour long mission.
It's not something I use personally so I can't speak to how much of a real-world benefit it would be, but it seems like it could make sense for getAllHitPointsDamage to have an option to return a hashmap rather than that awkward-looking array of arrays
Same for namedProperties
well the hit point names and their order don't change so you can just make the hashmap yourself once and reuse 
also there aren't that many hit points, so find isn't slow to use, so a hashmap doesn't give much benefit here
you can already convert that to hashmap yourself
using createHashmapFromArray
looks like this is possible to be converted into a hashmap directly as well
the new syntax for the insert command supports this: hashMap insert [splitArray, [keysAndValues]]
Unusually blatant spam comment: https://feedback.bistudio.com/T163723#2287745
well, thanks for making it obvious I guess
thx, nuked
nuked by KK
Can we get a command that returns the ID of a terrain object directly please? So that we can use it easily with nearestObject without any awkward workaround?
https://feedback.bistudio.com/T163749
but it returns 3 arrays, how do you figure hashmap would look?
it's not really needed anyway
we have getHit, getHitIndex, etc.
you can use the hitpoint name, selection name or its index directly
and if someone needs it for something else they can just build their own hashmap like I mentioned
i know, i am curious how you pack 3 arrays into hash map
Well, the arrays are related to each other, right? So the key would be the hitpoint name, and the 2 values would be the damage and...whatever the other thing was
have you optimized this recently (or something else related to this)?
it takes a really short time for me now 
I'm pretty sure a few versions ago it took ~9-10 seconds on Altis. Now it's ~2-3 seconds (which was on Stratis before) 
thanks, dealt with
can you post comment link for comments please? otherwise I almost always bug on the ticket itself before checking further π
I typically don't even open the ticket... 
I just rclick on the ticket in the main page and copy its link π
no
hitPointName -> [selectionName, damage]
But at that point, not sure if it makes sense.
Creating the hashmap might be more expensive than just doing the array lookup, as these arrays are very small
I literally even posted the results here: (~6 months ago)
#dev_rc_branch message
I'm confused. but I'm also happy π
maybe I have a mod or something that's doing nearestTerrainObjects without my permission?! 
no also faster in vanilla... 
nope just checked to make sure, no change.
mhhh... although
could potentially be related to the moving instead of copying change, but only if your thing returns lots of things.
And also there is still a bad thing in that command specifically where it copies twice instead of moving once
but only if your thing returns lots of things
yeah ~1.8 M objs (everything on Altis)
I'm pushing a improvement tomorrow, but probably not gonna do much
boom
https://feedback.bistudio.com/T163756 not spam for a change but a suggestion which I think could have been used a lot by now.
Ugh
The inventory system is less than nice to fiddle with
I wanted to do that a couple times, thought about how much effort it'd be, and then just didn't do it
You can do it! You will become immortal just for fixing this one bug.
the company will make a pizza party for you*
The inventory system needs you. There are more than one thing in it that are clunky. Maybe you stumble over it on your way to fix this one bug. π
Maybe now it is not the time as there are more meaningful optimisations and critical fixes to be made but in the future the Inventory system really needs some love that is for sure.
Maybe if we wait a while Leopard will be promoted to honorary dev and he could take a look at it. He is quite good at UI stuff and ded could really use a couple more hands. Our beloved cat is greatly underused almost exclusively as an ultra competent and technical tester but he could be so much more with some BI love πͺ€
what bug? suggestion for the event handler?
what are the stacks?
If you take two grenades of the same, there will be stacked in the inventory in one slot but with a number.
when you click on one stack by right click as long as the stack is transfered to another place, the stack doesn`t stands still in the order of the items in your inventory. It "jumps" through your inventory (change the position). You have to check after every right click if the stack is already under your mouse cursor or not. When not, you transfer a unwanted item.
huh? never noticed, could you make a video?
I assume it's because it's re-sorting the inventory after each click, but yeah, it's quite annoying.
Highest-weight piles go to the top or something.
not at moment. but the steps to reproduce are really easy
ok gimmie
I think it is related to the order of taking the items
i need some repro
I know it is mybe not a big thing in SP but MP with many loot like "Dynamic Bulwakrs" or "Escape from..." were you change whole suits or backpack and transfer the content back. It kills you.
in the Tracker-Item
start VR
create OPFOR Soldier with Katiba (player)
create Ifrit
start scenario and open inventory of ifrit
select your inventory and put random different items in vest (for example many smoke grenades but in random order) to create different stacks of items
then put the stacks back into the ifrit by right clicking on the stacks in your inventory (put back single items)
-> position/sort of items/stacks in your inventory change (pseudo?)random by right clicking
no canβt repro, slots donβt move, only disappear when you take all items then stacking changes but not the order
i need video
- make your vest empty (into ifrit)
yes
- take in random order only some grenades to create maybe 4 stacks
- put stacks back one by one per right click into ifrit's inventory as long stack is gone
-> position of items change in your vest sporadic
The worse inventory glitch in Escape is where you swap your vest with one from a corpse and then can't pick up or open the one on the ground.
For me this is the second one π
game-wrecker, that one :P
I've had so many inventory bugs that only perhaps a full rewrite would probably fix them all
the ticket is already assigned to @untold sky
I swaped a backpack inside a vehicle for my own backpack and my old backpack got deleted! Lost everything in it....
oh yeah, that's another one. Haven't hit that much myself though.
And don't even get me started in multiplayer, plenty of items duplicate and/or aren't pick able while they should...
A couple months ago there was an Ai body with a rifle, all the player in our MP session could pick up the rifle and it would not delete the original one so it would duplicate the rifle on every player pick... and then players couldn't drop/unequip the rifle anymore.... suddenly there were like 20 of the same rifle on the floor and on the player's inventory and they couldn't drop it! π€£ π (Only for me, the host, was that Ai's inventory working properly)
is there a ticket with repro, if not, you are wasting your time complaining. For the 100s time, no repro - no fix
You are right, just doing some catharsis here because at this point these bugs have become funny rather than annoying... will do a ticket+repo when I get those again...
workaround is to close the inventory and reopen it or use drag and drop instead of right click. that's help most of the time. But yeah it is annoying, too.
I feel like more recently that's less likely to work.
Hell, it might even be common enough for a repro these days.
now I wonder, would you be happy with sorting by count of items?
They would still change order then π€
I guess name is better
I think best would be name + some kind of item type. So grenades close to grenades. Mags to mags etc
By group # maybe and not item # ?
OMG OMG the Inventory CDLC finally happening? π
nuked by a JointOps w/ KK
The vanilla Hatchback, specifically, is immune to towing for some reason https://feedback.bistudio.com/T163769
do not tamper with anti-theft devices!
I didn't know they fitted portable black holes as anti-theft devices :U
I'd be interested to learn what actually caused this once it's fixed. I wouldn't have expected the Hatchback of all things to have special properties that make it behave differently in this respect. (Especially properties that aren't inherited by the Hatchback Sport)
I wouldn't have expected that either
Enemy AI, will always know the location of the player who put the explosive/mine
https://feedback.bistudio.com/T163771
this sounds very much like a fix Ondra did late in A2/OA - maybe it wasnt merged?
@waxen tundra do you still have that on one of your lists?
Dont know if this is the right place for this but it seems to be a general bug / feature which is pretty annoying. I ask myself if someone can confirm it. I played the OA SP scenario Jackal where you have to blow up enemy tanks and stuff. After having placed all satchels undetected I went back to...
