#arma3_feedback_tracker
1 messages Β· Page 42 of 1
https://feedback.bistudio.com/T165845
Move to DayZ Feedback. Or remove altogether cuz it's more bullshit than actual feedback π
ah that's sad, I can fix it
player setMimic "smile";

To import a preset, drag this file onto the Launcher window. Or click the MODS tab, then PRESET in the top right, then IMPORT at the bottom, and finally select this file.
1 sec discord be bugging out
2 small changes I'd propose
To import this preset, drag this file onto the Launcher window, or click the MODS tab, then PRESET in the top right, then IMPORT at the bottom, and finally select this file.
a -> this because its talking about this specific modpack, and . Or -> , or since afaik sentences shouldnt start with a connection word thingy like or
(at least thats what its like in german but I think the same applies in english)
doesnt make much sense grammatically methinks
as a sidenote this change could also be made
To import this preset into the Arma Launcher [...]
but thats propably not necessary
TIL that apparently sentences can be started with conjunctions in english
not ideal but acceptable
but not ideal π
Is there a reason for this error when using setOwner? It appears when running on a vehicle that's occupied
Script command setOwner cannot be used for object 'LIB_Pak40'. Use setGroupOwner instead.
setGroupOwner does not work for this purpose
Despite the error setOwner works correctly
vehicle thats occupied is always owned by its driver (or the driver of towing/slingload parent)
Yeah - but this is a static, which doesn't have a driver. so it's usually owned by the server
basically I was just trying to transfer ownership to the gunner cuz of the lack of driver
which Im pretty sure is the reason this is bugged too: https://feedback.bistudio.com/T153212
Do turrets also have hasDriver=-1 entry?
or 0
I could update the engine locality decision thing to take command or gunner passenger if no driver exists
hmm not sure? i can see 'hasGunner' in turrets but no hasDriver
but that's just looking at existing configs and their defaults
Mh... looking at code now.
It takes copilot as owner first, and if no copilot present then it takes driver
but wtf, driver should win, not copilot
that's strange
maybe it's an override that's disabled until the copilot uses the action to take control
I can easily change it that if no copilot AND no driver, it takes effective commander instead.
Which I think would be gunner
depending on the vehicle set up yeah
We can test that on profiling branch. need to update it today anyway
maybe copilot "is a copilot" only when control is taken?
oh it was already written, discord did not load the history. 
I see you guys like the taste of canned worms
here is the setLightDayLight ticket we were speaking about earlier at #arma3_scripting
Answered it
spam account & ticket https://feedback.bistudio.com/T166043
can this ticket get closed?
https://feedback.bistudio.com/T164620
I replied to it https://feedback.bistudio.com/T165990, excuse me for the long reply
heads up still not fixed
Iβm sure there is a bug somewhere how the light is calculated in relation to global light, but that is another matter and unlikely to change as this could affect a lot of missions. The command in question has very simple implementation this is why I couldnβt find anything wrong with it.
only when 2.10 is released no?
It was implemented in dev branch and we found a bunch of problems that got solved
So I guess its up to you to see when its fit to close it
I understand, just wanted to point out why I thought it was turned off, many thanks
And I guess the only workaround is a loop that checks what is the time and adjust the light intensity accordingly but that's too much honestly I would just do it at night time
yes, its already in feedback
since this can't be fixed https://feedback.bistudio.com/T165990
can we get this ticket closed?
I don't see why that can't be fixed. If underwater lighting brightness calculation is broken, and thats all it is, that is fixable?
Not sure whats involved in it, would need to look at it first to see if its fixable or not, but as it seems noone did that yet I wouldn't call it unfixable
@sinful kettle
https://feedback.bistudio.com/T151907
What is difference between cell size and heightmap resolution?
Or rather, what is "cell size" for you?
diag_getTerrainGrid is cell size. I got the name from:
https://pmc.editing.wiki/doku.php?id=arma3:terrain:grid-cell-size
heightmap resolution is worldSize / diag_getTerrainGrid
it's a bit redundant but I figured it would be nice to have it in there
No getTerrainGrid returns heightmap resolution
by heightmap resolution I mean 2048*2048
Ah. Well its the size of one pixel
resolution of the heightmap texture (only the 2048 one)
You mean worldSize*terrainGrid then.
But thats also wrong, worldSize returns landGrid size, which isn't necessarily the same as terrainGrid
really? 
wait I'm a bit shaky on the terrain terminology. what is landGrid?
its what worldSize uses, and dynamic simulation and AI and other stuff. Not sure why its seperate
but in the end landGrid*landRange and terrainGrid*terrainRange
and the other one or two grids that there are, always end up at worldSize
idk but when I use this equation I get the hightmap resolution (e.g. 2048 for Stratis, 4096 for Altis)
tested on many terrains (vanilla and non-vanilla)
[32,256,4,2048,0]
Here on stratis.
Land grid cellsize and width.
terrain grid cellsize and width
and sealevel
isn't landgrid used for the quadtree? 
but anyway, the second pair of values is what I meant in the ticket, 4,2048
btw seaLevel is real time right?
No one said it canβt be fixed, if this is not going to suddenly whack up the brightness of light sources in existing missions, by all means do it!
yeah, if it changes that would change
He said it can't be fixed
Is this answered? I think I've found a such issue
reyhard does it from time to time yes. but he doesn't have much time for it
Kidzone kid said it has to do with global lighting and might break some missions? Idk I have never programmed games for real lol
Of course if its fixed it would be real good, I can finally create my SOMA type of Arma 3 missions
2Γ link to the messaaages plz
all nuked
Could be related to how underwater lighting keyframe are written?
cleared it all, thanks π
Spam comment: https://feedback.bistudio.com/T166041#2331635
no rest for the wicked
Experienced AI Helicopter tries to climb infinitely when there is some attached objects:
https://feedback.bistudio.com/T166266
found a crash if #(rgb,8,8,1)color(...) is improperly used
waituntil { !isnull (findDisplay 12 displayCtrl 51) };
findDisplay 12 displayCtrl 51 ctrlAddEventHandler ["Draw",
{
params ["_map"];
_map drawRectangle [getposatl player, 250, 10, 0, [1,1,1,1], "#(rgb,8,8,1)color([1,0,0,1])" ]; // Crash because of array format used
}];```
or something like this will also crash: "#(rgb,8,8,1)color(asdf)"
sometimes there are error messages and no crash but I don't know why only sometimes
Ah if the rectangle is being rendered then arma crashes
if it's drawn/located outside of view then arma wont crash
yeah it seems drones don't suffer from this
They do
can i post pic link here?
ok
Idk if this is known issue but this happened to me today: https://postimg.cc/mzc94Dh2
what's the issue?
only the guy's head is rendered
do you use setObjectTexture or something?
never seen that happen 
guy setObjectTexture [0, ""]?
