arma3_scripting

[granite sky]

Note that if you run createMarker with a marker name that already exists, it returns empty string and the rest of the commands do nothing.

[spare idol]

what's the correct way to write the ampersand without the code freaking out?
I just want to have it inside a displayname: "" and it cuts the text.
I've tried "&&", "&" and "\x26" as suggested by some AIs but I can't get it right. This shouldnt be so difficult...

[warm hedge]

Post your code

And don't rely on AI

[lean dragon]

IDK what I'm doing. I'm trying to stitch together some sort of script but I'm not a scripter. I use AI, but dont trust it.
Anyways,

Objective: To create a script that pushes a boat on a dedicated server that does not rely on boat size or if its in the water or not. You should be able to push it if its 1000m on a mountain or on sealevel. This is what AI has come up with.
I add these odd requirements because publicly available mods make dependent if its in the water or if its too heavy you cant push. None satisfy all of these requirements

fn_pushBoat.sqf

    Author: Community Script / Adapted for Dedicated
    Description: Pushes a boat in the direction the player is looking.
*/
params ["_boat", "_player"];

// Ensure this runs on the server
if (!isServer) exitWith {
    [_this, "fn_pushBoat.sqf"] remoteExec ["execVM", 2];
};

// --- Configuration ---
private _force = 10; // How hard to push
private _lift = 2;   // Slight upward lift to prevent getting stuck
// ---------------------

if (isNull _boat || !alive _boat) exitWith {};

// Calculate direction from player to boat
private _dir = [_player, _boat] call BIS_fnc_dirTo;

// Apply force
_boat setVelocity [
    (sin _dir) * _force,
    (cos _dir) * _force,
    _lift
];