[88850] Fixed: AI was sometimes able to reveal a unit which has detonated a satchel
Torture Dedmen idea: screenToWorld but it also takes distance from camera
1000 screenToWorld getMousePosition```This returns 1000m from the camera
what does it even mean?
Well, I'm suck at English /s
Instead of the surface pos xy as the regular syntax does, this returns position xyz. My usecase is to detect object at where my mouse is
positionCameraToWorld can recreate the thing, but it's not really reliable as far as I made
Instead of the surface pos xy as the regular syntax does,
screenToWorld only intersects with terrain tho
That's the point, so it's not reliable source to make a lineIntersectsSurfaces execution when the camera looks the sky
oh so that's what you mean 
You know my brain is not really working well
Alternative non-existent so far syntax solves it
well if you know the screen distance to camera and fov you can generate it manually π
But not really sure if it does for literally everyone else
Since everyone uses different settings and screen
wasn't there a command that returned FOV and aspect ratio? 
getResolution?
ah right π
@alpine tulip
fnc_screenToWorld = {
params ["_screenPos"];
_screenPos = _screenPos vectorDiff [0.5, 0.5];
_res = getResolution;
_m = _res#4;
_aspect = _res#7;
positionCameraToWorld (vectorNormalized [
_screenPos#0 * tan (45/2) * _m * _aspect,
-(_screenPos#1 * tan (45/2) * _m),
1
] vectorMultiply 1000);
};
onEachFrame {
_p1 = positionCameraToWorld [0,0,0];
_p2 = [getMousePosition] call fnc_screenToWorld;
drawLine3D [_p1, _p2, [1,0,0,1]];
drawIcon3D ["", [1,0,0,1], _p2, 0, 0, 0, "0"]
};
works on my system at least π
it's a bit slow tho... 
was always reliable when i used it
he means for screenToWorld "recreation"...
but works pretty well with what I wrote here
Hm, good to know
I just removed some redundant stuff. now it's fast (not perfect tho...) 
Spam in comment
https://feedback.bistudio.com/T163154
comment in spam removed π
https://feedback.bistudio.com/T163708
This one seems pretty important. I uploaded a scenario in which the bug occurs even in ideal circumstances.
Oddly enough it probably works better with RHS that way.
Vanilla tanks seem to take a good chunk of hull damage even if the round doesn't penetrate (IIRC six frontal PCMLs to kill), while RHS tanks can just soak non-penetrating rounds indefinitely.
Setting invalid pitch 0.0000 for B Alpha 2-2:1 REMOTE
Setting invalid pitch -32257595280994599000.0000 for C Alpha 2-1:1 REMOTE
previously only seen these with 0. the negative value is new AFAIK. something worth to report?
also noticed that, wanted to look into it
some garbage value of unitialised variable that causes undefined behaviour
user land variable or engine internal?
engine maybe. undefined behaviour is just that, so it was 0 and now it is -gazillion tomo could be +42
Is βretexturable proxyβ aka βweapon accessories with hiddenSelectionsβ really a won't happen thing?
βconfig oneβ?
config hiddenSelections
no setObjectTexture script command
Well I mainly asked this for the weapon accessories so setObjectTexture was not a thing after all, no?
I meant I tried to clarify that I didn't thought about the possibility of setObjectTexture but just hiddenSelections and hiddenSelectionsTextures for accessories
hiddenSelectionsandhiddenSelectionsTexturesfor accessories
I thought this is possible already
@untold sky please give this ticket request a last consideration after NikkoJT's suggestion which seems quite valid π https://feedback.bistudio.com/T163781
Possible, but that doesn't really fix the issue that both engine and mods expect it to be a mod class.
Engine I can fix, mods eh
If you manage to find a way, not for now by much much later when you have less pressing matters, it would help a lot! π
Mission makers usually don't want to go though all the trouble of creating and maintaining a mod just so they have have custom and proper marker icons for their missions, and what ends up happening is that they are forced to use military icons to represent things on the map that the players end up being confused about π¦
Would it really be that much different from description.ext CfgSounds, CfgMusic etc.? I guess I don't know what goes on behind the scenes to make that work, but my impression is that that counts as a valid config class for both engine and mods.....?
no, it wouldn't. It would be about the same amount of work
for both engine and mods.....?
It just reads both, seperately, after eachother
This might just be because I'm not very smart, but here's what I'm seeing:
- description.ext cfgmarkers is technically possible, depending on priorities
- the engine will accept description.ext classes as valid classes
- mods will still be able to define custom classes as before by modifying the main cfgmarkers
- mods should be able to accept both ordinary and description.ext cfgmarkers classes as valid
- this fills the original request for missionmaker-defined custom markers
...what's the unsolved problem, other than the amount of dev time required to implement?
What is the issue with mods? Classnames collisions?
mods should be able to accept both ordinary and description.ext cfgmarkers classes as valid
Mods need to be adjusted to ALSO look in mission config and also handle correctly when entry exists in both
this fills the original request for missionmaker-defined custom markers
Which I already denied a couple times due to the above mentioned issues
As much as I understand classes defined in description ext can not be appended to main config?
correct
So mods will not break. They just will not know about new markers. Maintainers can add additional config lookup without rush
They will break if they cannot know about new markers
They will know about the new type when a marker has been placed down.
Then fail to retrieve that info which is impossible because the game gets the info from same place so it cannot be missing
Its not worth the effort
there is a CfgSFX and CfgSounds etc override
I presume they have all this check system along the way yes
@limpid rune Moving the conversation here about https://feedback.bistudio.com/T151270
What information would you want/need as params that isn't easily gettable by using commands on the (player) unit? That is for the EH that triggers when the player goes in and out of ADS, meaning switching from CamGunner to CamInternal/External and vice versa.
Currently I pass the unit and a boolean to indicate if the new state is ADS or not.
probably not necessary but maybe include the current optics mode too, like:
[object:unit, string:opticClass,bool:isADS]
how early does the eventhandler fire by the way? One of the flaws with cameraView is it only switches over when the camera transition is complete
Issue with opticsmode is that it will not work if you are in a vehicle (currently), as it will return for the unit and not the vehicle. I am not sure if you meant for it to also work in vehicles, and which kind of information you would expect in that case. But I am at least certain that the opticsMode from weapon attachments is not directly 1:1 for vehicle optics, and a bit more complex to get than what I currently do.
You could still use the getOpticsMode command on the unit to get that information, so it is not too difficult to get it.
The switchOptics EH I currently have triggers as soon as you hit the RMB, so when it starts the transition.
oh right I forgot the getter was added
awesome
then I can't think of anything else currently, vehicles are a bit different (usually defined in OpticsIn/Out classes iirc rather than OpticsModes)
about closing spam tickets:
can we also mark them as "closed, resolved" so we know from another ticket that they are dealt with? e.g spam ticket
What about FFV?
from another ticket
another ticket?
Wouldn't hurt, I always do admin only and remove projects so that they don't show up in any searches.
Spam tickets being referenced by other tickets hasn't occurred to me yet
In what way? Whether opticsModeChanged works for that, or if switchOptics does. In both cases yes. Or in some other way?
For opticsModeChanged it will detect the opticsMode change on the weapon you are holding, which in FFV is the same weapon you would hold when not in a vehicle. For switchOptics, it triggers whenever it would switch from CamGunner to CamInternal/External and vice versa. Which in FFV would still be ADS on the weapon you are holding.
would it be possible to swap the x and y in in-game description and hint of atan2?
if you don't read the description and just read the hint, it's confusing because it shows x atan2 y, making it seem like actually the x-component should go first
also conventionally tangent is shown as y / x, not x / y (technically both are correct tho)
in other programs this convention is observed as well (e.g. C++ shows it as atan2f(float Y, float X))
I already changed this on wiki
https://feedback.bistudio.com/T161983#2289602 spam/bot account
lol is this a bot or human spammer?!
https://feedback.bistudio.com/T163912
gone
Thanks!