Possible to see such when your camera clips into something, but it doesn't right?
yeah no, seems camera-related
try _unit setUnitLoadout configOf _unit
not using that
yep (and I should read the chat better)
i already closed the session but can return to this if needed
well if it was a texture problem that'll reset it
it must be clipping issue somehow
maybe 
like I said happens only in certain angle
oh didn't see that 
what object was that house?
it could be an occlusion problem or something
lemme see
should be i_house_small_01_b_pink_f.p3d
wow i just realized it happens to whole map, everything stops rendering. only ground left
shift + - then type F L U S H
that's definitely an occlusion issue
this I mean
get out of the house, see the stuff being drawn!
it only happens when I stand at certain position in the house
which map and what pos?
and at certain camera angle
Malded 2035. [7312.34,7919.9,0.740433]
yeah happens to me too 
wow
make a ticket
well good it isn't just my PC π
There you go: https://feedback.bistudio.com/T166297
Spam comment: https://feedback.bistudio.com/T166244#2333224
Huh so this happened while tinkering with the controls, cant replicate it though
probably same as scrolls bug 
stable build btw, forgot to mention it
That happens occasionally with dropdowns, can happen in Editor attributes too. I thought I saw a report for it a while back π€
yeah I also experienced that
specifically pylons loadout
usually fixes itself when you click it again
Yeah it did
incorrect geometric occluder
mh no repro on internal data.
Maybe already fixed
Can someone check on dev?
was that on dev?
no that house model wasn't edited. Why do I not get it
There is no occluder on that wall, da heck
yes
only happens in 3rd person
as I said in the ticket
go back all the way back against the wall
then try in 3rd person
omg. Did you just turn this from a encoding into a programming issue π
No lucky.
The occluder is juuuuuuuuuuuuuuust infront of the wall. instead of inside it
the weird thing is I couldn't repro when I spawned the object myself on another terrain 
quite a bit infront of the wall infact
yeah just tried again and indeed it doesn't happen
Correct object?
a3\structures_f_argo\civilian\house_small01\i_house_small_01_b_pink_f.p3d
"a3\structures_f_argo\civilian\house_small01\i_house_small_01_b_pink_f.p3d"
yeah
I put the object on Stratis airfield and opened the door and kept circling the camera when I was at the end of the corridor but no repro
Well the issue is visible in the model that'll be sufficient
I think it's one of those "object placement" related issues
The position of the house looks bit undone because one of the windows is facing a wall. dunno if that can be part of the problem
like I showed you before
and there was also a bug with animations
the game definitely does something wrong
I guess the way it puts objects in the quadtree or looks them up 
if you feel up to it it's worth investigating imo
the whole house is hastily placed https://postimg.cc/BPHs5qgp
P.S: that bug is not specific to animated selections
I've seen it happen several times with completely static objects as well
it seems the object's position is an issue
that same door object in that ticket works fine when I spawn it in 3den
wait it works on one side but not the other 
check this out:
(both doors are open)
if you stand on the side the gate opens to it no longer works
so definitely something wrong with the "quadtree lookup"
here's the kicker: even bullets pass thru! (and they don't if you stand on the other side)
it's not just lineIntersectsSurfaces!
There are tons of reports of that.
Not sure if fixable though.
#Ignored
boooo #devsDontSupportTheirOwnGame #ArmA3isDed #BIstealsMoney
You didn't make a ticket right? if no then I don't need one.
Interesting in eden editor you get a popup that the formatting is bad.
I guess its specific to drawRectangle then
Found this comment right above the crash
haven't made ticket. you get popup if it doesn't crash (in game). the drawRectangle seem to be what crashes
because if the rect isn't in the view it won't crash
usually commands verify that what you gave them is valid.
But the map draw ones don't
ok
Well fixed now
great, ty π
Spam comment: https://feedback.bistudio.com/T166266#2334041
you know it would be useful to have some command that gives debug info of how long some scheduled scripts are running. Then you could more easily optimize your scripts. there's diag_activeSQFScripts but it doesn't give any details about performance.
what you guys think? π
the duration of scheduled scripts is kind of meaningless
because it varies depending on scheduler loads
I can write a script that does hint "bla" that takes 1 minute to execute
doesn't mean it's slow
@sinful kettle But you could get info that running your script took 1ms to run and use that info to spot heavy scripts
or your script took the max of 3ms to run, meaning it took maximum time available
like I said taking n ms to run in scheduled environment doesn't say anything about the script's performance
well if one script takes 1ms to run and other 3ms I think that's useful info
well it should only measure the time it take to process the code. when sleep comes that's the end of that run
well it should only measure the time it take to process the code.
the time it takes to "process the code" is the same as the time it takes in unscheduled environment
roughly
really? I think different length of scripts/code take different time to process...
right
Having that measurement on every code ran and saving that info to somewhere would be good
it maybe bit problematic with sleeps though, as it could measure only the time between sleeps if there is any sleep used
There is a tool for that and it works on unscheduled code, some really cool dude made it π
https://github.com/dedmen/ArmaScriptProfiler but it'll only really work with unscheduled code but generally it is enough to infer where issues lie
never learned how to use that one π and I would really like it for scheduled
@solid marten Hi.
I'm pinging you about that EH ticket. You said if Dedmen doesn't claim it I should ping you.
https://feedback.bistudio.com/T166230
ok thanks
@untold sky found a bug with the selection name returned with lineIntersectsSurfaces, while the intersect command does it correctly:
https://feedback.bistudio.com/T166368
you return the parent bone
child bones are never returned
may my new todo list be reliable
with https://community.bistudio.com/wiki?title=Arma_3:_Mission_Event_Handlers#EntityCreated wouldnt it also make sense to have EntityDeleted ?
you?