// Optional: Say something
// _boat say3D "push_sound"; ```

initPlayerLocal.sqf
```// Add action to all boats for all players
if (hasInterface) then {
    player addAction [
        "<t color='#00FF00'>Push Boat</t>",
        {
            params ["_target", "_caller", "_actionId", "_arguments"];
            [_target, _caller] execVM "scripts\fn_pushBoat.sqf";
        },
        [],
        1.5,
        false,
        true,
        "",
        "objectParent _target == objNull && _target distance _caller < 5" // Only when close and not in boat
    ];
};```

[proven charm]

thats AI code? maybe use https://community.bistudio.com/wiki/addForce instead

[lean dragon]

the problem is is that it doesn't show up in-game

and I am too ignorant to troubleshoot this so I come to the experts

[proven charm]

its ai code?

[lean dragon]

yeah from ai. I dont know where it got the code. it didn't source anythign relevant

[errant jasper]

Many issues. The technically broken ones

• The action is added to player. This is not inherently wrong, but nowhere is the boat captured, say via cursorObject or cursorTarget.
• The action runs on the player, but the script only operates on server.

[proven charm]

ai codes typically dont work 😄

[lean dragon]

usually I look at the links the ai gets me to confirm if it interpreted correctly according to the original author but idk who i t is

[lean dragon]

it did once, lol. I figured it sourced it from someone

[errant jasper]

Then you probably also want:

• Ensure the boat is not occupied (at least in the "driver" seat) (if not you need locality-run it correctly), or already moving when pushing.
• Maybe a push cooldown

[lean dragon]

well the one area I can isolate it to is initplayerlocal.sqf as the add action option doesn't even show up. Therefore idk if the rest of the script is trash

[errant jasper]

I think actions are also not preserved across respawn.

[lean dragon]

ouch

well I'm convinced someone somewhere has pushed a boat on a dedicated server

[errant jasper]

objectParent _target == objNull
^^ will never be true

that's why you don't see it.

[lean dragon]

oh should I modify this line or delete entirely?

[mortal folio]

objectparent can be null

but == objNull isnt the right way ofc, should be isNull

[lean dragon]

to confirm. I'm going to edit this line to:

object _target isNull ?

[mortal folio]

isNull objectParent _target

[errant jasper]

if (hasInterface) then {
    player addAction [
        "<t color='#00FF00'>Push Boat</t>",
        {
            params ["_target", "_caller", "_actionId", "_arguments"];
            private _boat = cursorTarget;
            private _bump = vectorNormalized (getPos _boat vectorDiff getPos _target) vectorMultiply 10;
            _bump set [2, 2];
            [_boat, _bump] remoteExecCall ["setVelocity", _boat];
        },
        [],
        1.5,
        false,
        true,
        "",
        "_target distance cursorTarget < 5 && cursorTarget isKindOf 'Boat_F'" 
    ];
};

Should get you started.

[lean dragon]

dude how do you learn this code? Its proprietary

its not like they publishing books about it. I'm impressed with you all

[mortal folio]

there's a wiki

so technically yes a book is published lol

[lean dragon]

lol I guess so. Like is it based on any other languages?

Python?

[mortal folio]

it's not really based on anything so it's technically Properietary in that sense, also it's a scripting language so cant really be compared to a complete programming language, tho that's also why it's much easier to learn, easier than python even, since all you gotta worry about is syntax and game commands

probably the only way sqf is comparable to python is in its performance (i havent measured but it probably peforms better than python tho 😂 )

[real tartan]

how to disable EndMission respawn template? want to prevent end of mission when respawn tickets reach 0

respawnTemplates[] = { "MenuPosition", "Tickets", "Spectator" };

[lean dragon]

do you know how to fix it so it reappears on the respawn?

[errant jasper]

AddAction_PushBoat = {
   params ["_unit"];
   _unit addAction [
        "<t color='#00FF00'>Push Boat</t>",
        {
            params ["_target", "_caller", "_actionId", "_arguments"];
            private _boat = cursorTarget;
            private _bump = vectorNormalized (getPos _boat vectorDiff getPos _target) vectorMultiply 10;
            _bump set [2, 2];
            [_boat, _bump] remoteExecCall ["setVelocity", _boat];
        },
        [],
        1.5,
        false,
        true,
        "",
        "_target distance cursorTarget < 5 && cursorTarget isKindOf 'Boat_F'" 
    ];
};

if (hasInterface) then {
   // Add Now
   [player] call AddAction_PushBoat;
   // Add again on respawn.
   player addEventHandler ["Respawn", {
       params ["_unit"];
       [_unit] call AddAction_PushBoat;
   };
};

[lean dragon]

I want to again thank you all. Once I test muzzleflash's fix I think all issues will be good. I should publish this on the workshop and give you all credit. Thanks again

[faint burrow]

Add it to the boat, not to player.

[errant jasper]

That is more proper. Though, I assume "complex scenarios" where it might not be known how and where boats will be spawned.

[faint burrow]

https://community.bistudio.com/wiki/Arma_3:_Mission_Event_Handlers#EntityCreated 😎

[mortal folio]

probably still more performant to keep it on the player, idk how expensive the internal engine distance checks are for addAction, but close to a dozen or more boats it's probably cheaper to just run the distance and type check on the targeted object

iirc that's how the "Unload incapacitated" action works (which might be from CBA but i cant recall)

[faint burrow]

Not sure.
Plus, according to BIKI, the action's condition will only get evaluated if the player is closer than 50m to the object and is looking at it.

[faint burrow]

It's vanilla action.

[exotic gyro]

... For what it's worth, I'm like 90% sure Ace already has a push boat thing

[mortal folio]

that's the check i meant actually 😄 there still has to be something internally that runs which determines if closer than 50 metres before enabling the condition check