Yeah you make it really easy to make a account here: The only control for "Are you a robot?" is to tick on "I am no robot" at creating the account. No picture searching and such.
You don't understand how captchas work
If the server thinks you're a human, it presents you a very simple check box all good.
If it thinks your suspicious the challenge gets harder
you are a normal user doing normal human user things, so it doesn't give you a hard challenge
Then the server "thinks" the wrong things. 
What I want to say is that one bot has tricked the other bot.
No
They are probably using human farms to click captchas
So a human is confirming the "are you human" check
dealt with, thanks
BTW I am so hyped for next week perf/prof inventory patch. π
I still see the ticket 
https://feedback.bistudio.com/T163915 this one?
ah yep, Ded fixeded it
Hi. I posted more info about "warlords" bug, about ammo box when it is called mid air.
Player game freeze when ammo box get smash, there is infinite loop going in warlords script.
https://feedback.bistudio.com/T155317#2291613
nukeded, thx
Saving undefined enum value -2147483648 / 7, context /Vehicles/VisibleNear/Vehicles/Item7/Identity/Diary/Pages/Item1/Log/Item1.state
Saving undefined enum value -2147483648 / 7, context /Vehicles/VisibleNear/Vehicles/Item7/Identity/Diary/Pages/Item1/Log/Item2.state
Saving undefined enum value -2147483648 / 6, context /GuerrilaCenter/Groups/Item8/Waypoints/Item0.combatMode
is this indicating a bugged savegame?
Client: Unhandled user message Type_389
is this new by chance? cant remember ever seeing this before
nop. I also found that a couple months ago. Its been broken for years, is on my todo
can you confirm that you got that during JIP?
yeah you must've
possible from the rpt context but cant say for certain
can provide full client rpt plus server console (and rpt) if needed
no need
This has actually ballooned to be bigger than expected, might just not fix :harold:
setVariable on a object with a target clientId won't set the variable when the client JIP's by going back to role selection and rejoining from there without disconnecting.
And in general object setVariable with target clientId is completely broken JIP wise
thanks for the insight
people used going back to lobby to spawn as another unit from the groups when respawn ticket limit was at 0
afaik only global public vars are persistent, specific client id was never jip compatible
Well it never worked right at least, but it's in JIP queue
Spam comment: https://feedback.bistudio.com/T163998#2292946
gone!
This spam is super sneaky:
https://feedback.bistudio.com/T164017
π
Also spam:
https://feedback.bistudio.com/T164018
or is it an actual feedback? 
I can't think of anything in the game that links to a casino website
Maybe armaholic? 
I actually laughed as hell about that one π€£
moar spam: https://feedback.bistudio.com/T164023
nukeded, thanks
After the two patches are implemented in the last dev build it is time to complete the "Make the "Inventary" great again!"-Trilogy:
https://feedback.bistudio.com/T164025
"after random time"
But you gave repro steps that should be reproducible within a few seconds, why random time?
And the fix for your thing is to just close and reopen inventory and it works again?
Yeah in moded scenario "it seems random"
because there is no correlated feedback (nobody looks on the white bar on the bottom). And if right click not work, drag and drop works -> it seems random. I edit this.
nobody looks on the white bar on the bottom
I do.. If full then full
Does closing and reopening inventory fix it?
No. if full then you walk after that slowly.
this is inteted
Mh I don't think I ever got that one
but mods switch that slow walk thing off. Then it seems that only drag and drop after a cirtain "time" is working.
The problem is bigger if you mod the invantory to be bigger. So after these border only drag and drop work. This feels like it is a bug.
No. After reach this limit only drag and drop is working. Reopen inventory doesn't change something.
can someone change format into + or joinString or something in the BIS_fnc_addStackedEventHandler code?
it compiles a new code using format which causes long codes to fail... 
can you make a ticket with repro? will need that for qa
ok. I'll do it tomorrow
@solid marten
https://feedback.bistudio.com/T164036
I made something for now but I didn't test it myself 
ok it "works" (fails to compile)
thanks
https://feedback.bistudio.com/T164049
This has been around for a while but I didn't find any posts about it on the tracker. If it's a dupe sorry!
I'll have a look no promises
busy Monday, spammers go to work early
nukeded, thanks
I had this weird graphic bug on a few different maps and servers and also heard it from others.
Sometimes there is a weird line/break in the sky as you can see in the first screenshot.
fogParams visible at the top of both screenshots.
https://steamuserimages-a.akamaihd.net/ugc/1843668602666592836/BEA8A12918FEBAF44F8D4601B3CAB261EEF0A049/
I found out, you can fix it by executing 0 setfog [0,0,0] on the client and after a few seconds it syncs back to server values and the visual bug is gone.
https://steamuserimages-a.akamaihd.net/ugc/1843668602666593230/EB90395B6385904F06C7B1B48FC1E68B2F99DCD2/
Is this a engine bug or script/value problem ?π€
I've had people mess things up by Zeusing all the fog sliders to the bottom.
which is actually [0, -1, -5000]
Player vision seems normal but it stops launchers from locking.
Our weather is script controlled, no user input possible.
And the fogParams don't change on the server, i just force the client to resync and it's gone
hmm yeah, different issue.
Another one π°
When i drop a uniform/vest/backpack on the ground and relog, their inventory is empty for me after reconnecting (tested client prof & stable).
I wanted to test setUnitLoadout for our loadPlayer function instead of multiple calls/functions to add the gear and while testing i stumbled upon this weird behaviour.
The player unit is created on the server and setUnitLoadout also executed on the server.
Only behaves like that with setUnitLoadout 
Any plausible explanation ?
Before dropping on the ground and re-logging: https://gyazo.com/c651969e4b049f8db9375c16aa23b4af
After re-logging (Inventory of vest/backpack/uniform is empty) : https://gyazo.com/98308fa1224f3dc12eb2ba3c797e06db
Edit: To make it even more confusing, if you take the seemingly empty gear and relog again with it equipped. The inventory is correctly filled afterwards 
So again get/set unitLoadout is done by the server and after re-logging everything is there again, so it's just a "visual inventory problem" since the data is valid.
https://gyazo.com/e3a6d6a557997365ac00a3a84778eb47
Better make ticket
In case someone has issues on current dev branch with ammo/group eventhandlers that aren't fixed, please remind me of the issues in case they went under
what about this one? (useless objects being reported in knowsAbout)
https://feedback.bistudio.com/T161515#2281406
I didn't test it after that so not sure if the problem still persists
did't fix that. The game does that internally so π
They might be checking if the objects can be used for cover or whatever.
in any case, I doubt anyone (for scripting) cares if the AI knows about a fridge... 
that EH would normally be used for when AI detects targets
basically only valid targets (whatever targets/targetsQuery commands return) should be returned imo
Here you go https://feedback.bistudio.com/T164086
We had similiar situation: Mission spawns loot (uniforms, vests, backpacks). Loot can taken but is not seen on player (also not seen from POV). It seems it is only modpack(in these case CUP Units) uniforms. Also after some playing dropped vests cannot retaken. only other player can taken dropped vest.
server: v11
client: v11/vanilla
We don't have this kinda issues with AI (No HC & AI created on the server). Only player gear sometimes, i guess that's some form of "desync" since i can see it more often on players that live far away from our server or have bad internet.