Just use EntityCreated and add a Deleted handler
Hello,iam new to arma would someone like to show me how to play, i mean i played SQUAD, but i don't know how to play arma and whats the common thing to do.
Main Menu -> Tutorials -> Field Manual @ebon bone Also wrong channel
What I did was I just fucked around with stuff in the editor and eventually figured it out
There's lot of stuff to explore
lots of things to do
Ok tnx guys i hope that would help
except deleted EH doesn't work with simple objects 
still not gud enough! du eeet!
stop ping
π π
and entitycreated does?
Issue with setRain (RainParticles version) and rainParams when used on dedicated server: https://feedback.bistudio.com/T166399
yes
it triggers for all objects
it even used to trigger for footsteps and insects 
wait it still triggers for footsteps 
I thought Dedmen said he removed it 
Wasn't the suggestion to add a second filtered handler?
Although I can't imagine why anyone would want a realtime insect feed :P
it was denied
yeah that was the idea
but Dedmen said one of RHS UAVs or something was an "insect" (CfgNonAIVehicles)
it's also used for proxies too iirc 
Arma is a mess :/
The rainParams is available on dedicated, dont know why you get empty array, please provide video/repro/dunno. I explained in the ticket it doesnt automatically sync as it is not meant to change. I wont change that. Instead I will think of scripted solution and post on wiki
I think the only time it would work out of the box if you mod the rain config, i.e snow map, then it should sync fine. If you use script to change that then you need script to sync
So global execution is required to sync across to all players? If that's the solution I can work with that. I'll create a video of repro for the empty array tomorrow if still wanted.
This would be great time to reintroduce voting on feedback tracker. 2016 when they reworked feedback tracker we lost that feature that showed what players prioritized.
I hope for its return one day
Or maybe the devs can make the pool right here....
There's a difference between bugs and feature requests...
Bugs don't need voting, it's broken so it needs to get fixed.
Features are nice, but only AFTER the bugs are fixed and IF it fits the scope of the game.
If BI runs out if ideas I'm sure they'll ask for it, but that is not their focus right now. So no need for voting.
And since most people have no idea how to report an issue or request a feature properly, it's a good thing they removed that feature from the tracker.
Any community voting system would quickly lead to ridiculous requests being heavily upvoted (e.g. "add more maps", "remove CSAT rubber suits", "make the AI play like humans"), leaving actually achievable or important issues (like every ticket I make :D) at the bottom of the pile - thereby making the system useless.
In theory it sounds nice to be able to resurface pet issues to get them more attention, but there's no way it ends up actually working out like that.
raises a zombie army to vote up "make the AI drive better"
that and any new features coming to a3 are likely going to be ones a good chunk of the community will not immediately notice also meaning they would get considered "boring" suggestions
If you were around 2016 you would know that Add more maps or remove CSAT rubber suits were not among voted stuff. What was for example was "Ability to walk on moving vehicles" "Fix for mid range textures" actual important engine changes that no one is talking about right now. All because we
have no way of voting on them.
While I obviously don't have access to the code base I'm under the impression there's a engine-based variable that keeps track of when the player is trying to aim, is it possible this could be added as a scripting command similar to cameraView etc?
reason its noticeable and why I ask is if you right click while sprinting around the moment you stop sprinting the game will aim down sights for you, basically buffering your previous input
unfortunately simply using a mouse button down handler and switching between true/false on a variable is unreliable for reason I cannot work out (usually involves double clicking like mad) and the scripted variable vs. the engine ADS variable get out of sync
Only where it shows that rainParams is []
Making the video now
I've added it to the ticket.
Could you do this on dev and not on 2.08 and no need to diag_log, just type rainParams in debug console and press server exec
roger give me 10 mins
Added.
added note to the wiki, this is not changing
@paper lichen Im adding BIS_fnc_setRain that will handle JIP, it doesnt need rainParams
Cool, I'm guessing it would still require server execution though? (the BIS function)
you can just put it in init to be executed by everyone, but if you want to change something later, server exec yeah
There is a eventhandler for ADS? Maybe new in 2.10?
yeah... unfortunately it has the same issue, doesn't trigger if you click while sprinting or any other animation that blocks aiming
I managed to get something reliable using keyhandlers
Maybe thats something we can still fix @neon swan 
But knowing what a mess it already was, probably not π
Oh god, the headache comes back trying to think about how the aiming logic works.
im surprised its such a mess, it seems pretty reliable with banking/buffering the input until the unit is able to aim
It is all part of the cameraview loop
I thought I fixed this, btw? Will have to check again.
hmmm if you did I didn't see it in a changelog/ the ticket. i honestly haven't tried it since
lemme boot up the game and see
Second to last point of this changelog
https://forums.bohemia.net/forums/topic/140837-development-branch-changelog/?do=findComment&comment=3460555
hmm ok I missed it then my bad, trying it now
hmmm thats new the enum isn't working
OpticsSwitched right?
Should be, yea
oh no it's Switch
the wiki is incorrect
ooh yeah it seems to work
works very good infact, it triggers soon as I stop sprinting
I just BIKI corrected
I'll try a few other edge cases I noticed when I was using my key handler but it seems to be fine so far
nope works when you switch weapons too, goes false and then true again when the weapon switch animation ends and the gun is raised again
so all good, ty very much sorry for not noticing it got fixed till now π¬
should this have been removed Ded?
probably yeah+
about lineIntersectsSurfaces.
if the intersected thing is in multiple selections, its probably better to just return array of selection names instead of choosing either first or last
a surface can be in multiple selections?
they're solid tho
as in they don't deform
The one you sent has two
if it works like the intersect command I'm good with it
There was a difference to intersect. intersect command tries to find the name of the skeleton bone.
lineIntersectsSurfaces finds the name of the selection that contains the part that you hit.
And it could be in multiple selections
I changed it now to try skeleton bone name first, and if not fall back to getting selection names
yes I did that, if parent class is CfgNonAIVehicles, its ignored
getting selection names
does it mean it can be array now?
or string like before?
well I just needed the bone (only 1) 
well seems to work fine
thanks π
indeed footsteps didn't have a class.
fixed now
maybe both? 
would be more futureproof to change it now.
