#arma3_feedback_tracker
1 messages ยท Page 38 of 1
@untold sky https://feedback.bistudio.com/T159828
could someone of the devs please take a look of this ticket if it can be done? https://feedback.bistudio.com/T159474
thx ๐
Could a GC8 please stop rebumping the same ticket constantly
sorry but I thought no one noticed it
set to "Acknowledged".
thx ๐ well atleast I know it has been read
I read all new tickets
oh ok thats nice
dealt with
indeed - account was already disabled but the moderator didn't cleanup
how about making the pathfinder run on its own thread and unit waits for the path? just an idea...
a wat that does wat
multithreading
wat
you threw out a buzzword so I threw out one too
regarding optimisation, don't worry, they explored enough about AI
k ๐
ok no more ideas about that then
I'm afraid they would be at best already explored, at worst unfeasible
(and be assured they did spend time thinking about it)
yea, but isnt arma old engine from OFP times? when probably not much MT was done
yep, but then another and another and another team worked on it
things that weren't multithreaded got multithreaded
but some key aspects for stability/time constraints/resources issues
alright
it's not laziness or "didn't think of that" ๐
though what happens these days is "oh yeah this command works, but OH WTF WHO WROTE THAT" and it gets rewritten and gains 800% perfs, and since it is something used very often it saves lives performance
benchmark & structure-wise, I trust Dedmen\ยฎ
ok
small team...
after 8 years, it's good to still see it updated and the community listened to
it's maybe less good not to have a new instance by now in industry's standards, sure
but hey, Enfusion's cookin' ๐
yeah good to see it updated
Cooking because it can't handle Arma 4 ? ๐ ?
pathfinding is already multithreaded
cool ๐
that's your CPU ๐
Yeah it's old ๐ข
it already does...
yeah Leopard20 just told me ๐
I reply faster than I read
I guess Dedmen's write/read routines are also already multithreaded ๐คฃ ๐ง
Some of your pins might be bent there... check your motherboard ๐
๐คฃ
sorry race condition
that's racist
I got a request: is it possible to add the Head Bob from 1st person POV to 3rd person POV? Thanks.
Oh god please no
That's gonna be first class motion sickness ๐คฃ
https://feedback.bistudio.com/T155872#2144204 Any update on this? @finite anvil added more info.
cleared
Lol no.
Not on a saturday
I think displayctrl should show error if null display is passed
I dont know of any command that shoes error when you pass a correctly typed null.
hmm doesnt isnil do something like that?
no
no command shows errors when a Null is passed
ok I must remember wrong then
Yeah. Of course not on the week end.
trying to load config stored in profilenamespace seems to crash arma
is this known bug? I think it has actually been pretty long
it's been 16 minutes come on, give them a break
didnt mean to rush, just wanted to say I may have past problems with this thing
I meant it as a joke, based on "I think it has actually been pretty long" from your last message ๐
aah lol, got it ๐
FT ticket with repro
yes
ok
i havent been able to make it crash but there's definitely something wrong when loading class created in description.ext
trying to load config stored in profilenamespace seems to crash arma
i havent been able to make it crash
wha
I had to stop saving configs in my mission because it would crash
save configs?
i can still make a repro, just not crash
yes custom classes
I don't see why anyone would need to store mission specific config references in profileNamespace which is not at all mission specific :u
well here's the ticket. hope it makes sense: https://feedback.bistudio.com/T159901
its just that the class print is wrong after load. which is probably a big problem if it starts crashing as well
hm, why save the config instead of the classname and values?
well I wanted to be able to access the class later if needed
if I may askโฆ why profileNamespace?
saving mission there
gniiiiii
I might be wrong of course, but I sense something wrong about the design ๐
(not saying that a fix should not happen though)
its just that the class print is wrong after load.
maybe it just reads as configNull?
If it reads configNull its probably not a bug
it would be worth it to say "profileNamespace cannot save classes" or whatever it is on the wiki
it reads nothing, not even quotation marks. but at saving it prints as supposed to
thats to dedmen ^^^^
config classes can be saved. the problem is with description.ext classes
it reads nothing, not even quotation marks.
try this
systemChat str [profilenamespace getVariable "cfgTestSave"]
maybe its nil then
prints []
nil, no?
maybe its null?
then it should print something like <null>
[confignull] prints []
:u
[objnull] prints [<NULL-object>]
is it making any sense?
should be <NULL-config> then
ยฏ_(ใ)_/ยฏ
lol ok
well I already changed my code so that it doesnt try to save those custom classes
just saving the class name
But.. buuut... Game updater is a 3rd party :U
hue hue intensifies
MovesType CfgMovesMaleSdr load time 17623.0 ms
Is it possible to make it fffffffffffffffast?
no
its usually loaded from cache, because its so slow otherwise
that means in most Arma starts it has basically no impact
thanks 
Is it correct behaviour of intersect that this code ...
...
_projectile enableSimulation false;
addMissionEventHandler ["Draw3D", {
private _pos = _thisArgs # 0;
drawLine3D [_pos, _pos vectorAdd (_thisArgs # 1 vectorMultiply _thisArgs # 2), [0, 0, 0, 1]];
}, [ASLtoAGL getPosASL _projectile, vectorDir _projectile, _maxRange]];
private _is = [_target, "FIRE"] intersect [ASLToAGL getPosASL _projectile, ASLToAGL getPosASL _projectile vectorAdd (vectorDir _projectile vectorMultiply _maxRange)];
systemChat format ["Intersections: %1", count _is];
...
```... produces the result shown in the screenshot?
`_projectile` is the rocket (top right corner in the screenshot), `_target` is the Slammer.
(the problem is that count _is is 0 in the situation depicted in the screenshot)
use lineIntersectsSurfaces
also this code doesn't draw correctly anyway
(it always draws the same line)
That's... what it should do.
I thought you wanted to draw a line to the intersection point...
I specifically freeze _projectile (and the line) to see how my vectors look at the point in time when intersect determines that there is in fact no intersection with _target.
intersect only works if there's an intersection with a selection
Ah, now this makes more sense to me 
@vocal abyss it works fine for me
the green icon is the one from intersect command (the selection is shown as hint)
code:
onEachFrame {
_p1 = player modelToWorldVisualWorld (player selectionPosition "proxy:\a3\characters_f\proxies\weapon.001");
_p2 = _p1 vectorAdd ((player weaponDirection "") vectorMultiply 20);
pp11 = [[_p1, _p2]];
{
if (getModelInfo(_x#3)#2) then {
_inters = [_x#3, "FIRE"] intersect [ASLToAGL _p1, ASLToAGL _p2];
if (count _inters > 0) then {
hintSilent str (_inters#0#0);
pp1 = [_p1 vectorAdd (_p1 vectorFromTo _p2 vectorMultiply _inters#0#1)];
}
}
} forEach lineIntersectsSurfaces [_p1, _p2, player, player, true, 1, "FIRE", "NONE"];
}
but as you can see, there's no green icon here because there's no selection there
Yea, makes sense.
but there is intersection using lineIntersectsSurfaces:
btw could you tell me how? I haven't been able to figure it out.
I mean there is a proxy selection: "proxy:\a3\characters_f\proxies\weapon.001", but it's not where it should be...
it is exactly where it should be according to your screenshot?