It happened to me and iam sitting 2m away from my server (the thing with uniform is not seen). A players from extern had the thing with the vest on ground cannot retaken but from me it was possible and also uniforms wasn't seen.
I will do check with vanilla executable in the next days.
I recorded the bug to make it more clear.
https://www.youtube.com/watch?v=bwxybZaVcx0
I'm curious to know whether the issue you describe here is linked to this one
https://feedback.bistudio.com/T126030
Spam comment: https://feedback.bistudio.com/T164083#2294490
Spam comment: https://feedback.bistudio.com/T164035#2294715
all cleared, thanks
this wasn't spam: π¨
https://feedback.bistudio.com/T164086
oh nvm 
I thought you had closed it... 
Please make a ticket so i don't forget
Do you think this could be related to the Ai checking the nearby objects like stones etc causing some micro sutterings exactly when they get shot by a bullet?
https://feedback.bistudio.com/T164110
I'll test other EHs too to see if they have any problem as well
ok first test: EnemyDetected doesn't trigger at all π
test 2: UnitLeft: if you do [unit] joinSilent grpNull the EH triggers before that unit has a group (group unit is grpNull). not sure if that's intended.
seems to also happen even if you do [unit] joinSilent someGrp the group of the left unit is shown as null 
please put in same ticket
so i have bulk for next week
all of them in the same ticket? ok
I guess I'll put them all in this one... 
Before i open a request... Can someone confirm that removeAllMissionEventHandlers does NOT reset the event handler creation index to 0. ```sqf
_evh1 = addMissionEventHandler ["Draw3D", {hint "y";}];
systemChat str _evh1; // Output: 0
_evh2 = addMissionEventHandler ["Draw3D", {false}];
systemChat str _evh2; // Output: 1
removeAllMissionEventHandlers "Draw3D";
_evh3 = addMissionEventHandler ["Draw3D", {false}];
systemChat str _evh3; // Output: 2
it shouldn't
it's correct
"This command also resets the Event Handler creation index to 0."
that's not valid anymore
I know. like I said it's intentional
see:
https://dev.arma3.com/post/spotrep-00102
Fixed: Mission Event Handler ID would reset to 0 when all handlers were removed
which was bad, because�
Was actually pretty handy... no clue why its removed
Can it be reverted? I dont see the issue with resetting the index to 0. For example it still works for display eventhandlers.
no. that would cause issues
if you removed all EHs and started adding all over again, and a mod was using the old values to delete its own EHs, it would delete the new EHs instead
Not resetting is better for avoiding bad code killing the wrong EH. Resetting is better for figuring out whether ACE installed its shit yet :P
Aha, so because of the modding community :p
added a huge feedback ticket. have fun π€£
seeing it this way, I can get behind
don't get behind me π π
too late π
blocked, cleaning
Still some spam from this bot, he is blocked though. https://feedback.bistudio.com/p/Nishaknapp/
And spam: https://feedback.bistudio.com/T164123
done
and done
done
This can be closed since reflector lights were added in a recent update: https://feedback.bistudio.com/T77584
And...spam comment: https://feedback.bistudio.com/T164043#2296201
May I ask for an 3den command that switches to another map without user confirmation?
Something like:
set3denTerrain "VR"
I need the user to be on the VR map when my mod does something and I don't want to ask them politely π
If possible please let me know to make a ticket
well... @sinful kettle
Filtering for real "targets" wouldn't filter out buildings and props, as they are indeed listed in targets command π
nvm then 
I was always using unit targets [true] 
But anyway why are they listed as targets?
Β―_(γ)_/Β―
AI being able to target buildings makes sense to me. They can engage them as weapon targets (...sometimes) and I think they can also be used in movement orders
Sure. What about garbage washing machine?! π€£
I don't think it's that specific. It's hard to procedurally determine what's a potentially useful object, and the more rules you put in to try and narrow it down, the more likely it is that you'll accidentally filter out something you do need.
And it's not just those. E.g:
decal_mud_road_1.p3d
That's a decal
It's a purely visual object
So are helipads and I'm pretty sure targeting those is important
Btw Ded high command is broken right now 
https://feedback.bistudio.com/T164112
ye its been months ago
Also regarding this comment:
EnemyDetected:
The event handler doesn't seem to trigger at all.
(group enemyAI) addEventHandler ["EnemyDetected", {systemChat str ([_thisEvent]+_this), copyToClipboard str ([_thisEvent]+_this)}];
-> ["EnemyDetected",O Alpha 1-1,B Alpha 1-1:1 (dedmen)]
Did you fix it or it just works as it is?
I tested it several times both on my group and the enemy group but it never triggered 
I just tested and it works
I also tested it when I made it and it worked
I just placed AI behind a wall, walked around the corner, and when AI saw me it triggered
Maybe in my case they already knew about each other then (but knowsAbout was low so they didn't react) π
So I guess the EH only triggers when an enemy is added to the targets array, correct?
it triggers when a new enemy is added to target list, that hasn't been known before
I thought it triggers when they actually "detect" the target (e.g. react by aiming, changing stance, etc.)
So that explains it 
[Request] Expose cover points of objects/buildings via sqf command
https://feedback.bistudio.com/T164226
@sinful kettle did you make a ticket for that already? i searched with cover point[s], but couldnt find any
i thought to remember you had brought them up recently, didnt you?
No. What are cover points? You mean outer points?
E.g. the edges of geometry where the cursor sticks to? (shows a line)
cover points are where AI takes cover π
the engine generates them based on the geometry
like edges or alongside
pretty sure Suma/Ondra had a blog post when he introduced the tech in A2 days
sadly most of his blog posts are gone with BI changing their website and they werent mirrored in the forum or else
had often good technical depth to understand how the engine works
Improved AI commanding - selection of cover point on position where you are sending soldier to
https://forums.bohemia.net/forums/topic/150500-development-branch-captains-ai-log/?tab=comments#comment-2387939
In case of low / distant danger, civilians try to hide in user predefined cover points.
...
In situations where there is an insufficient number of cover points, especially in the outer parts of the area, AI might run away from danger, then calm down and move back searching for cover point. Then, on the way back, it gets too close to the danger it will start to flee again. This is obviously not ideal situation, so pay attention to proper placement of the cover point and use Debug mode to see where unit is heading and how terrified is.
https://community.bistudio.com/wiki/Arma_3:_Civilian_Presence
Wed like to try using this thread to track clear updates about changes to the AI. But first, some all important disclaimers! Not all the changes made to the AI can be logged as and when they happen Sometimes the changes made by the programmers require other departments to update or pack their dat...
could find only these. seems in the past some called them also cover positions
I was wondering if we could get an alternative to the allMissionObjects command, because it's so slow and useless for frequent usage (it has to be used frequently to update the list of mission objects, because we don't have an "objectCreated" mission event handler sadly)
the entities command is super fast in comparison
I was hoping for a command that returns only mission placed objects that are "significant" and not "temporary" (such as projectiles, footsteps, insects, etc.). in other words:
- all units, agents and vehicles (alive or dead). we already know this is fast
- all simple objects. this is also fast
- all misc "vehicles" placed during the mission, such as buildings, ammo boxes, etc.
if only there was also a command that returned the 3rd case I wouldn't ask ofc. or maybe we could just get a command that returns the 3rd case?
because we don't have an "objectCreated" mission event handler sadly
yet
I was hoping for a command that returns only mission placed objects that are "significant" and not "temporary" (such as projectiles, footsteps, insects, etc.). in other words:
wouldn't make the scanning any faster
doesn't the game keep a list of the 3rd case?