If there is a case where multiple may happen, then it will just work.
Otherwise it can't be fixed anymore as we can't change syntax after release
as in the returned intersection array should be:
[pos, normal, proxy, obj, bone, surf, [selection1, selection2, ...]]
the reason I said that is that if you return the selection name again it can become wrong for me
it has to be an animated bone
so either only return bone, or both bone and selection
I personally say the former, because it's easier, and it's all I need
but if you want to return the selections too then the latter

I mean either:
[pos, normal, proxy, obj, bone, surf]
or
[pos, normal, proxy, obj, bone, surf, [selection1, selection2, ...]]
not
[pos, normal, proxy, obj, boneOrSelection, surf]
that boneOrSelection (uncertainty) causes problems
I planned
[pos, normal, proxy, obj, [selection1, selection2, ...], surf]
if there is bone its always first
btw can a vertex be animated by multiple bones?
they can in soldiers
but I've never seen it with other objects
The code is written for multiple

actually no it doesn't 
e.g for that apc, both the otocvez and poklop_gunner are animated bones 
it returns only one tho
even on a human unit I can never get more than 1
on the small military container, if you aim at the glass it doesn't find any skeleton bones.
But then it falls back to finding named selections and finds glass_1_hide
Does it make sense to pay the overhead for array if in 99% of cases there is just one 
annoying
but the glass_1_hide is animated
its animation type is hide or something like that
if it's always 1 then why return array? π€
when can it be more? multiple selections but none a bone?
But its not found as bones
For the chance where its more than 1
I noticed that I can't run/sprint but only walk, if I overload than regular loadCoef allows, even if player setUnitTrait ["loadCoef",-10e30]; or such, even the stamina bar suggests I have enough stamina to run. Bug or intended?
It should probably respect loadCoef
So ticket worthy?
It was reported that daylight flares on lights are broken.
But they are not really. I just tested with PAPI lights, they don't show at daytime, because their intensity is so low that they wouldn't be visible in ambient light.
Their base intensity is 60, but they need about 50000 to be barely visible during daytime
Sadly just upping the intensity, breaks them at night :/
For papi lights, just force increasing the intensity during daytime works fine.
I guess I will do that for all.
it will cause the "side effect" that lights that were configured to use daylight flares, will now correctly use daylight flares... I hope noone will get annoyed by that...
this is the arma community there'll be someone who will
Hey people, send me tickets 
can the game hide specific LODs? if so this feature would be very nice:
https://feedback.bistudio.com/T166409
many people have been asking for disableCollisionWith to work with multiple objs
but if you can just "hide" a LOD there won't be a need for that
just send list of stuff I'll pick stuff out tomorrow
well as per usual at the top of my list are these (they're for my AI mod)
- vehicle handling commands: (maybe you could create a similar "controller" as you did for cruise control stuff?)
https://feedback.bistudio.com/T150072 - AI aim/head direction commands. I know KK said before that they can be used as aimbots, but we have
lockCameraTowhich does the same thing for vehicles and somehow that doesn't count as "aimbot"
https://feedback.bistudio.com/T150070 - ability to jump to specific "frame" of unit animations:
https://feedback.bistudio.com/T164731
also one by POLPOX:
https://feedback.bistudio.com/T157753 - ability to extract info from
model.cfg, namely, bone hierarchy and animation parameters
https://feedback.bistudio.com/T150103
1 & 2 essentially complete the missing parts of my mod (and I know, they're the hardest ones
)
say3D's first person in vehicle bug > all
If the player is in first person view inside a vehicle at the moment of command execution, the sound created is greatly attenuated.
make setPlayable great again?
I know it's impossible/would imply great changes so it's a no, but one may still dream
- force rotor shear command
https://feedback.bistudio.com/T163193 - system/global chat execution is slow on HC in comparison to other environments (2ms+)
https://feedback.bistudio.com/T155902 - Vision Modes Not Working With switchCamera
https://feedback.bistudio.com/T122792
I guess this is also related a bit although not flares, https://feedback.bistudio.com/T165990
ticket title is a bit misleading, video will show the issue I'm talking about more clearly
That's one pretty good light if it's still working at that depth π
Simple sqf only function fix: https://feedback.bistudio.com/T164264
Small string only fix: https://feedback.bistudio.com/T162306
*unstabilizes your MBT cannon*: https://feedback.bistudio.com/T162061
_light_emitter_0 setLightIntensity 17000;
_light_emitter_0 setLightAttenuation [15, 4, 5, 0];
true but it it should be more bright, far more bright
Ye its true either way lol
is that setLightDaylight or whatever that command is? can you retest that on tomorrows profiling branch? There will be a change to it
yes sir
_light_emitter_0 setLightDayLight true;
I will test it tomorrow too
if you still need tickets: https://feedback.bistudio.com/T161933
it doesn't? i was able to make one man vs 4 men collision disabled
Each object can contain only 1 reference to the object it disabled collision with. Thus trying to disable collisions with multiple objects will overwrite object previously set for disable collision
yeah i know what it says...
and I've tested it and it's true
hmmmm
maybe unit-unit is different 
I can walk thru 4 dudes π
no
Spam comment: https://feedback.bistudio.com/T166429#2335625
makes sense if those 4 units have disabled collision with you
so how does this contradict what leo quoted?
wasn't sure if it did
wait, let me show some code?
you should read biki page first
i did...
here's my code: ```ll = player nearEntities ["Man", 50]; { if(!isplayer _x) then { player disablecollisionWith _x }; } foreach ll;
why am i looking at it
wasn't sure which object had the collision disabled registered
player or the _x (AI)
and here is your answer https://community.bistudio.com/wiki/disableCollisionWith
the wiki doesn't seem too clear but maybe it's just me π
it seems like you can have multiple collisions disabled but only one such case registered
buut it's probably just me not reading it right or something , so nvm π
It said there in the first paragraph that collision is disabled for both objects, does it say something different for you
says the same to me
I still have this one open: https://feedback.bistudio.com/T163128
ping on reply being on by default is the worst decision Discord's made since the new logo font
I remember that one but can't find an open ticket. Do you mean very quiet? That was fixed in 2.08
whats the problem with it?