I wanted the muzzle end 
the weapon is in players right hand, looks correct on the screenshot
well then you need to get the weapon model, and find the relative position of the muzzle end
yeah but how?
there is no command for that
selectionPosition
I mean
get the weapon model
"model" entry in CfgWeapons
oh you mean create the model using simple objects?
yes
I guess unit GetWeaponMuzzle "weaponClass" is in someone's head ๐ฌ
configfile >> "CfgAmmo" >> "ShellBase" >> "soundFakeFall0"
https://ibb.co/k2sQr7Y
what is 1000? distance?
https://community.bistudio.com/wiki/CfgAmmo_Config_Reference#sound....5B.5D
biki doesn't mention this
And I noticed when firing from artillery beyond a certain distance, there is no sound of the projectile falling
It's about 1 km
yeah why didn't I think of that one?! 
I thought you meant there's a direct way using selectionPosition 
@untold sky I think selectionVectorDirAndUp may be broken
I'm holding a pistol and pointing it straight forward
but why does the command return that?
it worked fine for the rifle
launcher was "rotated" by -90 degrees ccw around the X axis
and the pistol one is just...
or is it because those selections are actually rotated?
yes, this is the distance when the sound is heard
but I mean that there is no sound at all if shoot further ~1 km
I created a ticket with this https://feedback.bistudio.com/T159917
pistol doesn't have its own bone and it is glued to rightHand instead. I guess it is a relative vector
huh. that explains many things!
thanks!
yeah the launcher is just rotated 90 degrees
that one is easy
so is it returning vectorDirAndUp from t pose?
so I'll have to find the transformation matrix for the pistol...
no. but it's not pointing correctly. I was expecting Dir to be the same as weaponDirection, but it's not
the dir and up are just pointing to nowhere...
(the one pointing up is dir, and the one to the side is up
)
how does it work with right hand?
sadly this doesn't work properly.
when your aim is not level (aiming up or down), the selectionPosition command has a strange shift, which causes the muzzle point to be calculated incorrectly.
there's no way to measure this shift either 
so I guess a command is necessary
I tried memory only
It should give you exact position of the weapon proxy in your hand
and the weapon is rendered inside the proxy
what do you use for rotation?
selectionVectorDirAndUp or however i called it?
yes
you mean "ViewGeometry"
nono
or one of those "numbered" lods?
oh
ok
because I tried "ViewGeometry" and it didn't work either
I'm gonna try the rest
oh yeah! it worked! thanks! ๐ (I used res 0)
btw I encountered something strange:
player selectionNames 0 returns []
also, player selectionVectorDirAndUp ["proxy:\a3\characters_f\proxies\launcher.001", 0] returns [[0,0,0], [0,0,0]]
but how come this works?
selectionPosition [player, "proxy:\a3\characters_f\proxies\launcher.001", 0]
the other two commands seem to suggest there's no such selection in LOD 0 
huh
turns out LOD 1 is the best lod
it already has the transformations applied
the pistol works fine too (no need for my rotation matrix anymore)
w00t w00t
I wonder if the 001 in the proxy names (e.g. "proxy:\a3\characters_f\proxies\launcher.001") was actually referring to that all along! ๐
could be the fact that this script command is taking literal LOD number
most of the characters have something like that
you can start from 0 too - engine doesn't matter
and some older characters/models could have some custom LOD numbering. In past LOD number was also used to determine how fast LOD change is happening so people were doing LOD like 0,0.5,4,12,24
how about elseif command? I bet this has been asked before ๐
yes. the answer was no
but you just said yes!
btw Ded, remember when I told you that attachTo's followBone was not working correctly when you aimed up or down? so I guess that too uses the memory LOD right? I remember you asked if we wanted more params for attachTo, and you went with followBone. What about a LOD parameter too?
weapon jam command would be cool ๐
you can script it
really? take away bullets?
aah but thats for player. I meant for all AI units
Please stop constantly requesting garbage... ๐
garbage? weapon jamming is real in real life
thought arma was a simulation ๐
just use ace
Its just annoying getting these constant requests from you that are either
nonsense
were already denied
are not viable
Arma is in end of life support stage, how did you get the idea we were gonna add a whole weapon jamming system now
they are not nonsense but I didnt know you are closed for new ideas
Speaking of annoying requests I have been holding back on this one, but would a behaviour EH make sense?
maybe you should read the entire message 
closed for new ideas
no. but the problem is that weapon jamming requires a change in game mechanics, which won't happen anymore
Arma is end of life (EOL). So only fixes and nothing new
That was requested before and I think the answer was "maybe". But don't quote me on that.
Behaviour and roe ehs would be nice indeed.
EH behaviour will happen
โ @daring wagon

Thanks.
hm, would it be worth a ticket to ask for a way to replace e.g onPlayerKilled.sqf event scripts by (scripted) event handlers?
When the game is changing it's weather randomly if the values are e.g. from 0.21 to 0.58, when it reaches 0.58 it immediately calculates overcastForecast, but if the overcastForecast is 0 or 1, when it reaches this values it takes longer to come up with a overcastForecast and stays at 0 or 1 for much longer than any other values.
Also all weather changes take the same amount of time, e.g. 0.18 to 0.1 takes as long as 0.9 to 0.1.
If you could review the code would be awesome.
I ran this for observation:
[] spawn{while {sleep 0.1; true} do{hint str [overcast, overcastForecast]}};
setTimeMultiplier 120;
Yeah. I believe all "event scripts" should have mission EH counterparts
e.g.:
onPlayerRespawn.sqf -> "PlayerRespawn"
very useful for mods
Yeah I understand, just didn't know that was the case with new features
maybe weapon jamming for arma 4 then ๐
Arma 4 is going to be about medieval wars and fruit picking, so only sword and strawberry jam I'm afraid
Oh thanks for mentioning jam, just noticed I forgot my breakfast
feed
for we need alive dedmen wait a minute
Strawberries weren't cultivated until the 18th century. Already BI is ruining Arma 4 with military inaccuracies!
oh dang, I should have known that particular NDA was for me to not spread bs around
lies!
Wild Strawberries are documented as early as 1324!!
So the fruit picking will be part of a foraging system, not a farming system?
Time for me to make a 2 hour youtube video with this news
a taming system, as they are wild
buh *ahem* enough #offtopic_arma, back on topic ๐
you guys keep offtracking, it's unbelievable!
Clearly we don't have any bugs to report. Good job team!
@untold sky when you have time, could you check why deleting super simple objects causes the game to freeze for several seconds?
To test, just use the "code performance" button to run the code many times.
Simple object: No freeze; returns ~0.3 ms
deleteVehicle(createSimpleObject ["Land_Axe_F", [0,0,10]]);
Super simple object: returns 0.017 ms, but the game freezes for several seconds during the execution.
deleteVehicle(createSimpleObject ["A3\Structures_F\Items\Tools\Axe_F.p3d", [0,0,10]]);
if I don't use deleteVehicle, there is no freeze at all, and the code performance shows nearly the same number. so the problem is definitely deleteVehicle
there is nothing in the log either
Not sure what others think about this topic.
De-Ebo Enoch data in base A3 Platform
https://feedback.bistudio.com/T159955
TacOps too (everything but missions)
afaik de-EBOing single-player-only content will never happen
De-EBO will happen whenever it happens. And I don't think requesting it to be done has any influence
made a ticket:
https://feedback.bistudio.com/T159956
btw, I also noticed other bugs (e.g. deleting a super simple object that had class = pond never deletes the "water" it creates)
it seems like deleteVehicle can't handle super-simple objects properly
pond objects I might just label as "unsupported" ๐
ok, but what about the performance thing?!