That is just vehicles command
because it seems the first 2 cases are indeed arrays
- also has a command for it
no. vehicles doesn't return buildings. only actual "vehicles" are included
π€
based on your reaction I assume it's not intended π
so I made a ticket:
https://feedback.bistudio.com/T164242
seems like all objects with simulation = "house" are not included 
maybe the command filters them manually? π€
anyway I think adding a new command instead is better. that way you won't break current stuff that rely on vehicles not returning buildings...
I'll just make ObjectCreated eventhandler and tell you to go away

could you please look into it at least? π
maybe it's not intended behavior...
no 
Actually ObjectCreated EH would be quite performance critical π€
Maybe its too dangerous to give that to scripters
maybe you could filter out some objects?
e.g. insects, footsteps, projectiles, sounds, lights, etc.?
but where to start and where to stop
for example I imagine people will want projectiles
I can't think of a good way filtering method either. allVehicles it is then π
I can only think of a "special" event handler, like addObjectInitEventHandler, with special filters... 
can hack it and use eventhandler arguments, which missionEH has
Is it acceptable for an arma 3 server to do donations for ingame but don't have monetization because I've seen project reborn do this and I've reported them on the forums
Thats not #arma3_feedback_tracker
But, servers are allowed to take donations. REAL Donations, for real donations you get nothing in return, So no ingame items or anything, without monetization approval.
Many servers love to call things donations, that actually are plain purchases.
Our monetization people don't read forums, report to monetization@bohemia.net
they are selling vr suits that noone has access to but only of you pay for it
Thats not a donation, thats a purchase. Thats not allowed.
Since something monitoring projectiles is likely to be fairly focused on monitoring only projectiles, perhaps you could have separate EHs, one for non-projectiles only (objectCreated) and one only for projectiles (projectileCreated?). Then both things are possible, but any script that only does one (likely to be most scripts) gets better performance.
comment removed π
The filter idea would work with that
btw how should the filter be?
e.g. special strings that are internal enums?
I'd do it like regex params
"*" any
"P" projectiles only
"PU" projectiles and units
That would be efficient to scan in engine
what if someone wanted to pass actual args to the EH? π€
they're still there btw
fixed
Tickets reported as fixed and clear to close:
https://feedback.bistudio.com/T124332
https://feedback.bistudio.com/T124569
closed
AFAIK there is no way to determine what title effect a number relates to (unless you check all the source of all scripts)
possible to have a SQF command that would return the associated code?
practical use: trying to determine when to show a cursor - you dont want to mission intro/outro/cutscene effects, but often title effects are also used for HUD in various ways
gone - it's nice how a spam posted and another spam added a spam comment to it
killing two birds with one stone π
I don't see how title affects are similar to calltrace's?
I think he means he wants a command that gives a similar return to stacktrace/calltrace, but for title effects
Hence i used quotes for "similar" - in other words it would be handy to
possible to have a SQF command that would return the associated code
i used the stacktrace/calltrace references as they help you understand the code context
what is "associated code" here? the code that created the title effect?
yep. the id is the layer, so there must be some internal ref the engine holds to the code from what i understand
it holds a ref to the layer that was created
but probably not to the piece of script that created it
i see. what about the properties/setup of the title effect? with the info you could at least determine easier where it come from
Spam: https://feedback.bistudio.com/T164287
Weirdly...opinionated? spam comment: https://feedback.bistudio.com/T117650#2297997
the alternative would be to have a way to see if there is a black/white screen from some title effect
there is no way to check for that right now, is there?
Spam comment: https://feedback.bistudio.com/T150733#2300150
Spam: https://feedback.bistudio.com/T164431
gone
btw Ded I have a feedback on group EH "CommandChanged".
It only shows the group name and the command. Is it possible to show the subgroup that received the command? only the leader of that subgroup is enough
probably
Leo do you have similar issues?
https://forums.bohemia.net/forums/topic/238374-lineintersects-not-working-in-dev-version/?tab=comments#comment-3458365
You'd probably have noticed
I don't use those 
surfaces version works fine
using it rn
that issue is bogus
he's checking under the terrain π
pos for those commands is ASL
and he's doing that in VR
VR terrain height is 5
not 0.8 or 1.xxx
so obviously those are under the terrain
if you feel like fixing lineIntersectsXXX commands they do have a bug with animated selections tho (sometimes intersection is not returned)
but you said you don't want to fix it before π
huh? it's really close to the water level, are you certain of the 5m thing?
also the only water on VR is near ~[7500,7500] grid pos, and you can tell it's actually that close to the water level
I thought it was smaller than 5m actually
because when a vehicle explodes the water splash effect triggers π¬
oh. I thought you meant 5m is too close π
spam comment:
https://feedback.bistudio.com/T164440
Spam comment: https://feedback.bistudio.com/T163398#2301058
Is there a trick to deleting agents created with createAgent? Ive spawned some on the map, and then delete them all later using - { deleteVehicle agent _x } forEach agents; They disappear, but theyre still marked on the map in Spectator mode -
should agents still return <null> references?
and should
{ deleteVehicle agent _x } forEach agents;
still show up in the observer mode
or worth reporting as bug?
spam comment:
https://feedback.bistudio.com/T164497
Double spam in comments https://feedback.bistudio.com/T164449
It is worth requesting an alt syntax for allGroups -> allGroups <side>
perhaps even better groups <side> to be consistent with units <side>
I'd like that
cannot do alt syntax for a nular, needs to be new name
groups is 
I know
https://feedback.bistudio.com/T164521 groups <side> request
This is still here
Does it seem plausible that boundingCenter could get additional parameters similar to those added to the other boundingBox commands? (and/or boundingCenterReal)
@willow finch
If you find a bug, instead of putting it on the wiki, plz report it here first (or make a ticket) so that it can be looked into. If it was confirmed to be a bug but wouldn't be fixed, then put it on the wiki.
I am referring to this:
https://community.bistudio.com/wiki?title=setCaptive&diff=309045&oldid=303351
sure I can do that
clipping type?
eh. I'd say no, center is center
I'm not sure about that. Sometimes "unclipped" bounding boxes are oversized in a particular direction (e.g. too tall) which would result in an off-centre "centre point".
I don't make 'em, I just have to deal with 'em :U
But wait, there's more:
https://feedback.bistudio.com/T163571
https://feedback.bistudio.com/T164511
https://feedback.bistudio.com/T164122 (closed but not removed for some reason)
https://feedback.bistudio.com/T163663
Would anyone be able to make a bug ticket for me if I provide all the info? I'd rather not make an acc just for this one thing ^_^
what is it?
Was being discussed over tehre
setUnitRank is used to change the players rank as they change role.
If a player leaves the server, then another player respawns, once the player reconnects the units rank is no longer what it was, and defaults back to private. This video demonstrates this: https://youtu.be/AfBEyMrWfOk
Reproduce:
Load this mission: https://steamcommunity.com/sharedfiles/filedetails/?id=2793205082
1. Join server and have both players select role through menu
2. One player disconnects, other player dies/respawns whilst the other isn't on the server
3. Player reconnects, rejoins the squad and see's the player as a private again (what the unit is in the SQM)
Got a nice little video for it too π
The one other thing which may be worth adding, is the other player still sees themself as the same rank, its only the JIP player who see's it as private
the other player still sees themself
you mean the player that didn't leave?
if so that's obvious
Yes, I agree its obvious but I wouldn't want a piece of info like that delaying it
well the issue I made says "it's not JIP compatible". players that are in the game are not JIP so it's not really important detail
Fairs
Its just odd as I wouldn't say its not JIP compatible, as it is as long as they didn't respawn when you weren't on the server
indeed
I'll edit the title then
partially JIP incompatible π
just don't die
reworded the ticket a bit
Devs, is there any change we could get OnScriptError event handler?