that distance check would probably get more expensive with many many actions compared to just one check for whatever youre observing

[faint burrow]

Maybe.
Anyway, I think the game engine does it much faster and more efficiently than the scripted condition.

[mortal folio]

fair

[spare idol]

I'm just trying to have a displayname = "K & H Assault Rifle"
However, since the & gets interpreted as a function or something, in-game it breaks and stops at K. How can I write the & without this happening?

[warm hedge]

I said post your code, not your summary

In what context is also important

[spare idol]

Its either on config or an sqf, it doesn't matter. Not at the PC right now give me a few

this is the important line, but I'll paste the whole code below too:

displayName = "AR-12 K&H assault rifle";

it gets cut on K

[old owl]

I could be totally mistaken since been a while since I've had the issue but maybe you just need to escape the & if you're localizing or something?

[spare idol]

You mean doing &&?

[old owl]

Ah okay found it
https://community.bistudio.com/wiki/Structured_Text#Hyperlink

Should be & if you're using structured text

[spare idol]

without the semicolon?

I have tried it with the semicolon and it just literally wrote &

[old owl]

With semi colon sorry

[spare idol]

would this be the correct way supposedly?
displayName = "AR-12 K&H assault rifle";

it works

so, the problem was that before, I was checking on the arsenal

[winter rose]

it's better if you use a stringtable still

[spare idol]

apparently the arsenal and the in-game display behave differently

I was checking here

but it doesnt matter really

all I care is how it looks ingame after all

[spare idol]

how would that be?

[old owl]

If you intend to publish it in the workshop, stringtable is neat as you can localize the name to other languages :)

[spare idol]

yeah I dunno what that is, I literally know 0 coding haha

im managing how I can

[winter rose]

see https://community.bistudio.com/wiki/Stringtable.xml
it would turn out as displayName = "$STR_My_Translation_Key";

[hallow mortar]

A stringtable is a table...of strings...associated with a key, e.g. $MY_LOCALISATION_KEY. Instead of writing plain text in your config, you write the key. Then when the game is looking at the config, it finds the key, and looks it up in the stringtable. From there, it can pull out the string associated with that key and with the currently-selected language, and substitute in that string wherever the key is used.

You can use plain text if you're really not going to provide alternate languages, but if you do want to, then a stringtable is the way to do it.

[spare idol]

Ahhh okay. So it keeps the content of the text separately and easily modifiable or localizable

Would also help make my scripts look less bloated

[old owl]

We don't often use a string table in our mission since we require use of English to understand and abide by our rules, but if you plan to publish on workshop I'd recommend it. Kinda wish the gamemode we hosted was more inclusive in that way.

[hallow mortar]

* a side benefit is that if you're using the same text in multiple places, you can just use the same stringtable key, then if you want to change the text you only have to change it once

[spare idol]

I mean this is mostly a personal project I started doing purely for my own enjoyment. But this is good to know. I like the idea if only just to keep things more organized and flexible

Thanks for the help!

[cyan dust]

How can I find the path of the sound for playSound3D command? 🤔
Like, I have sound name that works with playSound but not with playSound3D

[old owl]

To find the path, you can look in it's CfgSounds which can be found in it's associated config: missionConfigFile, configFile, etc in the config viewer. That will give you an idea of what PBO to look into if you're trying to find the actual file.

[cyan dust]

Yeah, thanks, I was able to find it in CfgSounds with sound[] = attribute having path to file (without '.wss' extension though)
If someone else will search that question on discord:
Better check https://community.bistudio.com/wiki/Arma_3:_Sound_Files page, it has it all

[warm hedge]

addonFiles is also your friend

[versed ledge]

Can vehicles be scripted to move in reverse between certain waypoints?

[hushed turtle]

By default AI reverses only if destination is close enough

[split ruin]

what move command is responsible for flying ? 🤔 how jets fly actually ?
PS: I see ... vector array

[hallow mortar]

Aircraft respond to the same movement commands and waypoints as any other vehicle.
If you want to control altitude, you can try flyInHeight and flyInHeightASL.
Note that the basic AI movement behaviour is done by the game engine, not by commands.

[split ruin]

obviously I want to manipulate flying vector 🙂

[thin fox]

see if this can help you
https://github.com/PiG13BR/PIG-CAS_Menu/blob/beta/Missionframework%2FCAS_Menu%2Ffunctions%2Ffn_planeAttackPlan.sqf

[cursive tundra]

id like to know too - itd be super helpful for for my tank script...

[versed ledge]

trying to work something out with the LCC inside the LPD. LCC fits much better with the bow in, as far as entering, getting out of, and loading, etc. but then I want to try and script it to work as a taxi or support delivery ship. I think I can figure out the scripting of the support stuff, but it seems like it needs to start with the LCC moving out of the well deck in reverse.

I just recently started trying to really learn scripting and it's awesome! so much better than using the editor imo.

[versed ledge]

so probably trial and error to figure out the AI behavior?

[hallow mortar]

You can try sendSimpleCommand, but it's not as easy as "go to this position in reverse"

[versed ledge]

what fun would it be if it was as simple as "do what I want you to do" 😂

thank you, I'll check out that command!

[old owl]

I've said it before but man- every time I read anything here related to AI, sounds like black magic

[past wagon]

I'm encountering a VERY strange and disturbing bug. If anyone can help me figure out what's going on I would be very appreciative.