I didn't say anything about that
ok 
I found an interesting bug
when you scale an object, vectorDir, and vectorUp return a vector multiplied by the scale
however, as per wiki, they should return a normalized vector
oh noice!
while benign, it should be normalised indeed
vectorModelToWorld is affected too
but I guess that one is not a bug (or is it?)
I reported this bug a long time ago, but I never knew why it happened
it just hit me now! ๐
I'll update it
wiki is wrong ๐จ
scale is stored in orientation
so you should observe the same when calling vectorDir/vectorUp on scaled terrain objects
so won't fix right?
no
https://feedback.bistudio.com/T150422
can someone close this as won't fix?
hah! closed it as not a bug ๐
REMOVED
it was active since 2020-02-28 in DayZ tracker, ofc nobody thought about reporting it so it had a dozen of posts
fought back by sending the invite mail multiple times. probably does nothing, but it soothes me and potentially pisses them off.
quick question - I've got one thing that I wanted to suggest which is not a bug, but more of a "feature/update request", as it would be a nice-to-have thing from a modder's perspective. Is it still a matter for the feedback tracker? The Maniphest really only allows bug reports and it's not quite that
yes
The Maniphest really only allows bug reports and it's not quite that
it is called bug report but in it you can add more details such as "Feature request"
Lovely, thank you
https://feedback.bistudio.com/T159957 aaaand I did it, hopefully it makes sense
closes it
removed
https://feedback.bistudio.com/T159972
Spam ๐ฆ
thanks, gone!
thanks! all gone โ๏ธ ๐งน
Spam comment https://feedback.bistudio.com/T159895#2219901
gone, thx
robux time! https://feedback.bistudio.com/badges/view/1/
wat
plz close this as Not a Bug: https://feedback.bistudio.com/T159956
it was my own mistake 
sorry @untold sky
done
But it is a bug
Just not something you observe in normal gameplay. But its still broken and bad
Well what about cases where there can indeed be a lot of object deletions, such as a heavy fire fight? (many bullets fired and deleted)
or ambient animals spawning and despawning?
Thats why I asked you about that in the ticket
When does that legitimately happen in normal gameplay outside of performance testing?
But didn't get an answer beyond the ticket just being closed
Well I wasn't sure if it impacted it that much
But I tested it by creating a fixed number of objects (simple/super-simple/normal) and they were indeed slow just as much. So I figured it's not a bug
At least not what I thought (I said it can't handle super-simple objects efficiently, which is incorrect)
Its a performance inefficiency -> bug
ok 
I just figured that's the way it is
well you can reopen it can't you?!
If not I'll open a new one
the internal ticket stays open
but I wanted to set the FT ticket to low priority unless someone comes up with a reason why this is important
Resolve is fine I guess, but "Not a bug" is wrong
Well what about cases where there can indeed be a lot of object deletions, such as a heavy fire fight? (many bullets fired and deleted)
or ambient animals spawning and despawning?
These aren't good enough right?
I don't think so. Even heavy firefight doesn't have more than like 5 bullets per frame being deleted
How much do you think it can be optomized? (in %) if not much it's indeed not worth it
btw why is it slow anyway? does it update the "available ids" or something?! 
recalculates bounding sphere for all objects in the grid chunk after every delete
Simple O(Nยฒ) if you delete 5000 out of 5005 objects in the chunk
is each grid chunk the same size as the cell size?
no
path finding I think
no
emptyPosition?! ๐
no, I have no idea
It has to do that after a delete. But really only once after all the deletes are done. Not after every single delete when you delete 5000 in bulk
hell yeah
can you do it in the "next frame" if some object was deleted? 
(depending on why it does that anyway...)
isn't it because of the line intersection things? (plus other stuff, like AI visibility checks)
collision maybe?
yeah that too
I really don't know ๐
how would that be any better?
ah you mean the bounding box
No that would mean during one frame everything will be broken
try to disable it, just to see if something actually breaks? ๐
(j/k even though I'm curious, if one doesn't know, one doesn't touch ๐ )
Anyone know if there is a ticket about.
When you are in eden
Have different vehicles in cargo of some other vehicle (for example multiple small vehicles in a big plane)
You save and reload the mission, suddenly the order of the vehicles in cargo changed randomly
https://feedback.bistudio.com/T160044
Here it is sir @sterile nova
closes it
gnii
The thing I can imagine with this value: heat haze and winter visible breath, and rain/snow but engine driven? It's okay if it is โnot worth to torture RV engine anymoreโ
heat haze would be great
I scripted that
but it would be way better engine/render-side yes
yet people would complain about AI ๐
Isn't Tanoa in the tropical region with barely any annual temperature range or daylight variance anyway?
Indeed... but still, Tanoan summer (Nov-Feb ish) temp in the current config doesn't make any sense
not anymore ๐
https://feedback.bistudio.com/T160059 reported by @stark pier - setUnitLoadout adds an entry to the JIP queue
How about this idea?
Make the command SP only. For MP, do this:
add a variable to description.ext like allowAimCommand = 1 or something (disabled by default)
this variable sets an internal (engine) global variable on the server, which will be sent to the clients when changed.
also add a similar command that is server-exec only and cannot be remoteExeced at all (and cannot be executed by remoteExecing codes either, by checking the remoteExecutedOwner), which can modify this variable during the mission.
you can also add other restrictions, like don't let the command run if the unit is being controlled by a player in MP (even if allowed by server)
you can also keep a hashmap of recent player-controlled units in format: <net_id, float>, where float is last control time (or maybe system time if mission time is not good enough)
if the unit is player-controlled, or was controlled in the past 3 seconds, disallow the command
also even if that's considered an "aimbot recipe", the command for setting the eye (look) direction is not ๐คท
use cases:
- controlling AI
- making cutscenes
- I've always wanted to fix my eye direction in helicopters when I use freelook (because I don't have TrackIR), which can be done using such command
oh wait. that was item 2 on my list 
1. a command for setting weapon/aim direction (units and vehicles). maybe with an additional "commit" command if it's not possible in one frame. (similar to controls/cameras)```
player setDir (([player, _target] call BIS_fnc_dirTo));
sorry can't get it to format
but thats fairly close to what you want probably, except it doesn't control camera
I use that snippet for when players buy vehicles, it stops them from guessing what direction their car spawned in, and stops them from guessing which one is theirs, not sure who made that snippet not mine fyi.
wat?
Try
player setDir (([player, _target] call BIS_fnc_dirTo));
it should do what you wanted in line 1
not so much line 2 but line 1 maybe
it sets the players dir to whatever it is you want to face
yyyes, but if the target is above or below, fat chance
PS: isn't BIS_fnc_dirTo obsolete?
done! thx
@untold sky
openMap [true, true]
this code now wont open the map just prevent "M" to be used. revealed In dev build, dont know how about in stable.
tried the following```sqf
[] spawn {
sleep 0.5;
openMap [true, true];
sleep 5;
openMap [false, false];
};
fixed already
w00t w00t!
regarding the deleteVehicle thing, seems like the actual deletion doesn't happen till the next frame, correct?
because lineIntersectsSurfaces still works after deleting the object 
so I have to use something like _obj setPosWorld [-1e4, -1e4, -1e3] or something before deleting it
just the current/next frame mayhaps?
current yes
btw why does the game still want to "update the grid chunks" when I delete an object at [-1e4, -1e4, -1e3]?!
shouldn't it not freeze anymore, since the object is in nowhere?
yes, its "marked for deletion"
its still somewhere, just a weird position
but shouldn't it check if it's within the map area first?
or are there still "grid chunks" there?
isn't [[-1e4, -1e4, -1e3] just [-10000, -10000, -1000]?
yes 
soooโฆ it's still within the calculated area, no?
https://community.bistudio.com/wiki/Position
-50km, +500km thing should be "the grid", no?