It would be cool to be able to create your own error dialog
with call stack and variables displayed
with call stack and variables displayed π
u dont have to look in .rpt
and I believe .rpt doesn't show variables just call stack
Supposedly mission is not accessible, any ideas? Worth me just adding a zip file of the mission? I can get it from steam workshop and steamworkshopdownloader.io
https://github.com/DomT602/Doms-Arma-3-Mods/tree/main/Dynamic Groups/DynamicGroups.Malden
Could you add that link to the description, and this line as a response to where its used: https://github.com/DomT602/Doms-Arma-3-Mods/blob/main/Dynamic Groups/DynamicGroups.Malden/Groups/fn_setupPlayer.sqf#L51
Thanks π
this function does not work in arma 3 (the note in the website was added at the time of this message being wrriten)
https://community.bistudio.com/wiki/BIS_fnc_scriptedMove
going to the source code, its quite easy to make it work for Arma 3, just replace line 64 that had _nic = [objNull, _guy, rPLAYMOVE, _walkAnim] call RE; with
_nic = _guy playMove _walkAnim; will this replacement be possible for a future update?
preview of the function's source code : https://sqfbin.com/siyuxumavocacaferovu
ft ticket please
thanks
btw KK do you happen to know if the game keeps a list of placed buildings during the mission (not terrain)? because vehicles command doesn't return them.
if they're already available as an array in the game I'd appreciate a command that does return them (allMissionObjects can be used too but it's super slow to do it that way)
I asked this from Dedmen too but he wasn't interested 
and xxbuilding commands return them?
iirc there's only nearestBuilding command but I didn't try it since it only returns 1 instance
Also it returns terrain buildings
But I need mission objs only
It's not just buildings. vehicles command doesn't return any object with simulation = "house"
Such as rocks you place in eden. But if you place a "thingX" object such as a wrench it is returned 
I was hoping for a fast command to fetch those missing "house" objects (comparable to commands such as allUnits, vehicles, entities, allSimpleObjects, which are fast, since the game probably keeps them as an array or something)
if you do allMissionObjects in an empty mission it takes several milli seconds to complete (vs several micro seconds taken by those other commands) 
Plus I don't need 99% of what it returns anyway 
Make a ticket plz
I already have
https://feedback.bistudio.com/T164242
Just wanted to know if it's possible.
As in if the engine stores them in a faster container than a quadtree or whatever
@sinful kettle so there are lists of vehicles: fast, normal, slow, very slow each of the categories is divided to visible near/far and invisible near/far. What exactly do you have problems with, just the houses?
I didn't test all.
the houses one was the most annoying
would you like me to check all?
also does it depend on the category, or the simulation?
e.g. a rock is simulation = "house" and it's also not returned
I think it also applies to weapon holders
not sure tho
simulation will determine frequency of simulation, so houses will be slow vehicles i suppose
and is the reason not included into vehicles as those are normal vehicles
ah
well I want all I guess π
since houses are probably the slowest ones
just out of curiosity, what other objects are not returned?
what do you mean, by vehicles command?
fast, slow and very slow not returnes
e.g an object with simulation = "thingX" is returned
slow are simple objects and houses, very slow - lamps, fast I guess bullets and particles
thingX would be normal
lamps
light objects (e.g#lightpoints) or lamps?
simulation is wide concept with rules including update frequency, so it would be huge subject
street lamps
10sec
how slow would it be to return all of them? (excluding projectiles)
except for engine "hidden" ones (e.g. sound objects, light objects, etc.)
they are separate lists so this would mean iterating through all of them so depending on how many objects you have could be either lightningfast or pretty heavy
I guess what I want that is missing is slow and very slow
lets look at the problem, user placed houses?
yes
the house is determined by presence of the pathway
let me experiment with slow and very slow to see the ranges
@ KK: (https://feedback.bistudio.com/T163286#2304221)
Per https://community.bistudio.com/wiki/setFeatureType it seems like flying aircraft are probably considered to be Type 2 features and therefore always visible within terrain render distance.
@solid marten I was literally gonna update the request for a version with flags but I thought it would be difficult! you read my mind! thanks a ton! π
btw will it be available this week?
I can't wait to try it!
yeah should be in next update
would it be worth making a ticket to request extending
https://community.bohemia.net/wiki/ctrlAnimateModel
to match the parameters of
https://community.bohemia.net/wiki/animateSource
(doesn't have the 'speed' param)
I know UI models don't see much use but being unable to manipulate speed is fairly annoying in some instances
Don't need a ticket, give me a minute
nice ty
Okey one minute more while I take care of my head exploding over this absolutely mind blowing heap of whatever this is
I have to browser windows open, in the same firefox instance.
Left window, I open wiki, I'm logged out. I try to login, I get error message, and am still logged out.
Right window, I open wiki, I'm logged in!
WHAT
Same browser instance, same environment, shared cookies.. what the literal
Now I opened animateSource in right window an am logged out
But if I reload the ctrlAnimateModel page I have open in the same window, I'm logged in! WHAT
Holy crap
you sent me a bohemia.net link, that one is completely broken and doesn't work.
The other one is bistudio.com
holy shit @gray wharf FYI I guess
https://community.bistudio.com/wiki/ctrlAnimateModel Here fixed @limpid rune
...does it already exist and it was just the wiki missing it?
apparently so? @untold sky?
top speed
this was suspicious indeed π€£
I was just thinking it was copy paste between the two commands or somehting
This is used for CT_Object right?
kinda of an obvious question but I'm really new to the whole 3d controls concept
Hi everyone, I don't think this ticket has been picked up since 2017, but I think the ability to hide an object (especially a unit) but maintain its collision and ability to be shot would open up a lot of possibilities (and I'm experimenting for a project and selfishly need this functionality!)
short answer, probably not happening at this stage, hiding had many problems in the past, so doubt anyone will want to open this can of worms
That's a shame but understandable. Cheers @solid marten
there are invisible objects in A3 that have collisions also those made invisible with setting textures to nothing also have collisions I think
can we have a command like
allowThirdPersonView = false;
for now we must always create a loop for it
That's very true, I might have to make a minecraft hitbox by using attachto on an invisible unit's selection positions π
I don't think this will play nicely with collisions with environmental objects though π¦
It is already game option and also server option
yes but how would i force it in a scenario?
without a loop?
sp or mp?
mp
difficulty setting
and if i want to change it (3rd person camera availability) during the mission? (context: multiple submissions that may have different, say, "modifiers")
switchCamera
or maybe just switchCamera to internal and then disable inputAction:personView
Realised I could just use an RVMAT with following render flags:
RenderFlags[]={"NoAlphaWrite", "NoColorWrite", "NoTiWrite","NoReceiveShadow","NoZWrite"};
Now unit is invisible and still has collisions/can be shot etc.