I'm making a battle royale where players parachute out of a plane. The Blackfish (largest plane) can only hold 32 players, but the game must accommodate 70 or more. Since I dont want to use multiple planes, I'm giving each player a local plane, and all of these local planes follow the same path. I'm temporarily hiding the players from each other while they are in the plane so they don't see each other de-syncing through the sky.

Basically, I'm creating the illusion that all players are in one plane.

Here's the unexpected issue: when players are parachuting down, something weird happens. When they get close to the ground, or right before they touch the ground, or apparently sometimes right after they pull their parachute, they go FLYING IN A RANDOM DIRECTION AND DIE.

I have no idea why this is happening, but it started after I implemented the plane code, which is otherwise working perfectly. Heres a video of it happening: https://youtu.be/QaaOeklvZDU

[old owl]

I've never seen something like that before so I am guessing it would be some sort of script issue. Do you happen to use some sort of loop with isTouchingGround? I know that command can be kinda hit or miss- had issues when I was making a jetpack. Do you have code examples you're able to share?

[past wagon]

heres the script: https://pastebin.com/AT9R7kDC

the script is remote executed from the server:

[_planeStartPos, _planeEndPos, _pathMarkers] remoteExec ["TRI_fnc_doLocalPlane", 0];

[past wagon]

I don't have anything happening with isTouchingGround

[old owl]

I could be totally waffling as I am purely guessing- but have you checked to see if player action ["eject", _plane]; is being ran while the player is in the air and in their parachute causing unintended behavior?

[past wagon]

ahh

yeah maybe

[old owl]

Or maybe one of the deleteVehicle commands? Looks like you're doing that on the plane so I wouldn't expect those to be it but only thing I can really see off a glance atm

[little raptor]

I think it's related to creating a local plane and moving a global object into it

I think the server doesn't know about your plane when you do eject, so it will eject you from the parachute

[past wagon]

so this will likely fix it?

if (vehicle player == _plane) then {
    player action ["eject", _plane];
};

[little raptor]

no

[winter rose]

call all actions related to a local plane locally

[little raptor]

does it work correctly if the plane is global?

(I mean not created using createVehicleLocal)

[past wagon]

I confirmed that its this with a systemchat

the systemchat fires the instant the player gets "ejected" from their parachute

[past wagon]

this entire script is being remotely executed for each client

[old owl]

Oh wait

Lou was just here, sorry probably a unecessary ping

[sharp grotto]

better safe than sorry

[past wagon]

yeah this fixed it

thanks, that was way simpler than I thought it would be

[tardy osprey]

Hey, I'm having an issue with setObjectScale on a dedicated server. The script works fine in locally hosted MP but fails on dedicated. I've tried remoteExec with 0 and 2, neither work on dedicated. Putting it in the object init field doesn't work either on dedicated. The script is called from init.sqf. Anyone know how to get it to work?

Script:

[atc, 0.50] remoteExec ["setObjectScale", 2];

[little raptor]

setObjectScale only works on attached and simple objects.

[tardy osprey]

wait huh, wtf. thats interesting... what do you mean by attached btw?

[little raptor]

I mean attached using attachTo

[tardy osprey]

ah i see, yeah that makes sense. Alright lemme just quickly look into it then.

[little raptor]

also your remoteExec is wrong. it should be local arg not server:

[tardy osprey]

Okay that is a different ballpark entirely. Not quite sure what you mean by Local arg. I clicked the LA one, and tried to have a look, but it doesnt really make much sense.

[little raptor]

I mean it should be executed where the object is local: [atc, 0.50] remoteExec ["setObjectScale", atc];

[old owl]

This page also has a lot of really good info regarding locality 🙂

https://community.bistudio.com/wiki/Multiplayer_Scripting#Locality

[granite sky]

Local vehicles can and do have destructive effects on global objects. It's a common issue when you're making vehicle-placing UIs.

[tardy osprey]

Hm okay. I put in your version of it instead. found an asset with Simple Object, ticked it, but didnt do anything. Still works fine when hosted on my own machine SP and MP.

[tulip ridge]

I'm not really sure how to phrase this question, but I'm making a toxic gas-esque smoke grenade. Here I found some code from a WIP ace feature that checks if a unit is in smoke, but I want to be able what kind of smoke / particle is in the player's face.

https://github.com/BrettMayson/ACE3/blob/370da581027009d2e65e9e9be2b064741d4ea9aa/addons/irritants/functions/fnc_isInSmoke.sqf

Is that kind of thing possible? I'd really like to be able to check the type of smoke directly rather than just "If within X meters of smoke grenade -> damage player"

[warm hedge]

Theoretically, this is entirely possible, as you can really get at least the scale, velocity, position of a dropped particle as an object. But definitely not ideally

If the toxic gas is made from one particular particlesource, I would drop multiple dummy particle drops and track the dropped particle to get where it did go

[tulip ridge]

Haven't really messed with particles all that much, so I'm not really sure what you mean. Currently I just have a smoke grenade with a custom CfgCloudlets class

[warm hedge]

Which I meant was like, scripted particles either with setParticleParams or drop, but for CfgCloudlets thingy, ain't really sure

[old owl]

Might be good info here- I had a similar thing of needing to know where my particles were ending #arma3_scripting message

[tulip ridge]

I'm not entirely against scripting the smoke effect and just giving the grenade a dummy effect if it means I can then check if the player is in it

[ivory lake]

I just did a similar thing the basic way by having a set radius