Besides that its only possible during loading game state or before. not when game is started already
the wish to have it work in later stages :)
when can one use it? server-side preInit function?
Don't know if preInit is soon enough, might be
well, otherwise it's useless?
- force rotor shear command
If you are aware that rotors shearing off only exists in advanced flight model, which some players may turn off... Then probably doable
What determines which model AI helos use?
Dedmen (ΰΈ β’Μ_β’Μ)ΰΈ β Yesterday at 6:01 PM
Hey people, send me tickets :owoyay:
is this gathering suggestions for 2.12?
Seems like so. One of my subscribed ticket got a message with Ded's considering to do it in .12
Whatsoever, this game's not going to dead within years thanks to the dedicated developers
I'm not sure I understand, when the engine detects a "collision" of the rotor (via the blade radius) it destroys it in the basic flight model, as far as I'm aware the only real way to replicate that is to spawn some object that'll trigger the collision check. IIRC there is a parameter that is configured on helos that defines the radius the collision check is allowed to trigger within and I had that visualized somehow: https://i.imgur.com/Rf1Mhqp.jpg
Quite simple to do but ehhhhhhhhh I don't want to :/
which one is simple?
disableCollisionWith or hiding LODs?
its currently just one value thats disabled, could turn it into array
generally stuff. If small enough I might do it for 2.10
but rotor destroy is not rotor detach. Its still attached, just broken
alright
Hm that definetly sounds nice
a few bugs with repo
[BUG] AI cargo only replaces dead driver once [if not in the same group]
https://feedback.bistudio.com/T162686
[BUG] DisableAI "COVER" makes AI do weird moves and triggers Combat Shouts
https://feedback.bistudio.com/T155908
[BUG] displayAddEventHandler [KeyDown?] remains available after mission restart - could be abused?
https://feedback.bistudio.com/T163625
[BUG] AI [tank gunner] fires (under player command only?) first shot too early and thus always misses first shot
https://feedback.bistudio.com/T152879
My bad then, wrong terminology. Was asking for a method to trigger the rotor destroy behavior.
Remind me tomorrow to revisit that, I'll try to remember but my brain is full of soup
I told you to stop after that 5th bowl
hah that chat stuff.
yeah don't run systemChat, in a while true loop, inside a eachFrame handler, on a HC.
It spams so much into the log window, that the log window gets so full, that Microsofts code that tries to process scrolling in that window, is so slow that it basically just freezes
Need to limit length of log window
is this profiling update you are talking about released now? (just to make sure my test is right)
my bad never knew about that channel
ok sir 
Thats a easier fix than expected π
That makes sense, before I disabled and replaced code with external logging it was executing every 1 sec with FPS/AI counts to the window.
is this for the underwater lighting issue or for something else? (I'm not the brightest at coding)
lighting, brightest⦠there's a joke in there, I'm sure of it!
It was for the logging stuff :U
yeah I tested it on the latest profiling branch, will keep any updates here so we don't bloat the chat:
https://feedback.bistudio.com/T166453
my bad :v
This chat is made for bloat
true but Leopard will delete me from the globe (they are a nice person and is trying to keep everyone safe)
windows server log window is so bad :/
There is no max limit in there, and the more text is in there, the slower adding text to it becomes.
It can get to the point where posting a chat message can freeze the server for half a second
btw Ded did you make any more changes to the lineIntersectsSurfaces command?
the thing you wanted
I'm getting some weird results and not sure why
the bone change, nothing else
Even after adding a limit that removes text if it gets too large, its still too slow.
@livid sleet did you only test the slow systemChat on a headless client, or did you also check on dedicated server?
normal server has the same log UI window, so it should be just as slow
Oh nevermind, actually dedicated server doesn't log chat messages.
I think that would make sense but apparently not
Will just have to make it async logging like I did with RPT, that should be fine. I didn't want to do it because I thought it would be more effort..
But now I spent more time on trying to find a solution without it, than it would've taken me to copy that stuff over π€£
IIRC I did test server vs HC. It's not a huge deal anymore since I've moved on to better things but I hoped issue wasn't indictictive of some bigger performance issue with using the command outside of that context.
Well yeah it is. Generally chat logging on HC is bad, so I'll fix that anyway
or rather, all chat logging is bad. But normal servers don't log as much stuff to chat so its not thaaat bad there
I'm now 100% sue that something else has changed. I even reverted to an older version of my code which I know for sure it worked before, but now it doesn't.
it looks like it's skipping some intersections now 
maybe you changed something for debugging but forgot to undo?
I'll DM you a video
display event handler is not deleting and it is not changing, some campaign missions rely on it, bad design? doesnβt matter anymore, it cannot be changed.
roger
couple of feature tickets if those are to be considered too
[Feature Request] Add some shading to diag_drawMode
https://feedback.bistudio.com/T163437
[Feature Request] Add Mission EH 'HandleRadioMessage'
https://feedback.bistudio.com/T161306
[Feature Request] New scripting command: getCurrentAutoCombatTimout
https://feedback.bistudio.com/T161510
[Feature Request] New scripting command: setAutoCombatTimout / getAutoCombatTimout
https://feedback.bistudio.com/T156436
[Feature Request] Add SQF commands to get all guard points, update its position or delete one
https://feedback.bistudio.com/T161625
[Feature Request] Improve guarded points
https://feedback.bistudio.com/T161138
"[Feature Request] New scripting command: setAutoCombatTimout" is particularly a useful one imho
βοΈ My one and only remaining ticket (If it gets implemented it would be a first for me π )
I believe Dedmen mentioned on the CONTACT channel that this was already fixed: https://feedback.bistudio.com/T160213 (Perhaps update to ticket to "Need Feedback"?)
And also this was previously on your TODO list too Dedmen: https://feedback.bistudio.com/T161132 (Don't know if it still is)
Flares still don't work during daytime it seems? https://feedback.bistudio.com/T152896
This is still open? https://feedback.bistudio.com/T164086
This would be awesome to have: https://feedback.bistudio.com/T152937
The last one seems like it would require a major overhaul of how vehicle ammo is handled and is probably out of scope (not to mention the backward compatibility problems)
KK!
https://feedback.bistudio.com/T118270 is control problem
And maybe this β https://feedback.bistudio.com/T160343 in same?