ยฏ_(ใ)_/ยฏ
no thats not how works
yes. EVERY object has to be in grid, its a quadtree
its a quadtree
so the game uses it for finding empty positions + nearest objects stuff? or do you still not know why it does that?
yes-ish
whats with spam lately, removed about 8 just today
this is what is says on biki, no?
yeah, sorry I didn't read that 
btw the setpos 0,0,0 error in diag exe is related to this white helper object that is spawned under the player's feet:
when you're touching the terrain, the object is teleported to [0,0,0] I guess, and causes this error spam
I'm using diag_drawmode "roadway" btw. that's why the helper appears
no
or Enhanced Movement
I do use Enhanced Movement yes. but why would it appear?
it only spawns its helper object when you use the climb function, which I didn't in this case
the object appeared when I walked onto the rock, then disappeared when I stepped onto the terrain
the error occurs in vanilla too
exactly when you do that (walk on roadway LOD -> error disappears; walk on terrain -> appears)
I'm guessing the game spawns the helper object to orient the player's feet like it does with terrains (a workaround to simulate it)
btw is it a bug that you can run over units with attached objects? 
I've attached this vehicle to myself and I'm just running
the vehicle doesn't collide with anything else (not even other vehicles). just the units
no
So I encountered something weird.
Why is the land contact of a simple object not the same as the original object?
is it stored in the object's config?
(super?) simple object is "just" the 3D object, like, no ground contact data iirc
so how come its getPosASL and getPosWorld are not the same? ๐ค
getPosASL uses land contact afaik
dunno ยฏ_(ใ)_/ยฏ
simple no has landcontact I think
so what does getPosASL use then?
_test = createSimpleObject [typeOf heli, [0,0,0], true];
_offset1 = _test worldToModel ASLtoAGL getPosASL _test;
_offset2 = _test worldToModel ASLtoAGL getPosWorld _test;
deleteVehicle _test;
[_offset1, _offset2, heli worldToModel ASLtoAGL getPosASL heli]
returns:
[[0,0,-1.17551],[0,0,-1.19209e-007],[0,-0.00278322,-1.77914]]
```for pawnee
magic
PositionWorld
Similar to PositionASL, however Z is measured from the sea level to the the model centre [0, 0, 0] of the object, rather than transformed boundingCenter or land contact vertices.
I'm gonna guess it's "transformed boundingCenter"? ๐
btw this equation that Anim Viewer uses is wrong:
_duration = if (_speed != 0) then {abs (1 / _speed)} else {0};
it should be:
_duration = -_speed;
if (_duration < 0) then {_duration = -1/_duration};
because:
speed <= 0 : -duration
speed > 0 : speed
This issue is very very interesting:
https://feedback.bistudio.com/T152438
I wonder if it has something to do with the deleteVehicle and grid update thing?
it causes a really noticeable freeze, and you only fire one bullet
Subscribed and comment, I hope it is dealt with eventually
There is SetFace error: class ... not found. RPT message. Is it possible to add player name to it?
Would it make sense to modify append so that it returns the modified array?
Right now I have to do ...
_a = ["A"];
_b = ["B"];
_c = ["C"];
_all = [];
{ _all append _x; } forEach [_a, _b, _c];
```... but it could be much more convenient:
```sqf
_a = ["A"];
_b = ["B"];
_c = ["C"];
_all = _a append _b append _c;
```(I have no use for the original `_a`, `_b`, `_c` arrays)
I think by SQF standards, when a command returns something to you it's always a copy.
wouldn't this cause a confusion?
altho I do agree that it is convenient
why not use +?
true ๐ค
I don't think the fact that it makes a copy matters in this case (it copies the array itself, not the elements, so same difference)
I think by SQF standards, when a command returns something to you it's always a copy.
eeeehhhhh... often, maybe?
Because + is slower as it creates a new third array, and I have no use for the intermediate arrays created this way either.
did you measure the performance?
I think they'll be the same
I didn't confirm it, but https://community.bistudio.com/wiki/append claims that + is slower ๐คทโโ๏ธ
it can be slower. but not sure if it is for what you do... ๐ค
obviously it will be slower as it makes copy, you want append return copy henceโฆ it will also be slower. internally + is append with copy so at least currently you have a choice to modify array for speed or return copy with +
I don't want append to return a copy, I want _a append _b to return (the reference to) _a.
merge has the same drawback as append, it also doesn't return anything and therefore can not be chained.
@untold sky Can this be changed to be able to override the action if you return true?
https://community.bistudio.com/wiki/addUserActionEventHandler
https://community.bistudio.com/wiki/a_%3D_b#Notes
is making a = b return nil ticket worthy? fixing this "pure null" value could be nice & easy I guess (that or return the assignation value, allowing a = b = c = d = 4?)
no
no
nice yes, easy probably, safe 
โฆah, I forgot that safety/stability matter thing ๐
btw, can the "event" be changed to "up"/"down"/"analog"?
activated and deactivated are too long 
and prone to mistype
even "on"/"off" works 
I can't wait to see onOn events ๐
onOff
speaking of analog, due to analog fluctuations we can expect it to even run every frame yes?
A activate doesn't necessarily correspond with a "down"
you are binding events to the action, not to a key, a action doesn't go down or up
yeah but the point is that they're long...
hard to type
also you always wonder
is that Activated or just Activate? 
If you had told me two months ago maybe. But not now
Just write on wiki
Just read on wiki, and you only write it a handful of times anyway
Beware of shy spambots. They post innocent comment like Thanks, this was really helpful, this was useful etc. Then come back much later end edit in ads with links
argh new techniqueโฆ it's not half-stupid actually
So Ded, I feel like my game has become ~20% slower (at least scripting performance, based on one of my codes)
iirc, even a single scripting command, like player, had an exec time of 0.00055 ms. now it's 0.00077 ms
either that or my system has gotten slower.
is there a way to go back to a legacy Dev build to confirm?
Legacy dev
Which I believe doesn't exist
yeah. I think the performance decrease happened since the last dev branch update. I'll have to test an older build to make sure
Ah, yeah no way to get previous dev version.
I hope the devs can send me one themselves 
I don't remember the version 
If it takes too much time to build, plz send the one from two builds ago (to be safe)
I'll see if I can find the version
versions are listed in dev branch changelog
we ran a dozen builds last week, no idea which was released ๐
Yeah I was just gonna look there 
I think 147890 should be good enough. I didn't see any major changes
Wait I think DnA forgot to update the version number for the latest one!
147890 is 13th july 
Yeah 

I was using the diag exe. That's why it had gotten slower since my last test 
I tested the new build that Dedmen sent me and it was faster, and I was about to report it...
Then I realized that I was using the diag exe when I tested the code yesterday. So I tested the latest normal exe again and sure enough, it was working perfectly.
sorry 
๐๐ฅ ๐จ
but anyway, the takeaway from this story is that the diag exe has like 20-25% lower scripting performance (I didn't know that...)
again, I'm really sorry 
@sterile nova
directionStabilizationEnabled says it returns nothing. it should say bool
corrected by KK
Crash ticket: https://feedback.bistudio.com/T160224
Took me a while to figure out ๐
umm, was serverNamespace added intentionally?!
it returns nothing! ๐
nvm, works in MP 
but doesn't return anything in SP (isNil {serverNamespace} is true), and the signature is:
nular__servernamespace__ret__nothing
should return namespaceNull :p
I think @solid marten wanted to remove it? Together with the changed server eventhandler stuff
I don't get the usage though, as in "specific namespace for network communication that has nothing to do with missionNamespace and runs in parallel?"