its not perfect but the particles I used werent affected much by wind

and tbh relying entirely on the particles might lead to some oddities

ie. the grenade being set off in a building affecting the floors above/below/through rooms etc

does kinda make me wish the 'fire' particle stuff could be opened up for scripting instead of just being completely engine driven damage

[tulip ridge]

Yeah I was trying to think of a way to hook into that maybe, but the ways I thought of would also mess with ace since I don't think there'd be a way to tell between them

[ivory lake]

you might be able to do some funky stuff like
use a set radius (but much larger than the 'expected' size of the cloud)

and checkvisbility to see if they're inside the cloud

but I dont think that'd be 100% accurate as other stuff might mess with the visibility

ah thats what the isInSmoke fnc does

I hadnt looked at it till now

[warm hedge]

My idea would be // psuedo code globalValue = []; private _dummy = drop []; // drop one along with your visible particle _dummy spawn { while {alive _this} do { globalValue pushBack [getPosWorld _this,getObjectScale _this]; sleep 1; }; };and refer globalValue as the effective range of the toxism

[sage kettle]

Looking for a script to prevent the AI from dismounting immobilized vehicles when the gun still works. Used to have one years ago but took break from the game and dont have it anymore

[warm hedge]

https://community.bistudio.com/wiki/allowCrewInImmobile

[sage kettle]

Thank you.

[little raptor]

you can use allObjects to get all cloudlets, but it's slow to iterate over them in scripts since there can be thousands of them

I once wanted to add an object filter command with many parameters (side, type, etc.). maybe if I add that it can help

[tulip ridge]

Hm, maybe. I imagine I'd have to fill out a decent number of the drop params so that stuff like how much it moves with the wind would match?

[warm hedge]

Ja. It wouldn't be so hard to do so, since there is a dedicated parameter to see how the wind (ultimately the air itself) can slow down/accelarate the drop

[tulip ridge]

Yeah, rubbing iirc
My main thing is that I'm going to be tweaking the visuals some more and it'd be kind of a pain to keep editing it in two spots

Could read it from config but kinda annoying to cache it

[tulip ridge]

That'd be great, Kerc also mentioned potentially opening up the damageType to script (which currently only has "Fire" as an option) could also work

Something like: damageType = "tag_fnc_particleHurt"; which gives the _unit that was hurt and then the _cloudlet (class) causing the damage? Could do the object too since that seems to be a thing with your allObjects message

[warm hedge]

!purgeban 934386452685529100 1d crypto scam spam

[lyric schooner]

*fires them railguns at @iron swallow* Ò_Ó

[young mist]

Whats a script I can make or use to make sure a Vehicles waypoint only activates once a specific individual is placed in it? Like a Helicopter.

[faint burrow]

specificUnit in (objectParent this)

[granite sky]

There is any built-in way to constrain waypoints with arbitrary conditions. So normally you'd need a monitor script that created waypoints, or toggled lockWP.

[hallow mortar]

Waypoints do have script conditions that can prevent them from completing. Rather than preventing the target waypoint from activating, you'd have a waypoint before it that is prevented from completing until the condition is met. (https://community.bistudio.com/wiki/setWaypointStatements)

[granite sky]

oh huh, I forgot setWaypointStatements had a condition.

I only ever used the statement bit.

[topaz lion]

Has anyone ever come up with or come across a script for getting ai to swap or equip gear that is in their inventory?

I’m looking into it for undercover squads as well as scuba teams, so they could be wearing just plain clothes in the undercover case and wetsuits + rebreathers in the scuba case.

Then perhaps via a radio trigger or something I could get them to equip or swap their vest slot to swap rebreathers for plate carriers for example? Not sure if I’m explaining it very well.

[hallow mortar]

The way to do that would be:

• save a list of everything that was stored in their vest
• add the new vest (this deletes the old one and everything in it)
• restore the contents of the vest from the saved list
  Commands to start off with: vestItems, addVest

Or another way would be to use getUnitLoadout, modify the resulting array, and then setUnitLoadout with the new loadout

[modest kiln]

im no great coder but can you guys find a way to make this better

{ unassignVehicle _x; _x action ["Eject", vehicle _x]; } forEach (assignedCargo vehicle this);

its basically supposed to eject passengers from a vehicle and it works to a extent but has issues

• helicopter does some kind of wierd landing and takes off again while dropping off paratroopers
• somewhat inconsistent
• usually needs flat ground to work correctly

[proven charm]

hmm for one thing change action eject to https://community.bistudio.com/wiki/moveOut not sure about the landing issue

[granite sky]

Generally if a unit wants to leave a helicopter then it'll force the crew to land.

[proven charm]

maybe [_man] allowGetin false; too

[granite sky]

moveOut does avoid that, I think.

Probably want to do moveOut before any unassigning, to avoid them requesting a landing.

[proven charm]

oh and unassignVehicle takes local argument https://community.bistudio.com/wiki/unassignVehicle

[modest kiln]

private _crew = crew _vehicle;
{ unassignVehicle _x } forEach _crew;
_crew allowGetIn false;

so i could do this but with assignedCargo instead of crew

[proven charm]

and moveOut

[modest kiln]

so like

private _assignedCargo = assignedCargo _vehicle;
{ unassignVehicle _x } forEach _assignedCargo;
_assignedCargo moveout true;

[proven charm]

this is the function I use, use it if you wish ```sqf
manLeaveVehicle =
{
params ["_man"];

_man remoteExec ["unassignVehicle",_man];

[_man] allowGetin false;

moveOut _man;

};

[sly cape]

!sqf

[wicked roost]

How to use SQF syntax highlighting in Discord