What if this is not a replacement, but an addition? Like:
After old reload, check if new canister is present, if yes reduce new canister count by 1 and add 1 canister to the old system.
sorry for the noob question but what does "set Ref Ticket to AIII-55256" mean in the feedback tracker?
does that mean they work on it? π
They use Roman numeric to call A3? π©
It means they've decided it's worth raising an internal ticket to track it, which is not a guarantee it will happen but they are aware and interested.
hmm ok, that's cool
thx for the info
I guess that doesn't make the ticket it self obsolete
The FT ticket contains the information on which the internal ticket is based, so no, it's not obsolete.
The connection between the FT ticket and the internal ticket allows them to find the FT ticket again when working on the internal ticket, so when (if) the internal ticket is resolved, someone will probably update the FT ticket to reflect that. If you pay attention to other FT tickets that are resolved, you'll see someone usually posts when the relevant change is committed with a revision number.
If you have important new information that's relevant to the issue, you should update the ticket with that so that they are aware of it.
ok thx, good to know
Maybe such autotune would be good for tracers as well?
We play TvT with seagull spectator without reviving and respawn, but sometimes death by bugs happen.(abrams delayed start blowup as an example) would be nice to have ability to spawn new units by admin.
In my experience, tracer visibility in daytime is already perfectly reasonable.
Tracers also aren't really light sources. They don't cast light on other objects or the ground and I'm pretty sure their glow is caused by materials, not a "real" light.
omg, sorry but no 
The server/HC slow console logging is fixed, but you won't be able to benchmark it as the buffer is only 32 elements in size. Good enough to remove any lag from chat messages, but not good enough to remove lag when something is really hammering/spamming the log.
I don't want to waste memory on something that normally doesn't happen
Maybe this is my settings or eyes but they are almost invisible at takistan day.
Does KK want tickets? π
https://github.com/acemod/ACE3/pull/8457#issuecomment-1062522447
When you on some object in building you hear outdoor SFX
https://feedback.bistudio.com/T153970
AI issues are generally meh.
Its always dangerous as it could break some campaign or mission, and the changes are usually very complex too, so I don't even like to attempt these.
any chance this can be looked into again
https://feedback.bistudio.com/T150837
π¬

arma 3 blueprint simulator
There now you can see stuff in diag_drawmode π
Need to change color of the lines
That's sorting problem, or not problem because this is how sorting by text does it
You wanted shading for diag_drawmode's, this should be sufficient
that's the roadway LOD isn't it?
try geom or fire
they're pure white
I did yes
so those wireframes won't be visible
first screenshot is that. they are visible but badly
well that's the point
try moving inside a building
I doubt you'll be able to see anything
looks fine to me
I can't tell what's what 
Well thats all you get 
Idk if you have checked the ticket or not https://feedback.bistudio.com/T166453
But the issue os still there after the profiling update, here is a video
https://www.youtube.com/watch?v=KEmGIy_dJGA
Info:
Launcher version: 1.5.149102
- Hardware acceleration disabled
Game version: 2.08.149650
Branch: profiling
Yes.
Also when actually implementing I use the ticket as guidance. So if its updated between setting internal ref, and actually starting to work on the internal ticket, it will be seen
lol wtf is this π€£ Nice find
Great thank you
does it happen with other buildings?
thanks for checking
just curious - do you refer to the disableAI "COVER" or the AI firing too early, or both?
all AI π
Perhaps simple Ai tweaks or improvements that can yield higher FPSs are still worth checking out. I don't know how resource intensive the Ai units become when in combat mode compared to when in safe mode but being able to reduce the time the Ai units spend on combat mode via a new command could yield some improvements in FPSs when there are lots of Ai units. (Just using that as an example)
I don't know how resource intensive the Ai units become when in combat mode compared to when in safe mode
very
so no chance for custom coloring like VBS has? ref: http://web.archive.org/web/20200606213438/https://sqf.bisimulations.com/display/SQF/setDrawMode
disableAI "COVER" seems used only in 3 places basically in A3: https://pastebin.com/K33L0WLz
- time trials
- law dlc
- TOH function/module (supposedly by the supports game mode)
Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.
nop
I think nearestObjects not supporting ammo parent classes is a bug:
https://feedback.bistudio.com/T166472
because when you don't provide any classes [] it does detect ammo objects
Yes
with all buildings
Spam comment: https://feedback.bistudio.com/T163703#2336466
The last one, I still don't know why it is not something can be happened
Isn't 161980 essentially covered by the Presets feature?
And....spam: https://feedback.bistudio.com/T166478
Not really IMO > 161980
https://feedback.bistudio.com/T78663
Plus a bit - can we have a love about this?
When we're playing Arma 3 in public multiplayer server, it is very hard to identify the unconscious player's location when they're using radio (not direct), and vice versa yet hard to notify, which is not an ideal way to communicate IMO. If it transmits the voice both 2D and 3D like major radio Mods, it is pretty
can everyone please post the ticket name if you link them [spam excluding]
will also increase the chances a good deal anyone to check the linked ticket
https://feedback.bistudio.com/T120423
Not the author, recommended I post here by a veteran named Pi123263
From 2016, reposting because of a relevant issue I encountered wherein if I saved a scenario and loaded it in lots of the mod-added items would be completely removed, the large majority of them.
TL;DR Loading up a save in the scenario causes lots of mod-added editor items to be removed from the unit panel meaning I can no longer access over half of my units
Found such error in RPT
21:02:27 Error in expression <if (local (_this select 0)) then {[(_this select 0), ", [], false] call bi>
21:02:27 Error position: <[(_this select 0), ", [], false] call bi>
21:02:27 Error Missing }
search through whole arma folder bring me to
class Land_RepairDepot_01_base_F: ReammoBox_F
{
...
class EventHandlers
{
init = "if (local (_this select 0)) then {[(_this select 0), "", [], false] call bis_fnc_initVehicle;};";
};
I think it should have """"
its Arma 3\Tank\Addons\structures_f_tank.pbo
it's correct in the latest dev version
what is your definition of "loading up a save in the scenario"
do you mean multiplayer savegames? or using zeus in singleplayer?
for the missing faction panels
I think I fixed that recently-ish. There was a bug with search I think, where the faction icons would totally disappear.