Leave it, you can set variable in it from server event handlers and read later in mission
Both return types corrected rev 148061
so nil in SP & on MP clients?
didn't test on clients, but I'd say yes
it seems to only exist on the server
@solid marten nil wanted & confirmed, I can engrave it on the wiki?
just add server execution only sign
I did both, but I want to know and document such behaviour
https://community.bistudio.com/wiki/serverNamespace confirmed ๐
was there a plan to have some reliable input detection for key combos [and ability to block that input] (as replacement for inputAction / keyDown)
ie disableUserInput ["reload",true];//["userAction",on/off bool]
was there a plan to have some reliable input detection for key combos
already done
and ability to block that input
no
could be useful if done via handle EH
already done
that is?
would be the new https://community.bistudio.com/wiki/addUserActionEventHandler stuff
Why does ```sqf
systemchat format["%1", nil >> "test"];
why not a smol bug mayhaps
stahp doing that
but that could be a sign of bigger problems
true, ticket pwease
It returns "config" typed nil.
And its stringiziliatization is implemented as printing the this ptr.
I think it should be <config> ?
i thought it was weird printing pointer addresses
also the congfigNull still doesn't print anything at all when it could print <NULL-config>
I think I wanted to fix that right? Apparently I forgot 
you could do the same for namespace while at it, i mean make it print which namespace it is when you stringify namespace
@untold sky
I don't know if the namespaces know their own name.
But might be worth the 8 bytes of memory space ๐
is a command that allows setting the anim progress possible?
as in
_unit setAnimProgress 0.5; //% of total anim duration
jumps halfway through the current animation
and maybe with _unit setAnimSpeedCoef 0, it will allow manual playing of the anim
Anim, that can be set by switchMove or such commands?
https://feedback.bistudio.com/T157753
Yeah, I had similar wish
...Oh you already subbed to this
Does anyone know why some damages "get passed" the handleDamage EH?
Mostly damage from falling, terrain, water, etc.?
like this issue:
#arma3_scripting message
I'd noticed it before too. it's kinda annoying.
can this be fixed so that _veh addEventHandler ["HandleDamage", {0}]; acts like _veh allowDamage false?
https://feedback.bistudio.com/p/Aleksey2306/ this guy keeps on making public tickets reporting exploits
Then tell him to create private tickets 
I don't want to clutter it here, but i updated that linked post with some new details.
this ticket has become obsolete: https://feedback.bistudio.com/T159901
not sure if I can close my own tickets =/
nope
want me to close it for you?
yes please

thx ๐
IMO if you are developing a mission for MP and you want to check how it behaves both on client and server, it's better to design it in a way that allows self-hosting, in that case you're both client and server, you can see what's happening with both client and server. And you can test it easily from the editor.
I'm hosting a server locally on my LAN/PC if that's what you meant
then when you need to see a case for DS+client, all right, DS is like regular self-hosting for your sqf code (well, most of it?) except it has no UI, so it makes no difference for your code. So you can run the server through editor and run a second arma instance (with different profile! set it in launcher) and use it as client.
It is VERY useful to actually not just stare at DS's console log and figure out what's going on, but to be able to access the debug console and whole SQF environment of the server, that's why hosting from editor is more useful for testing than hosting from a dedicated server
Wow, that's very useful
So I can have two clients on the same machine without enabling loopback?
I'm not saying that you are doing it wrong, you should test how it runs on dedicated server, but not constantly IMO (but rhather when you need to see some specific things?), for quick fixed and prototyping it's better just to do as much as possible from editor
just create two profiles in arma main menu, then in the launcher select different profiles (it won't let you start the game with same profile AFAIK), and that's it
without enabling loopback?
well there is no such thing as 'enabling loopback' for regular arma3.exe client I think, it's only for dedicated server
So I can indeed have multiple instances of the game running if I use different profiles. The more you know! 
Thanks for this very valuable piece of info ๐
yeah np
no idea why people don't know this so often. Actually I figured it out pretty late too, when I was wondering how TF do I iterate my changes faster without restarting the game every time, rebuilding, etc
Yeah, that is/was my issue too ๐
I make a lot of small mistakes when I code (I'd strongly prefer a full blown IDE with a nagging syntax and type checker due to it ๐ )
So when the iteration time is so huge, it becomes a neverending fight against bugs and that's what kills my motivation to create stuff in Arma
Gib static typing plz
Hm? It closes the first Arma instance when I launch the game with second profile @naive ledge
what ๐ค It definitely worked for me across a few PCs at least
Do I need to have a second PC around and use it to launch the second instance?
no, it was meant to work on same PC
Yeah
It closes the first instance
Maybe it was used as exploit and was disabled?
Servers check for double UIDs by default though... I think
I just tried again and it works ๐คท are you sure that you selected different profiles in launcher?
Ha! You have debug mode on
yes
Didn't even know that such option exists... I've spent too much time in the Arma2verse ๐
and I was always running it without BE to make my DLLs work ๐
yes that's right, with BE it closes the other instance
There we go! Thanks ๐
I'm stunned by the fact that I have two instances of Arma running simultaneously while they're even on the same server lol
Feature request: unit disableAI "FIRE" -> making AI unable to fire the current weapon (by themselves, not affecting scripting commands such as forceWeaponFire)
Possible? Not possible? Won't do? Might do? Should I make a ticket? etc...
I made the ticket anyway 
https://feedback.bistudio.com/T160263
only the current weapon or no weapon at all?
any weapon it tries to fire obviously. they can only fire their "current weapon" tho 
https://feedback.bistudio.com/T75521 can be closed?
https://feedback.bistudio.com/T135718 @untold sky let's close this compile memory leak ticket?
https://feedback.bistudio.com/T160265 created a ticket to add CfgFunctions an additional attribute serverOnly so server-side functions are not compiled on client - for mission's CfgFunctions
if you donโt compile them you have variable that is free to be compiled for you with malicious code. i say no
oh you mean, TAG_fnc_TheFunc is an "available name slot" on the client?
oui
hahaha ๐
well I see the concern, but a mission designer that only wants to run code on server's side should be able to
then: can the client compileFinal {} ?
what's the point?
If you only want to run the code on the server run the code on the server.
saving client's compiling time
That's miliseconds.
if so, then it's unworthy opti
if it's a full framework though, can't functions take seconds to load?
here's to hoping config entry reading is not eating that won time ๐
They can use preprocessor to only enable their code on server, so client will compile empty file
or simply, use preprocessor in description.ext and simply hide the whole CfgFunctions section of stuff you don't want
then the client will outright not see it
โฆcan we preprocess isServer?
doesn't matter, just define whatever macro via start parameter
Yes?
They can just design it to remove the functions on client if they want to then
tl;dr if you need such optimizations there are tools to do them.
99.9% of mission designers do not need them.
it's that I didn't get
but I don't see themโฆ? is there a __IS_SERVER__?