```sqf
// your code here
hint "good!";
```
↓ turns into ↓

// your code here
hint "good!";

[topaz lion]

Great thank you, certainly a starting point I can use, I’ll probably be back when I’ve figured out a bit more

[old owl]

Hi friends! I was thinking about creating a Mod that would contain a UI with some tools. Before I start though, I've been wondering what's the best way of spawning that UI without causing a compatibility issue between other mods? Obviously plan to use my own variable prefixes and stuff but not sure really how to approach adding a control or key to open the menu without potentially causing conflict with another mod someone may download.

[sly cape]

Use #arma3_scripting message to make it format nicely in Discord.

[proven charm]

so you can do { _x call manLeaveVehicle } foreach (assignedCargo _vehicle);

[hallow mortar]

You can't outright prevent a possible clash of keybinds; there's only so many keys on the keyboard and some mod will be using just about any one of them by default.
What you can do is register your key as a proper keybind, so people can easily change it and will be warned about possible conflicts (at least if the other mod has also made proper keybinds...)
https://community.bistudio.com/wiki/Arma_3:_Modded_Keybinding

[proven charm]

@Milo its not too difficult just use unique dialog Ids and class name

[old owl]

Gotcha, that was my worry with keybinds; especially assigning it to a specific key. What you posted there would just give the ability for the mod to have custom keybinds in the default keybinds menu?

[hallow mortar]

Yes

[old owl]

That's awesome and sounds like exactly what I'm looking for- thank you for the help!

[modest kiln]

is it specific to normal arma helicopters because im using the blackfish vtol from apex and/or does it still work on waypoints

because i assigned a move waypoint and changed it to scripted as well but it doesnt work

im pretty sure im doing something wrong on my end though

[proven charm]

should work with any aircraft, the script just throws the men out

make sure your _vehicle or whatever is vehicle variable is correct

run the manLeaveVehicle function declaration code in init.sqf for example

[grand stratus]

I have issues on performance branch with VTOL ai behavior. If you're on perf try swapping and see if it helps

[modest kiln]

im not on performance but it just seems that the vtol generally has wierder behavior which im assuming because it was a wide turning radius compared to a normal helicopter

buts its just a guess

[proven charm]

maybe too small waypoint radius? 🤔

[hallow mortar]

The AI don't really know how to fly VTOLs properly, but it shouldn't make any difference to moving units in/out

[modest kiln]

that is a good point ill try to make a bigger one

[proven charm]

well i have to go, good luck with the script!

[modest kiln]

alr thanks

[unreal scroll]

Still actual:
#arma3_ai message

#arma3_scripting message

[rose pike]

do i have the wrong formatting or am i going about this wrong?
i have 3 tasks that once all are completed, a 4th task is supposed to show up
i have
taskCompleted task1 and taskCompleted task2 and taskCompleted task3;
on the final task trigger, but the 4th task is showing up at the start of the mission

[jade acorn]

&& rather than and

and ditch the ; at the end

[hallow mortar]

and is valid (it's exactly the same as &&) and the end ; won't break it

[jade acorn]

thought triggers specifically need ampersands rather than a word

[hallow mortar]

No. It's just a command. Triggers don't do commands any differently than any other part of the game.

[radiant lark]

how much is dedmen's backlog? shall I ping? 😅

[past wagon]

is there a way to prevent the command menu from coming up when players press a key like F3?

(the command scroll menu that pops up in the upper left)

[old owl]

Can't reference our code base since on my phone but could probably just check to see if the key for it is pressed by using actionKeys on it's associated userAction in a key event handler. Not sure what the userAction is off the top of my head but could probably be found here 🙂

https://community.bistudio.com/wiki/inputAction/actions

[molten yacht]

I was browsing the wiki today and I noticed that there are event scripts, onFlare.sqs and init3DEN.sqf

I understand what they do, the documentation is clear, but I wonder if anyone has any interesting examples of how they can be used? I also wonder if onFlare works with ACE3/mod flares

[rose pike]

then what's wrong 😔

[old owl]

Ah okay back on my PC now- this may be what you're looking for:

(findDisplay 46) displayAddEventHandler ["KeyDown",{
    params ["_displayOrControl", "_key", "_shift", "_ctrl", "_alt"];

    if (_key in (
        actionKeys "CommandingMenu1"
        + actionKeys "CommandingMenu2"
        + actionKeys "CommandingMenu3"
        + actionKeys "CommandingMenu4"
        + actionKeys "CommandingMenu5"
        + actionKeys "CommandingMenu6"
        + actionKeys "CommandingMenu7"
        + actionKeys "CommandingMenu8"
        + actionKeys "CommandingMenu9"
        + actionKeys "NavigateMenu"
    )) exitWith {true};
    
    false
}];

[past wagon]

this is what I already have, which isnt working

    if (_key == 61) exitWith {
        private _playerItems = missionNamespace getVariable ["playerItems", createHashmap];
        private _uid = getPlayerUID player;
        private _items = _playerItems getOrDefault [_uid, [0, 0, 0, 0, 0], true];

        private _redgullAmount = _items select 0;
        if (_redgullAmount == 0) exitWith { true };

        _items set [0, _redgullAmount - 1];

        [[-1, 0, 0, 0, 0]] remoteExec ["TRI_fnc_updateItem", player];
        [player, getPlayerUID player, 0, 180] call TRI_fnc_handleBuffs;

        true
    };

[old owl]

I would try what I tossed above, that should work

[dusk gust]

Keep in mind, if they rebind any of the commanding menus to say a double key, or mouse buttons, this will NOT work for it.

I've found that just using something similar to showCommandingMenu "" with an onEachFrame event is the only way to ensure that it never opens, or if it does, it will close immediately.

I believe a new-ish command can do something that you're wanting though, which is
https://community.bistudio.com/wiki/actionKeysEx

[jagged mica]

technically speaking

could the procedural textures, which come from the UI rendering, be used in rvmats in other maps?

[winter rose]

new-ish
2022-08-23 :p (I know, it's quite recent in Arma time :D)

[jagged mica]

as in, I want to have a dynamic normal map in my model,and while I am able to put it on the object as a texture no problem, the rvmat is refusing to render it

oh nevermind, it worked

[warm hedge]

Yeah it should, oh nevermind, you figured

[jagged mica]

just need to find how to use different shaders

I wanted to take a look at the refract shader, any idea how that's built and what the rvmat params are for it?

[warm hedge]

Clarify, why you need to? A shader itself is never meant to modify

[jagged mica]