This was a bug as of July 1st where if in a single player scenario I loaded up any save I had in the scenario a large selection of units would vanish from Zeus's placement menu that allows me to place down units, modules, markers etc, I had a ton of units from mods all no longer available in Zeus after loading up a save in the scenario. If I start a new scenario and never load up any save in it (iow don't save) all units will be available in Zeus
Oki I need to fix that, but thats different from that ticket
Needs new ticket with steps to reproduce, that can be followed step by step
Will do when at PC, thank you for paying it attention though
I didn't know what to mark it as in the dropdowns so something there might be wrong feel free to notify me and I'll just correct it
QA just tested and couldn't repro on vanilla :/
Couldn't repro with RHS AFRF either.
Even if you pick "official addons only", then if there's an AFRF guy in the mission then all the Zeus addons show, even after save+load.
So Zeus just has barely passable mod support...cool
So Zeus just has barely passable mod support...cool
what?
@idle arch Maybe specify a minimal modset for repro.
Will do when available again, thanks for the advice :)
Added an update to the repro steps, this one includes a mod-set (two mods) one made by me very quickly using malden 2035, zeus game master module and an empty zeus entity just if you need to follow along 100% and the other one contains some units that you ought to see disappear once the repro steps are followed
I merged your new ticket into the old one, you can edit tickets or post comments to update.
Will be tested tomorrow
Alright good to hear, thank you for listening and looking into it
BTW @untold sky I noticed something weird:
In this video, as you can see from the line intersection with the FIRE lod (the red line), the front part of the tracks is said to belong to 2 selections: "damagehide" and something else
But as you can see on the left side (the text that has [3]), the "damagehide" selection is getting animated (it shows the position and vectorDir/Up changing)
so it should belong to a bone shouldn't it? and shouldn't lineIntersectsSurfaces show only 1 selection as you said before?
Also plz add AIPathOffset to the returned array of getRoadInfo as well:
https://feedback.bistudio.com/T166542
oh wait @solid marten added that command not Ded 
https://feedback.bistudio.com/T147630
also I had a question about roadsConnectedTo
are runway road segments actually not connected?
the command doesn't return anything
their type is also "HIDE" 
should I make a ticket for a assignedVehicles <group> getter command? :3
yes please
I thought runway is not road
assignedVehicles <group>
https://feedback.bistudio.com/T166543 ticket created - Dedmen, quiiick! π
can group units be inside a vehicle that is not assigned to them? π€
they can be assigned to a seat, but the vehicle can be not assigned to the group (e.g a transport)
so assignedvehicles is different from "owned" vehicles
yes
so maybe assignedVehicles should be ownedVehicles or something else instead
maybe, there's something for you to think about π
you can currently do group _veh
a function is doable, but a command would be faster indeed
yep
private _grpVehicles = vehicles select { group _x == _grp };
```but meh
they are roads, but apparently they don't make "polylines"
oh ok
A vehicle can be in the vehicle pool of multiple groups at once. A vehicle itself has no "group pool".
I don't know if this is actually true because there isn't an assignedVehicles :P
A vehicle can be in the vehicle pool of multiple groups at once. A vehicle itself has no "group pool".
if so, what doesgroup _vehiclereturn π I suppose the last group that used it
if it returns something else than the driver's/commander's/gunner's group ofc
it should return the effective commander's group (which is not necessarily the commander)
and if there is nobody, grpNull? then my above request is absolutely not coverable by script
yes, just checked
so nope, useless here unfortunately
Empty vehicle has no group, even if they'd get back in when ordered to move far.
so https://feedback.bistudio.com/T166543 β add β if you love it π
You missed addVehicle :P
looks like first unit in a vehicle gets the vehicle as part of its group (group _veh)
doesn't matter which seat
no, not at all
updated π
πΌ
[Feature Request] New scripting command: assignedVehicles
https://feedback.bistudio.com/T153136
[merge the two?]
to complement
[Feature Request] New scripting command: unassignVehicle and unassignVehicles
https://feedback.bistudio.com/T153139
[Feature Request] New scripting command: assignedVehiclePositions
https://feedback.bistudio.com/T153161
[Feature Request] New scripting command: getOrderGetIn[State] and getAllowGetIn[State]
https://feedback.bistudio.com/T153157
If there is a bone, it will use that.
If there is no bone, it will return all selections, in your case seems to be two
apparently so
so Dedmen will do it quickly before I add another one π
I used search before, I swear!
Well here's my question. How can something move if it doesn't belong to a bone?
The position, dir and up of the "damagehide" selection is changing when I animate the object, so it should be a bone.
But lineIntersectsSufaces returns 2 selections for the front part of the tracks: "damagehide" and "something_else"
But "damagehide" is animated. how can it not be a bone? 
Oh wait I think I understand what's going on here.
that damagehide selection is not a bone but it's attached to a bone 
How am I gonna detect that now?! 
Can you plz add this command at least? 
https://feedback.bistudio.com/T150103
It saves me a lot of trouble
yeah but damagehide is not
it is actually
it's just not animated, it's a selection used to hide parts when the vehicle is dead
in this case the tracks apparently
but it does move. is that because it's attached to some bone that moves?
actually it moves with the turret
wait what?
the movement of damagehide bone is the same as the turret bone
oh are you monitoring selectionPosition on it?
cuz its gonna have a bunch of other crap inside it... like I think crew generally get put into damagehide
yeah and dir and up
yeah it'd be giving you the average position damagehide could have all sorts of stuff in it
if you move the suspension up and down via damper it'd also move
well this is what it looks like when I "reverse engineer" the animation based on pos and dir and up
(P.S: the model is generated using lineIntersectsSurfaces, not ripped
)
Ah so you want these commands so you can efficiently remake our models and animations π
no 
it's for my AI mod
I want to "cache" object LODs and their animations
so I can run stuff (e.g. lineIntersectsSurfaces, which I use for path finding, ray casts, etc.) in another thread using Intercept (because doing them in the main thread is a waste of time and FPS)
simple models are no problem (e.g. buildings with doors). complex models may give incorrect results (like that one)
plus how are some low poly models useful? 
and if you look closely that model has plenty of holes, which I can't fill because if I reduce the meshing step it takes a very long time
it's useless for pretty much anything besides what I described
if I wanted to rip your models and use them there are tools for that π
and if I did I sure wouldn't post my work here to get banned 
Besides using arma models on any other modern game would look terrible, not to mention the robotic animations which are even worst.