No
that it's quite rare, I get it
that I can do it mission-side, I don't see how? unless there is a way to #IF some SQF
(only works with a dedi then, which is also a bit portion of servers I guess)
just put #ifdef someMacro around your CfgFunctions or your sqf files
and only define that macro via server start parameter
so not mission-side
Ah I see, serveradmins would need to adjust for the mission
I mean by that "no need to tinker with server settings" yes ๐
out of curiousity, would this work?
#if __EVAL(isServer)
// stuff
#endif
well, apparentlyyy not.
always go in the else, I assume isServer is called too early ๐ฌ
As I repeatedly say.. __EVAL is NOT preprocessor
yet it's THERE, so how can you explain THAT ๐
https://community.bistudio.com/wiki/PreProcessor_Commands#EVAL
I explain that by me repeatedly telling you and other people that its NOT preprocessor and should NOT be listed on the preprocessor page.
But noone wanting to do anything about it
where does it apply then ๐
so something like this? ```cpp
name = __EVAL(if isServer then { "server" } else { "client" });
don't know if eval supports if then
its somewhat limited.
but yeah something like that
yes
okido, will try to figure a split ๐
but on the other hand, it pisses you sooo ๐
Can somebody else confirm that allowProfileGlasses = 0; fails when the profile glasses are set to None?
I have mine set to None and the Editor loadout's goggles get removed despite allowProfileGlasses = 0;. When I changed my profile goggles from None to something else it worked as intended, retaining the Editor loadout facewear.
So anyone interested in fixing this bug?
The caret jumps to the next line when it reaches the end of a line, which causes this behavior:
Yeah. Super annoying ๐ซ
I already looked at that but my thought was "meh"
so yeah make me a ticket, I already know where to fix it
I have this:
https://feedback.bistudio.com/T158384
But it also lists other issues 
oooh, the reason why tickets should be split
(I am guilty of that as well)
If you want I can make a separate one
He just fixes all issues ๐
There is a possible to add "isRestricted" param for
getAssetDLCInfo
or can we already return it with some command?
what is "isRestricted" ?
hm. I'd think only actually restricted items even have AssetDLCInfo?
isn't it isAvailable? vs isOwned
but, yeah, there is this in-game issue where you don't know if such item is DLC-restricted or just DLC-added
#workthoughts:
Is it possible to add these to attachTo:
- An arugment to make the attached object movement in render time scope
- An arugment to make the attached object's rendering the frontmost (if it is attached to the
cameraOnand the camera is in internal)
hope dis makez sens
yes but eh, don't think so
but but use render time scope positioning
pwetty pweaaase
pwetty pweaaase
pwetty pweaaase
FYI, I actually thinking about this because to like, โattach an epipen in one's hand, and see it in the FPPโ
Is it possible to get ellipse and rect marker border coordinates?
Border coordinates?
I meant want you to describe what is it not +
you mean "marker X, at angle Y, get border distance/position" or something?
I want to spawn objects at perimeter of zone marker
Isn't it possible right now?
ah yep, it would be nice
but you can still find that with maths
is a ticket to make BIS_fnc_kbTellLocal read missionConfigFile worthy? for Conversations
https://community.bistudio.com/wiki/Conversations#.bikb_BIS_fnc_kbTell_example
Spam (on a restricted project!?) https://feedback.bistudio.com/T160302
oh, Arma moderators probably can't touch that ๐คฆ
fixed, thank you
At ACE GitHub dedmen said that there is an option to create and use custom ropes. But I canโt find info at wiki where to define classname on rope creation. Is it wiki outdated or Iโm blind?
use discord search and search for "rope" in #community_wiki
is it possible to have a command that sets the angular velocity? (in rad/s)
_obj setAngularVelocity [_wx, _wy, _wz];
and a getter command?
currently there's no universal way to rotate objects. you have to do a loop every frame
there is that torque command, but it only works on Physx objects
(also PhysX is unreliable
)
made a ticket:
https://feedback.bistudio.com/T160321
ping - can a ticket about SQF edition be worth it?
also another question: does the engine have a way to set the acceleration? ๐ (except for PhysX objects of course)
if so maybe a command for that too?
using a combination of velocity and acceleration commands one can "fix" the flying tanks bug I believe... 
dealt with
Even without running any mods, I'm experiencing the same issue. As I was able to revert without losing much and things looked fine for the time being, I didn't think much of the corrupt savefile.
that was the comment notification I received.
when I came to see, the message contained an "apk" website link between "any" and "mods"
sneaky
Thats basically the same behaviour as on BIF
on BIF we even have some bots that post normal stuff and wait several days before they edit it to spam
but its pretty easy to detect.
They often post on the same threads, so when you look at it as moderator you see half a dozen banned spambots above. And at the bottom a brand new account from India posting "Thanks that was very helpful"
Instant bonk
If only we could see IP Addresses on FT.
The BIF ones are 80+% India
Then Vietnam and China
Spam comment https://feedback.bistudio.com/T160056#2226118
gone, thx
I saw this on the old feedback system a few years back but it appears nothing ever came of it, is anyone aware of whats going on with the VTOL max thrust glitch? (If you are in a group that you are not the leader of and the vtol engine is on the thrust will remain stuck at 100 and you're unable to do anything about it besides leave the group)
(it was reported and is known iirc?)
repro would be:
be #2 in a group
pilot
thrust is locked?
be in a squad and not #1
get in a VTOL as pilot
turn the engine on
I've had it happened in profiling
my setup:
placed player
placed unit
grouped player to unit
placed a V44
start
get in
turn the engine on
everything is normal.
so the above is not the issue case or is not a proper repro
pretty sure it needs to be a group where the player is the squad lead
ah, leader but not #1 indeed
nope, same
(same setup + group player selectLeader player)
let me rephrase that
What I was trying to say was that the squad lead had to be a player (on a server), and you're in that squad as another person
having a AI as the squad lead doesn't cause it
ah, ok
I was able to quickly pop onto a random server with people on and recreate it in a few seconds
its been a issue for years and is commonly joked about amongst the pilot community
gone
Bad texture format in GetPixel
Unsupported format palette
Unsupported format palette 8
any chance to get more context on these please
Bad texture format in GetPixel
No won't. Too expensive. But it triggers gamecrash, the crashdump might know what texture it was
Unsupported format palette 8
You have a texture in PacP8 format, that shouldn't exist anymore and will trigger gamecrash, the crashdump could tell what texture file it is.
Unsupported format palette
doesn't exist? there is only the one with the8
got'em all, thx
the above was generated during that fpsHeatMap script - no game crash for this (had some mem leaks in Eden but not here)
could it be from the rendering issues of infinite terrain?
eh, the init.sqf was generated?
Ah the script generated that
The messages you mentioned are all hard fail messages that should trigger gamecrash
this doesnt trigger a game crash either tho as far as i can tell
the Eden crashs i had was zooming out max, have the infinite terrain in focus, select a marker, and select "go here"
this seems to lead to a mem leak and eventually the game crashes with out of memory
Eh atleast I think it should 
so basically you are going far outside of terrain bounds?
I fixed that two weeks ago, not sure if its on dev branch yet
not sure as i couldnt see the marker anywhere
so it may very well be outside the visible area at least
you can try to disable clouds
if it fixed the crash, then thats the thing I already fixed
via video options?
yea
ok will test and report back tomorrow
Ah yeah seems like it doesn't crash indeed, it just prints the error message 
The place where the error happens doesn't have context about where it came from.
And I cannot add context there its too expensive (literally getting a pixel from a texture)
Check if you have any other Bad texture format. in your RPT, that one I could add the filename to.