@untold sky can you change the Addon Builder to perm delete the addon temp folder? 
my Recycle Bin is always full of those...
and a history list for previously used paths π
no
Because I say so
Because I don't want to
well say that in the first place... 
https://feedback.bistudio.com/T164350 spam in the comments
(profiling)
Created a ticket about RscShortcutButton rendering problems https://feedback.bistudio.com/T164820
I have an idea. is it possible to add scripted event handlers for tasks?
e.g. when a task is completed, a scripted EH added to the object/side that has that task is triggered?
oh wait setVariable doesn't work on sides 
but there are not that many sides so maybe you can do it via 4 hashmaps:
_sidesEvents = [createHashmap, createHashmap, createHashmap, createHashmap];
_sideEvents = _sidesEvents param [_side call BIS_fnc_sideID, createHashmap];
Does that make sense? Theres going to be a condition that returns true that makes the task completed right?
idk if this is the right place for this; My latest note on https://community.bistudio.com/wiki/if
I believe this falls under the category of magic functionality. Non-existence of code shouldn't change functionality, this is very unintuitive and definitely a bug, imo
I dunno, seems reasonable to me. What would you expect the missing else case to return?
missing case should not return
not possible, the way scripting is implemented, something has to return, even if it is nothing (nil). The behaviour is expected, whether or not it is desirable is up to the scripter to decide and adjust accordingly. the note is good
https://feedback.bistudio.com/T164946
Spam account
Some spam comments, https://feedback.bistudio.com/T163720
I cannot fix this even though I can see where it goes wrong, because it may break already existing image/button layouts which used hacks to work around the problem. In short, buttons shorter that width = 2 x height are rendered differently, but those > 2 x heght are also not great. I'm going to add a config param that will switch on different render
alright sounds good. what about the text vertical centering, can u do anything about that?
I need a ticket and a very good repro
ok will create one
same thing though cannot be fixed for the reasons stated above
ok
the control is weird one because the text position is defined by this class in the control class: ```class TextPos
{
left = 0;
top = 0.05;
right = 0.005;
bottom = 0.000;
};
shouldn't the text be automatically centered like in normal buttons?
now the top position is absolute and not relative to the button's height, so it will only work when specifically set for different buttons of different height
good?
ooh nice π
i hope this repo is good enough for you, did my best π https://feedback.bistudio.com/T165085
the animations of some objects cannot be changed using the animate command, even if they're simple objects or you disable their simulation.
for example, when I spawn a wind turbine, it keeps rotating with time and I can't force it to spin the way I want
would it be possible to get a command to force the object animation? like forceAnimate or something? or an optional bool at the end of the animate command args?
https://feedback.bistudio.com/T165117
p.s: it's ok if the effect is temporary (only lasts a frame). I just want the animation to actually change. for example with wind turbines, the following code should return true (it currently doesn't):
_animPhase = random 1;
turbine animate ["mxlprop", _animPhase, true]; //with new forced arg: turbine animate ["mxlprop", _animPhase, true, true];
turbine animationPhase "mxlprop" == _animPhase
setCombatBehaviour on group doesn't activate although it returns set behaviour.
grp1 setCombatBehaviour "SAFE";
combatBehaviour grp1; //"SAFE"
combatBehaviour (leader grp1); //"AWARE"
I'm on dev branch but I think it happened on stable yesterday.
why is this a problem? This command allows you to set it on group entity separately from units. Why, donβt ask me, community wanted more control over engine process
Hmm. On stable my experience is that the unit version didn't do anything. Use case was setting a driver to careless so they didn't freak out at enemy vehicles. Had to make a separate group for the driver instead.
I guess some AI functionality uses unit behaviour and some uses group behaviour?
Spam comment: https://feedback.bistudio.com/T165154#2313214
Please, consider to add Windows 11 in dropdown list of Operating Systems.
I guess it's a matter of perspective of what's expected from group applied behaviour. In this case it's more of a layered behaviour since it has no direct effect because the leader stays the same. It might have it's use cases however. Edit: thanks for updating the Biki, it's more explicit now.
is it not possible to delete the task? don't want to flood the feedback tracker with my ignorance lol
just make a comment saying it can be closed and someone will get to it eventually
Done, thanks.
nobody has commented or viewed this yet, I'm concerned if there is something wrong
been 10 hours
You haven't done anything wrong. There are a huge number of tickets on the FT and the devs are busy. Some never get comments. If it's important, mention it again in a few days.
Also, all of those 10 hours have been nighttime in EU (it's now ~9am). Definitely don't expect the devs to check the FT outside work hours.
i added comment
I thought I was asking for something trivial or already tried
replied
Please provide hpp vestion, i.e description.ext that works
alright wait one
iirc the problem is that RscObject controls can only be under the Objects class
ctrlCreate only works for those that go under Controls class
examples of working in controls not working in objects?
also please post everything in the ticket so itis all in one place
the example he posted in the ticket is correct. doesn't need description.ext
huh? I asked for description.ext example
I need to understand what is not working, for this I need 1. working example (hpp version) 2. not working example (sqf version) so I can investigate, capish?
there you go, https://feedback.bistudio.com/T165245 sorry for the delay and misunderstanding
thanks, I will have a look, if need more info I will ask
I would say, its the inback variables, zoom duration, and the z parameter? but don't count this on me because I'm not the brightest in GUI lol
these are missing sqf commands
vaguely, rscobject is in https://community.bistudio.com/wiki/Arma_3:_ctrlCreate/classnames,
and https://community.bistudio.com/wiki/ctrlCreate
says about the former page :
This list contains controls which are guaranteed to work with this command. However, controls not listed here should work as well - as long as they are properly defined.
controlName ctrlSetPosition [x, y, z], you are using the wrong syntax
my point was that the control is not even created. all you need is:
findDisplay 46 ctrlCreate ["RscObject", -1]
``` which gives you control null
but something like this works
findDisplay 46 ctrlCreate ["RscText", -1]
the problem is the RscObject control (CT_OBJECT type), which can't be created using the command
that's true but, I used this syntax you mentioned before, and it did nothing too
regardless, this is what you use for xyz
true true, I hastily made the example and overlooked this,
_RscObject ctrlSetPosition [0.5 * safezoneW + safezoneX,0.5* safezoneH + safezoneY,2];
_RscObject ctrlSetPosition [0.5,0.5,2];
still these do nothing
fixed the report page
yes, I'm aware, you can already tell when you create a display that its not existent
@sinful kettle there is also ctrlObject but afaik there is nothing in it but at least it creates something
rscXXX and ctrlXXX some weird stuff
just saying, I remember dedmen adding ctrlsetAnimateModel, but its not on the page, https://community.bistudio.com/wiki/Category:Command_Group:_GUI_Control_-_Object
ctrlObject is empty... 
count (configFile >> "CtrlObject") == 0
so technically it's not even a control π
we do indeed live in lovely days
sorry, I was a bit confused, it was missing a parameter, https://community.bistudio.com/wiki?title=ctrlAnimateModel&diff=309289&oldid=309212
configFile >> "ctrlObjectContainer" too
use utils 2
INCLUDE INHERITED ENTRIES: false
SHOW CLASSES ONLY: false
UNLOCALIZED TEXT: true
CONFIG PATH: bin\config.bin/ctrlObjectContainer
SOURCE ADD-ON(S):
*/
class ctrlObjectContainer: ctrlObject
{
};
How is it going, is it caused by missing commands or is it unclear?
I hope this doesn't sound rude lol
I dunno, haven't looked at it
alright, just so you know I really appreciate that you are working on this, thank you
I am not, well not yet, so I'd appreciate if you let me do it at my own pace, thanks
yeah no there is no rush lol, just showing me appreciation
assigned wrong ticket
I have a question. Is it possible to have a command that returns object animation parameters, such as source, sourceAddress, etc.