Just found a couple of pretty clear waypoint bugs, but I don't suppose there's any interest in fixing them:
- setWaypointCompletionRadius only affects progression of waypoints on the Zeus map view, not 3d view or group behaviour.
- The vanilla wheeled APC cannot complete some waypoints because it consistently stops slightly short of the progression distance.
looks like from here on out we're headed to v2.12 and v2.10 is closed 
which means tickets accepted from now on will probably not make it to v2.10
setDriveOnPath doesn't work while execution is paused (SP), which explains why people have trouble making the command work.
Not that they're missing out on much given how the command behaves when it's "working".
you can guide tanks and infantry pretty well with it
just need to put a lot of them in close proximity
which is why I asked for vehicle handling commands:
https://feedback.bistudio.com/T150072
It honestly feels like there's some low-hanging-fruit bugfixes to AI driving behaviour though. In the navigation failure cases I see there's nearly always a point where it's doing something obviously broken, rather than "yeah that's a hard pathfinding puzzle".
I can certainly see the attraction of overriding the stuck vehicle behaviour with a script that can do a passable approximation of a three-point turn, but that's only because the built-in version appears to have actual bugs.
I'd imagine depending on whats going on in the engine tweaking one thing can have adverse side effects to other behavior and to make changes requires rigorous testing to verify all behavior is still functioning correctly.
I'd love for the AI driving behavior to be improved as it constantly breaks in various ways that require lots of baby sitting to make them effective but its so far into the engine's lifespan (and has been an issue for so long) that I'm not expecting much on that front
Yeah iirc dedmen mentioned that a while ago as part of the reason ai changes dont happen mcuh / arent likely
The only real thing I'd imagine the community can do is providing tickets with 100% repro examples (test missions, scripts, etc) that are very easy to test but even with that making a change could still break something else.
IIRC tracked vehicles seem to have the biggest issues at the moment, wheeled vehicles seem to behave quite a bit better or maybe that's just my own skewed bias
I had a dumb idea that is similar to Leopard's request which was to make a script that let you select a vehicle in Zeus and then toggle into "driving" mode that would let you control the vehicle while in Zeus like a player commander would with AI drivers but I never looked into if it was even feasible with current limitations
Tracked vehicles have a weird habit of attempting to drive over houses.
kinda hard to replicate because it's generally not done under waypoints.
or generally just anything, almost always their own friendly units...
(when a mommy and daddy T-72 like each other...)

I consider that slightly more understandable given that other units move :P
I had an M113 on SAD yesterday that decided to mount literally the only building within 100m.
We basically cannot fix any AI driving dumbnesses, there are missions that depend on their driving timing
please at least give us some commands to fix it ourselves π
I mean this

I'd be fine with a mission file flag that gets automaticaly set for new missions post-update to divert logic into new behavior, its just frustrating that so much time is wasted wrangling AI and fixes seem to not be possible because of back-compat with older missions
I like the idea, but sounds like it could cause a lot of confusion
flag for improved AI driving would work.
we already have one that allows us to switch to the old one, works bit better on tanks IIRC.
man, even things like some unstuck routines would be god send, does not change "ai driving timing" but can help.
eg. if driving forward for 30s results in not changing the position maybe try driving backwards?
seen so many tanks trying to drive over a tree they can't tip over due to too low speed xD
@faint heart wrote us an "unstuck system" - among its features is to delete vegetation infront of vehicles when blocked by it for too long π
a few more tickets to consider for 2.12:
[Feature request] Add option in A3 task framework to hide tasks in 2d map
https://feedback.bistudio.com/T156986
[Feature Request] Add ability to adjust the number of tank tracks and foot steps shown
https://feedback.bistudio.com/T156441
Difficulty setting to -not- display helicopter/plane/tank without GPS on the map
https://feedback.bistudio.com/T151232
[Feature Request] Add ability to change weapon lock state
https://feedback.bistudio.com/T158616
[Feature Request] Separate default difficulty for MP
https://feedback.bistudio.com/T152901
[Bug][AI] When AI are ordered to get in a vehicle, the expectedDestination is always at 0 AGL height
https://feedback.bistudio.com/T153731
If I have a repeatable Arma crash on a CUP map (specifically with CUP interiors), should I report that on Arma's FT or just CUPs? Not sure on the expectations for crashing.
And to me probably too so I can see whats crashing and why π Instead of cup devs guessing around (not on the weekend)
does geometry of a cfgAmmos model effect its ability to hit? I believe I read somewhere it does for shotMissile. would it be possible to extend this ability to other ammo types?
would make stuff like AA proximity shells a lot easier as you could have it's geo lod be a sphere to represent the proximity detection
Afaik atleast normal shots (not missiles) just get calculated as a point, ignoring the model completely.
No I don't want to extend that
my bows & arrows mod!!
would it make sense to request the same for spawn?

it doesn't count as crossposting
I'm asking you
I was asking model makers there
also the timeline between the 2 posts is apart enough to not count as crossposting
and the two messages are completely different
but anyway I don't need your help anymore. I'll figure it out on my own
Don't be mean to the nicest guy around here. If you are the π§ here then leopard is the π« !
I think you released a profiling update? when trying to play a mission I made previously in eden, my game crashes
We just updated profiling to fix the issue. #perf_prof_branch
what has been the reason to limit this to Geo LOD only?
would be useful for any LOD and wouldnt leak anything sensitive, would it?
interesting gameplay issue ive noticed regarding Launchers. When a player fires a missile launcher, he reloads and needs to re-acquire a target lock. this adds a few seconds between effective shots.
When an AI fires a missile launcher, he reloads and does NOT need to re-acquire the lock, and is able to fire another effective shot quickly
at least, that is how it appears
Spam comment on my ticket π
https://feedback.bistudio.com/T166542
Spam comment down here too
https://feedback.bistudio.com/T166561
Sneaky spam comment: https://feedback.bistudio.com/T77829#2341448
Not anymooore π