If I see right that should always be logged if you actually have bad textures in a paa file
copy
c:\bis\source\dev\futura\lib\d3d11\txtd3d11.cpp(3239)(func: DX11::TextBankD3D11::ReserveMemory) [AssertNoDlg] MainThread=14424 item.texture->_levelLoaded<item.texture->_nMipmaps
c:\bis\source\dev\futura\lib\d3d11\txtd3d11.cpp(3239)(func: DX11::TextBankD3D11::ReserveMemory) [AssertNoDlg] ThreadID=2196 item.texture->_levelLoaded<item.texture->_nMipmaps
could this be related?
not getting a crash so far
quite some performance impact though
Could be related of course, something wrong with a texture, same as your previous error
we use this:
class OutsideTerrain
{
satellite = "...\s_satout_co.paa";
angleAltitudes[] = {{9,1}...,{9,1}};
colorOutside[] = {1,1,1,1};
analysisLength = 10;
enableTerrainSynth = 1;
class Layers
{
class Layer0
{
nopx = "..._nopx.paa";
texture = "..._co.paa";
};
};
forceMapOutsideColor = 1;
};
It could be any texture really ยฏ_(ใ)_/ยฏ
getting for all our terrains using infinite terrain with new angleAltitudes
could it be from the rendering issues of infinite terrain?
wuht just a edge where it stops?
mabe invalid outside textures or invalid altitudes
I don't really understand what these screenshots are supposed to tell me
the outside area has rendering issues when you use the new angleAltitudes options for specific view angles
SOGPF had no issues with that
They use it on CamLaoNam
fairly sure it also happens there - our terrain artists talks to theirs. will get it confirmed
Will update with photos but can 100% confirm it is on camlaonam as well
thanks, removed
we have encountered vehicle engine sound to stick if you remove a vehicle - not quite sure what exactly causes it. could be related to HC use. anyone familiar with this issues to help narrow it down?
Isn't this the same bug where if you pack a UAV before its engine stops completely, the sound will linger indefinitely?
Thats probably a very good repro, ticket plox
for the UAV?
- place a UAV operator in Eden & press play
- place the UAV from the backpack
- start its engine
- disassemble it, the backpack will be on your back but the sound will remain where the UAV was
I know.
ticket plox
no ๐คฃ
I found another "bug"?, with AI and directions, which they're told to observe
the AI being physically in a village on the east of Altis (Kalochori), doesn't look at the top (north on compass), how the compass shows north direction (blue line), but looks up and to the left (red line)
how do you determine where it looks?
it's shown on the map, where he's looking at
you select AI, press 3 and 8 and select to observe north
- chemist, can I have a pill?
- which one you want?
- a white one
I'm near it and I see he looks where I think he's looking at
and the map confirms it
don't AI generally take at least a 45ยฐ angle? as in "watch North not being 0ยฐ only"?
so these lines are engine generated?
daheck ๐
no, these lines I drew myself, from where the AI is look at
is the red line the body direction (vectorDir) or eye direction?
so here goes my question again how do you determine which way ai is looking
I don't know what it is. it's basically where he's whole body and head ar pointing atbody
so not where it actually is looking?
please at least use eyeDirection + drawLine
no there are other commands all documented
man, I repeat AGAIN, just for you alone
he's looking where he's looking and on the map it's perfectly reflected, confirming the red line a drew on the map, instead of looking in the blue line direction
the way I see it the eye is looking a bit to the right, no? so it falls on the blue line
it might be looking not where you think he is looking the only way to find out is to query it with relevant getter
no
it's beyond what I can do, sorry
so the AI doesn't simply look up, to the north, but looks up and to the left, because he's in a village that is to the right from the center of the map, so not only north, but also the top center of the map
AI looks up, only when I place him in the center of the map
try this:
_map = findDisplay 12 displayCtrl 51;
_map ctrlAddEventHandler ["Draw", {
params ["_map"];
_map drawLine [getPosASL AI, getPosASL AI vectorAdd (getCameraViewDirection AI vectorMultiply 1000), [1,0,0,1]];
}];
(if you have access to debug console that is)
but if I place him in a village right or left from center of the map, he will look up, but also to the right or to the left, facing the upper center of the map, not just up, where the north is
bob = group player createUnit [typeOf player, [0,0,0], [], 0, "NONE"];
bob setVehiclePosition [player modelToWorld [0,3,0], [], 0, "NONE"];
onEachFrame
{
_beg = ASLToAGL eyePos bob;
_endE = (_beg vectorAdd (eyeDirection bob vectorMultiply 100));
drawLine3D [ _beg, _endE, [0,1,0,1]];
_endW = (_beg vectorAdd (bob weaponDirection currentWeapon bob vectorMultiply 100));
drawLine3D [_beg, _endW, [1,0,0,1]];
_endV = (_beg vectorAdd (getCameraViewDirection bob vectorMultiply 100));
drawLine3D [_beg, _endV, [0,0,1,1]];
};
```๐
weaponDirection currentWeapon bob
can just dobob weaponDirection ""
I cp'd from https://community.bistudio.com/wiki/getCameraViewDirection ๐
so Idon't understand, why the AI does look straight up, only when placed in the center of the map, but doesn't when placed left/right on the map and instead looks up, but also to the left/right
no need for scripts to tell the difference
when placed in the center of the map
can you try a high place?
like on a mountain?
maybe it's environment related, they won't look straight at a house wall I guess
well based on what you say (looking up) it sounds like the commanding menu is mistakenly using the ASL position instead of AGL
so altitude can matter
maybe he will look down as well, to the sea level, since he's above it, instead of looking horizontally
haven't tried
ah I see what's going on
they're looking at the center of the map
directly north (i.e [worldSize/2, worldSize, 0])
that's a bug indeed ๐
then don't say
doesn't change
anything
please, it skews the data
can you make a ticket about it, please? in a more professional way to repro it, since I'm not friend with scripts
but it also concerns all the directions you tell them tool at, not only north
like I tell him too look at N-E, but he looks at east
when on the left of the map, at AAC Airfield
yeah
so I have to tell him to observe North, so I have better chances him observing N-E, even if it won't be exactly N-E
ok so I did a test myself.
looks like the AI does look north, but they only turn their body enough so that north is within their "head range", after that they just stop turning their body and turn their head instead
so they only occasionally look north (or whatever direction you give them)
@fickle root
hope it will be fixed
I'm not sure if it's a bug. the AI just place the direction within their field of view, so you probably don't have much to worry about. They still do see that direction
is it an engine limitation, that when you get in a heli as co-pilot, you can take control, but you have no access to flares and if an AA missile is fired at your heli, there will be no sound warning and probably nothing on the passive radar
or it's still a passenger seat, like in A2, just renamed to co-pilot and only taking control of the chopper was made possible, but without access to flares and warnings of missiles incoming?
also if the pilot of the heli doesn't unlock controls of the heli, when he dies, the co-pilot can't take control of the chopper
this would be good to have indeed
regarding the seat, I believe it is a "simple" passenger one yes
so if main heli pilot unlocks control of the chopper, without the co-pilot interfering when main pilot flying the heli, if the main pilot dies, the co-pilot can continue to fly by taking control
copilot is a turret
if the controls weren't unlocked by the main pilot and he was killed, bye bye heli, even if there is a co-pilot, since he cant take control of the heli
copilot can have its own weapons. What you would want to have is reserved manual fire option, where gunner controls pilot weapons
but co-pilot has no access to flares and missiles warning, when he takes control of the chopper
only control is transferred, not flares/warnings
this still needs to be fixed - https://feedback.bistudio.com/T148556
List of vehicles with wheels sinking into the ground (locally)
I'll set it as Critical
this also needs to be fixed - https://feedback.bistudio.com/T148411
List of vehicles with commander turret not returned back by AI to vehicle's direction, when player switches to another seat
manual fire only take control of first turret
there is no inverted way manual control option in game (turret->pilot)
beside that, there is bigger issue with CM disappearing from commander seat if another player gets into driver seat. Afair KK investigated it and there is no easy fix to it
this also needs to be fixed - https://feedback.bistudio.com/T137283
ability to see through deployed smoke when looking through Strider's windows, being outside of it.