Basically these parameters in model.cfg:
class Animations
{
class FrontWheelR
{
type = "rotationX"; // The type of animation.
source = "wheel"; // The controller that provides input.
selection = "pravy predni"; // The name of the skeleton bone used.
axis = ""; // The name of the axis in the model.
memory = true; // Is the axis defined in the memory LOD of the model or not?
sourceAddress = "loop"; // Does the animation loop or not?
minValue = 0; // The minimum value of the motion range.
// i.e. The controller input when animation phase is 0.
maxValue = 1; // The maximum value of the motion range.
// i.e. The controller input when animation phase is 1.
angle0 = 0; // The starting angle of this animation, where the
// animation phase is minValue
angle1 = "rad -360"; // The ending angle of this animation,
}; // where the animation phase is maxValue
};
or does that fall in the "model ripping" category? 
e.g. a command like object getAnimationParams "animationName"
iirc I had a ticket about that and it was claimed but wasn't implemented 
let me see if I can dig it up
yeah, but it was more limited (I only asked for selection name in that ticket):
https://feedback.bistudio.com/T150103
it was claimed by Dedmen
so won't do? @untold sky π’
that was before I had proper ticketing practices that make sure I won't forget stuff
I don't know, would have to look at code to see if sensible.
And won't do that now
Spam comment: https://feedback.bistudio.com/T165245#2316939
Spam comment still up
https://feedback.bistudio.com/T74177
This issue is still a thing
How could go around it? I just want the effects of the guns firing, I don't need the rounds actually landing
Got around doing this:
_gungroup = [arty_gun_1,arty_gun_2,arty_gun_3,arty_gun_4];
arty_gun_1 doWatch gun_watch_1;
arty_gun_2 doWatch gun_watch_2;
arty_gun_3 doWatch gun_watch_3;
arty_gun_4 doWatch gun_watch_4;
{
[{
params [["_gun", objNull, [objNull]]];
_gun setVariable ["artyFnc", {
_this params [["_gun", objNull, [objNull]]];
if (_gun isEqualTo objNull) exitWith {};
_gun setVehicleAmmo 1;
_weapon = (weapons _gun) select 0;
_mode = (getArray (configFile >> "cfgweapons" >> _weapon >> "modes")) select 0;
gunner _gun forceWeaponFire [_weapon, _mode];
[{
params [["_gun", objNull, [objNull]]];
[_gun] call (_gun getVariable ["artyFnc", {}]);
}, [_gun], 10 + random 2] call CBA_fnc_waitAndExecute;
}];
[_gun] call (_gun getVariable ["artyFnc", {}]);
}, [_x], 10 ] call CBA_fnc_waitAndExecute;
} forEach _gungroup;
also added a fired eventhandler to delete the projectile
Is there any interest in crash dumps from Linux servers, or is that basically unsupported?
Antistasi+RHS tends to fall over after a few hours, fine on Windows servers.
linux crash dumps are hard to analyse. I need the core dump which is huge, and copies of relevant libraries that you used on your server :/
I don't want to go to that effort unless its really important
spam comment:
https://feedback.bistudio.com/T164321
Anyone up for fixing an annoying Eden Editor bug? https://feedback.bistudio.com/T165559
Steps included + video
is anyone up for this? should be very easy π
https://feedback.bistudio.com/T165584
all you need is something like if (veh.allowCrewInImmobile()) return; before the bit of code that disembarks the AI π
https://cdn.discordapp.com/attachments/141866231362879488/982138538403450890/IMG_7044.jpg
RscCinemaBorder uses safezoneX and safeZoneW instead of Abs variants, so this happens I guess
(ref: #arma3_troubleshooting message)
Spam comment: https://feedback.bistudio.com/T165628#2321532
He's using one monitor so abs variants should be the issue. I guess there is some hardcoded limit set or so
I keep attempting to play ops with my unit and keep getting the same error "STATUS_ACCESS_VIOLATION" and I have no idea how to fix it any ideas?
#arma3_troubleshooting is the way π
https://feedback.bistudio.com/T152946#2310033
Spam on my ticket π
btw is there still someone left that can fix incorrect stuff in models? (e.g. missing parts of LODs)
or is that not gonna happen anymore?
how long does it usually take for example https://feedback.bistudio.com/T165085 to come to the stable branch?
As long as it takes for a stable branch update to happen
if the last update was recent, then a long time. If the last update was a long time ago, then some shorter amount of time
ic, does it go straight to stable and not thru dev/prof?
everything goes through dev
Everything goes thru dev
The ticket you linked to most likely won't make it to prof
Only selected small, non-game breaking "platform" changes are pushed to prof (there should also be no data changes)
ic, thx, good to know π
The path is dev > stable. Sometimes things go to prof as well but this does not add wait time unless it turns out to break something.
I think you are thinking of this from the wrong angle. Unless it's a core feature of the update, the update will probably not be delayed to wait for something to go through dev, and the update definitely won't be brought forward to get one change out sooner. The update is planned for a certain time and whatever's passed QA at that time will be in it. It's not "oh this change will be on dev for X days then it comes to stable", it's "this change will be on dev for however many days it is until 2.10 is ready".
Depends on your definition of "wait"
https://forums.bohemia.net/forums/topic/140837-development-branch-changelog/?do=findComment&comment=3460555 it already went to dev 4 days after KK posted that reply on the ticket
it is in todays prof, but its not mentioned in changelog because its not back/forwards compatible, so technically people cannot reliably use it on prof
nice, missed that one
Would it be possible to make the animate and animationPhase commands take animation indices in addition to the names? (i.e. obj animationPhase index)
based on a quick test I made, it seems that animations are stored as arrays, because the performance of animationPhase seems to scale with the index of the animation in the returned array using animationNames, which means the game searches for the animation name first before returning the phase, which is a bit wasteful since object anims don't change indices
if so, using indices means the command works faster, which could save me a lot of performance since I'm using animationPhase every frame for many objects and each have many animations.
some objects (especially vehicles) have hundreds of animations so it could save a lot of performance
made a ticket:
https://feedback.bistudio.com/T165735
Hey is it just me or is the text that is automatically added to html preset files to explain how to use them pretty wonky?
"Drag this file or link to it to Arma 3 Launcher or open it Mods / Preset / Import."
Something like this would props make more sense
:
"Drag this file onto the launcher window or click the Mods tab, then Presets in the top right, then Import at the bottom and then select this file, to import the preset into your launcher."
If there's a chance of it being changed I'll happily throw up a quick ft ticket
that text is even hardcoded without localization. So its very easy to change it
Might not need a ticket, but person to sign off on it is unavailable, please ping me about it on monday
Ye thats what it looked like to me too when I re-discovered it earlier, from the writing it wouldnt surprise me if someone put it in quickly while making it and then never improved it 
will do
not a bug:
https://feedback.bistudio.com/T165747
plz close
lol I thought so π
I said "We'll just ignore it" today.
Indeed not a bug, thankfully Leopard was kind enough to guide me on how to properly set it up.
Could use Windows 11 as option
This creates a situation were it is impossible to script a solution
Eh?
setVectorDirAndUp didnt work?
Of course will work, and this is not even a valid feedback
it does. he wasn't using the model coords for his vectors/dirs
Spam comment: https://feedback.bistudio.com/T165783#2324519
I wonder who spams the forum
Really weird to have a support forum spammed instead of a more gullible site
Bots. They trawl the internet looking for any kind of forum they can sign up to and post their spam on. They don't know and don't care that it's a feedback tracker, they just know they can post on it.
Damn, well they can use text formatting and stuff, so I guess this feedback tracker is open source
It uses a relatively common markdown syntax which definitely does not require access to the software source code to use.