https://www.youtube.com/watch?v=wU6d0yi7Aaw
Hunter and Ifrit don't have this problem with their windows
reyhard, Structures (Livonia) -> Village - Wooden House (Small, v3) Land_House_1W03_F is still not fixed
despite changelog of 2.00 saying it was all fixed
https://cdn.discordapp.com/attachments/108187245529268224/710908155986706442/arma3diag_x64_2020-05-15_19-20-49-894.png
https://cdn.discordapp.com/attachments/108187245529268224/710908105294086174/arma3diag_x64_2020-05-15_19-20-58-701.png
in A2, thanks to the community patch, we made it possible to switch from "co-pilot" seat to pilot seat in the heli, if the pilot dies - problem solved
and also in A2 with community patch, it was made possible for any passenger to switch to pilot seat from the rear of helis, if there is a door or space that allows a passenger in the rear of the heli to access the cabin of the helicopter
this way you can take control of the heli as main pilot and have flares and missiles warnings and there is no problem not being able to take control of heli as co-pilot, when main pilot dies and hasn't unlocked controls before dying
https://feedback.bistudio.com/T154433
Tank AI commander with 0 ammo, makes player or AI as gunner reloading 7.62 mm coaxial machinegun after each shot, if/after enemy AI gets spotted
https://feedback.bistudio.com/T154445
BTR-K Kamysh has no possibility for player as commander to cycle through different armament types (Ctrl+F), when turned out, unlike AFV-4 Gorgon with same turret/armament
https://feedback.bistudio.com/T154446
Left Mouse Button and Ctrl+LMB don't work to manually fire or order the AI gunner to fire from 120/125 mm cannon, when player is turned out as commander (using Arma 3 Apex keyboard preset) in T-100 Varsuk and Rhino MGS (UP)
that's a really annoying bug. It makes smokes useless on most of the tanks.
IIRC there was some scripted solution you had in RHS but removed it for some reason?
Yes, there was scripted solution in RHS which had to be removed due to author of script leaving our team. Basically it recreated vanilla UI and handled everything by script
in any case, @solid marten should know more about this issue
still not finished checking all the buildings windows and openings in different objects through which one can't throw a grenade
- buildings/objects through which bots can see/fire, despite visual obstacle, but in bots point of view there is nothing that blocks the view
don't kill yourself with a grenade you try to throw )))
https://www.youtube.com/watch?v=swrltjNwGNc
PvP server/mission - here I come ๐คฃ
https://www.youtube.com/watch?v=sMvCHycvDvc
https://www.youtube.com/watch?v=VDFMJZgMhjc
https://www.youtube.com/watch?v=cl83JZ4XpzE
https://www.youtube.com/watch?v=d6QB6kHBA4k
https://www.youtube.com/watch?v=Pvn9TWwKHes
etc...
dozens of buildings/objects concerned...
inevitably die from your own grenade or if still alive - reveal yourself to the enemy you wanted to offer a nade as present )
manually testing to see whether grenades can come through, where they normally should, each object/building there is in official Arma and each opening/windows those objects/building have...
that's stressful...
there are also objects/buildings, where grenades can go through, although they shouldn't be able to, because there is no openings/windows )
and bots that can see/fire through stuff they shouldn't and don't see/fire through stuff they should
in some buildings, grenades can pass through glass, without the need top break it first
in other buildings, grenades can pass through window(s), only if glass is already destroyed by rifle fire
but there are also buildings, that even when you destroy glass, grenades still can't go through the window(s)
or buildings, where grenades can't go through glass and glass can't be destroyed by rifle fire
or buildings, where glass is visually destroyed by default, but grenades still can't be thrown through...
so many inconsistencies/surprises...
maybe it's because so many different people have worked on different buildings, at different periods of A3 life cycle...
https://www.youtube.com/watch?v=KyF6GzLZyts
I should never have written that script
๐
I've abandoned so many times, but now I'm determined to finish this to make a ticket with all of this crap
super time consuming and tiring
poor guy, whoever will have to fix all of this, if at all, reyhard? ๐คฃ
my list is super long and contains a video for each object/building ๐คฃ
500 Mb/s internet makes it easier to upload (vs. 20 MB/s 1 month ago)
at least you know it's used extensively, for a good cause - it will benefit the whole Arma community
Has actually one of these windows ever been fixed since you started reporting them ? ๐
no
but haven't made a ticket until now, since I want to check every object/building opening/window there is, in official A3 and make one big ticket with all of it, with videos for each object/building
just poking fun at the amount of videos, I know it is for the greater good ๐
is there a limit in number of letters/symbols in feedback tracker?
not that I know of, but perhaps, why?
I fear that all the words/symbols there will be, might not fit into the ticket
it's a super big ticket
cp it in a notepad/word, and if it doesn't fit in the FT, link a Google doc
or maybe split the list in three sections, description, steps to repro + additional info. maybe the list will fit this way in one ticket, without having to use external services
if it is a list of all assets, attach a txt to the ticket?
yes, the displayed name of each asset + class name of each + link to the vid for each of them
txt/notepad file
good luck with grenades ))
house (large, destroyed) Land_d_House_Big_01_V1_F
https://www.youtube.com/watch?v=8L28bhANp1Q
not much better
house (large, abandoned) Land_u_House_Big_01_V1_F
https://www.youtube.com/watch?v=KqZm7obY0Ys
literally not one single opening works for grenades, but at least they're visually present ๐คฃ
same goes for USS Liberty )))
https://www.youtube.com/watch?v=27cGE5H3qUM
Perhaps you should have written it earlier when there was still a full team of bug fixers on the team! ๐
that's just few examples of many and I haven't checked Malden, Tanoa and Livonia, yet
can't imagine maps ported from A1 and A2...
Hey guys, i'm not being able to output reverse [1,2,3]. reverse "abc" works though. Stable 2.04.
you mean having a variable instead, right? Didn't work either. I tried reverse arr1, which was defined arr1 = [1,2,3];
I'm in Dev right now (2.07), let me test
Thanks.
Wait, are you expecting [3,2,1] for sqf arr = [1,2,3]; reverse arr?
yes
That isn't a case, never been AFAIK?
thats the biki example
I mean there're 1, 2 and 3. Which?
_arr = [1,2,3];
reverse _arr;
hint str _arr; // [3,2,1]```
Works pretty fine for me
Returns nothing, as expected
so I need to write an sqf?
What?
where should I test it?
Alright, i'm sorry it works with that local variable, but why can't I reverse a global variable like arr1?
I can?
๐ I can too, but running a test.sqf, not from Debug Console.
reverse arr shouldn't work?
test.sqfsqf arr = [1,2,3]; reverse arr; systemChat str arr;And it prints [3,2,1]
i'm using the watch field on debug console.
What?
What are you expecting from this?
it should output below