so Im trying to make a glass material with some refraction on it, I made an rvmat with pixel & vertex refract shaders and passed my nohq map to it but my first problem is that it's animated, and it seems to be splitting the r & g channels, I was curious if there's a way to control this behaviour with a parameter/texture

also side fun fact, this sort of glass is possible if you write your rvmat wrong enough

here's the refract shader in question without any normal map passed to it

[jagged mica]

Unless someone knows if there's a way I can find an rvmat using this shader without having to go through each basegame pbo

[warm hedge]

"P:\a3\data_f\ParticleEffects\Universal\Refract.p3d" uses it. Probably that's not the shader you wished for though

[proven charm]

its possible to find .rvmat using this script if you modify it to find rvmats: https://gist.github.com/HallyG/fa7a6cda10abcb630b1dc325f0523553 HTH

[jagged mica]

Gotcha, I'll take a look, Maybe I'll learn something from it

Otherwise im gonna have to stick to a transparent super shader with a normal map for some illusion

[delicate tangle]

How can I get a trigger that is synced to a sector via a script? The sector is synced to an area logic which is then synced to a trigger. For instance: Sector—>Area—>Trigger.

[faint burrow]

https://community.bistudio.com/wiki/synchronizeObjectsAdd

[radiant lark]

Let me know because I also wanted to do this but ppeffects seem to be fullscreen

And I wanted to make a cool glassy HUD

[jagged mica]

well I found this

https://gist.github.com/dedmen/90691e56c2dd77bc152ce0a48caf47ba

there are more stages and it's explained what they are, although little info on how to actually work with them

i see that changing the refraction depth (Stage 11 in the shader) makes a lot of difference to how it works

[delicate tangle]

Wiki seems to indicate this only pulls logic and ai entities. Would that ignore triggers then?

[jagged mica]

one value gives blue and yellow, another gives a water shader, not sure what's the limit here tho

[hallow mortar]

First of all you probably want this command instead: https://community.bistudio.com/wiki/synchronizedObjects
Secondly, half the point of syncing is to work with triggers. The commands work with triggers.

[delicate tangle]

Im an idiot, the wiki notes it includes triggers. Reading is hard

[jagged mica]

I take it back, I have no idea whats going on with this damned shader

this feels more like a question to the BI devs if they could peek under the hood and check what's even going there and how we're supposed to set it up

[jagged mica]

I ended up making a bug report

[graceful kelp]

I need some assistance, not sure if im being stupid, I was under the impression config lookup was expensive, no matter what way i test it a config look up is quicker on its own, than to have a system which on first instance sticks it in a variable then calls the variable every time, has config lookup been optimised

[faint burrow]

Post your code.

[topaz lion]

Would anyone know of a script for getting the turret number of a vehicle? Trying to block off certain seats in a chinook but there’s so many seats I’m having trouble actually finding what number turret it corresponds too to stop ai from loading into that position, figured there would be a way get the position if I’m in said position

[errant jasper]

Who told you it was expensive? Though still surprised a variable isn't faster. Something does sound wrong.

[graceful kelp]

params ["_heli","_configPathArray", "_type"];

//Test variable for enviroment
_heli = vehicle player;
_configPathArray = ["CfgVehicles", "fza_ah64d_b1", "components", "SensorsManagerComponent", "Components", "ActiveRadarSensorComponent", "AirTarget", "maxRange"];
_type = "number";

private _configHash = _heli getVariable ["fza_ah64_configHash", createHashMap];
private _cacheKey = str _configPathArray;
private _configValue = _configHash get _cacheKey;
_configValue;

if (isNil "_configValue") then {
    private _entry = configFile;
    { _entry = _entry >> _x; } forEach _configPathArray;

    _configValue = switch (toLower _type) do {
        case "number": { getNumber _entry };
        case "string": { getText _entry };
        case "array":  { getArray _entry };
        default        { nil };
    };

    if (!isNil "_configValue") then {
        _configHash set [_cacheKey, _configValue];
        _heli setVariable ["fza_ah64_configHash", _configHash];
    };
};

if (isNil "_configValue") exitWith {-1};
_configValue;

the get variable on its own is quicker, however the moment I add more code to build the system around it, it ends up slower and the difference from a getvariable to a configfile is 0.0004 vs 0.0008, its so negligeble, its almost not worth thinking about

[errant jasper]

This looks really over engineered.

[graceful kelp]

im accounting for any config path as an input, checking if its in the hash first then getting it and setting the hash

[digital hollow]

get into the seat yourself and then run cameraOn unitTurret player

[errant jasper]

Well how often do you need to look up the values that the difference matters? Compared to read-only config data, stringifying arrays and hashtable lookups is not really free either.

So the caller knows the entire config path and the desired result type. I don't see why it would not be simpler and more performant for the caller to do:

private _heliConfig = configFile >> "CfgVehicles" >> "fza_ah64d_b1";
private _activeRadar = _heliConfig >> "components" >> "SensorsManagerComponent" >> "Components" >> "ActiveRadarSensorComponent";
// get actual values
private _maxRange = getNumber (_activeRadar >> "AirTarget" >> "maxRange");
private _minRange = getNumber (_activeRadar >> "AirTarget" >> "minRange");
// etc...

Compared to:

[_someHeli, ["CfgVehicles", "fza_ah64d_b1", "components", "SensorsManagerComponent", "Components", "ActiveRadarSensorComponent", "AirTarget", "maxRange"], "number"] call WhateverYourFunctionIsNamed;

And why store it on the helicopter. You just means multiple helicopters each pay the "cost" (that is not really that expensive) that you are trying to avoid in the first place.

[graceful kelp]

there is a fair amount of varying config lookup through the mod, I was attempting to build a universal call function that would check the requested path and store it for better future lookups

however if the config lookup is only 2-3x the performance hit of a getvariable, it may be a moot point, I was under the impression it was significanty more expensive to perform

[errant jasper]

Always measure first. And as my example demonstrate you can store intermediate config nodes in variables (or other things).

[topaz lion]

Perfect thank you

[graceful kelp]

Unless im mistaken I having looked into it believe config lookups were made more performant by dedmen last year

[old owl]

From my understanding config has always been expensive. Caching config data in a hashmap should about always result in faster look up time.

I believe hashmaps were made significantly quicker though in recent years.

[errant jasper]

Comparing these two on my machine:

private _heliConfig = configFile >> "CfgVehicles" >> "B_Heli_Attack_01_dynamicLoadout_F"; 
private _activeRadar = _heliConfig >> "components" >> "SensorsManagerComponent" >> "Components" >> "ActiveRadarSensorComponent"; 
private _maxRange = getNumber (_activeRadar >> "AirTarget" >> "maxRange"); 

Takes 0.00174 vs:

private _arg = ["CfgVehicles", "B_Heli_Attack_01_dynamicLoadout_F", "components", "SensorsManagerComponent", "Components", "ActiveRadarSensorComponent", "AirTarget", "maxRange"]; 
private _key = str _arg; 
private _result = cacheConfig get _key;

Which only takes 0.00141, but without handling the actual lookup part, only the happy pre-cached path.

[old owl]

The reproduction above was done utilizing configs we had in our missionConfigFile so it's renamed but I've always found that there is a significant performance advantage

[errant jasper]

Doing the raw config (without even getNumber) takes 0.00160s (which you have to at least once for the caching case too) of that 0.00174s

[errant jasper]

Are you "caching" the entire config tree?? Why would you nest it the hash map like that otherwise ?

[old owl]

In the example that I had, that's what I was doing. It was really just for the sake of messing around and testing config performance. The majority of our configs though we will cache individual data in hashmaps based off of a key. For example:

someMap getOrDefault ["className", ""]

Which for the sake of on frame code or event handlers that we intend to run quick, I've found storing config data in a hashmap is always worth it

[errant jasper]

I guess I live in a different world. I can't imagine why I would run config lookups per frame.

[old owl]

Well I mean we don't so we are probably living in the same one 🙃

[graceful kelp]

A config lookup is costing me 0.0008 vs a getvariable of 0.0004, not including all the other code to sort the input, output, hash

a config lookup is looking cheaper, from the looks of it dedmen has hashed config on the latter end of last year

[old owl]

Not a super common occurrence we have to even use hashmap look ups in our on frame event handlers, but for drawIcon3D we do have some stuff that makes it sorta necessary.

My results I had here #arma3_scripting message were produced December of last year, but no idea if I was on perf or stable at the time. Wish I'd written that down. Maybe there has been some changes since then.

[errant jasper]

Seem to cost about 0.000166875 per >> for me.

[graceful kelp]

considering the tests are running x 10k, I cant see why I wouldn't stick with a config lookup now, unless I am seriously misunderstanding something here

[errant jasper]

Lol / is like tiny bit slower than >>.

[old owl]

If it shows faster and you're on stable definitely use config. What I tested December of last year showed me the exact opposite but maybe something changed. If you're on perf though maybe that's the reason why. Honestly not sure.

[old owl]

I remember seeing that on the code performance wiki, not sure why, but interesting haha

[errant jasper]

Guessing it is something about the overload of / for division

[old owl]

Wouldn't have thought about that, but you're probably right

Probably similar to why # is faster than select, less syntax and usage elsewhere

Also interestingly looks like config paths stored in variables save performance. Makes sense, but not sure I've really quite thought about it tbh.
https://community.bistudio.com/wiki/Code_Optimisation#Config_path_delimiter

[errant jasper]

To be clear I am not saying suggesting caching is always dumb. Just have to be wary of how it is done. If it takes work to construct the key you are already chance of losing:
Like:

private _arg = ["CfgVehicles", "B_Heli_Attack_01_dynamicLoadout_F", "components", "SensorsManagerComponent", "Components", "ActiveRadarSensorComponent", "AirTarget", "maxRange"]; 
private _key = str _arg; 

Takes 0.00104 out of the 0.00140 to finalize the lookup:

private _result = cacheConfig get _key;

And thus we can also reason that a get "b" get "c" get "d" is not "free" either e.g. 0.00040 which is about 2.5 >> ops.

[graceful kelp]

I was on perf however same performance on stable, it looks no more expensive then any other operation in sqf really

[errant jasper]

So just be mindful that being 3 times as fast in your lookup doesn't help if the caller ends up taking 5 times as long to make the key.

[graceful kelp]

my whole script ended up being slower than straight config lookup

config was 1.12X faster than my script function to check a hash first

[old owl]

I understand and get that. Kinda just depends on whether the trade off is worth it. For us, we create those maps on players initialization while they're awaiting player data that is queried from a database, so it's not really making any kind of actual impact as they wouldn't be ready to play anyway.

Especially in the case of frequent look-ups, makes the trade-off worth it for us. You guys seem to be producing different results from what I had from December and before though, so perhaps something has changed too.

[errant jasper]

No, you are not wrong, the raw hash map lookups are fast, even for nested hash maps (Disclaimer: at least when they are small, these only have a single entry):

cacheConfig get "a" get "b" get "c"

only takes 0.00048s.
Just have to compared against entry construction (which matters litte for you at startup), and actual computing the keys themselves.

[broken pivot]

Gooood morning Arma nerds 😄
Im currently writing a vehicle load function. Really not a big deal, but somehow Im not able to write it
to the end.
The plan is to call

[_this] remoteExec ["EBER_fnc_loadChinhook", 2]

via the EventHandler. My problem is that tthe first addAction gets correctly crreated, but not the sesconnd.
Code:
https://pastebin.com/rdRcw9zY

[broken pivot]

To get a little easier into this:
I expect the issue to be somewhere in the params of the inner addAction (green line)
Can someone analyse my screenshot and tell me if its even possible, what I try there
I never saw something similar before

[warm hedge]

I'm not saying it's not broken, or it is, but why you can tell something is wrong? What do you expect and what you got?

[broken pivot]

Good questions! Ill annwser them all

[harsh vine]

_selectloadingtarget params ["_selectloadingtarget "]; detach _selectloadingtarget

[broken pivot]

The script is ment to run on chinook helicopters to create a while clock that checks
if selcted vehicles are in range. If so, the player should become able to load one vehicle
into the helicopter via attach to,
I expect the outer addAction to remove itself on activation. The outer addAction should create
the inner addAction .
The inner one should contain the unload interaction and the ability to rerun the script on the helicopter
so the scanner runs again and the player stays able to load up the vehicle

[broken pivot]

What am I seeing? Is it the red line?
What does it change and how does it work?

I see its the second syntax of https://community.bistudio.com/wiki/params
Can you ttell a little about its benefits and usage?

[warm hedge]

...And what do you got? What is wrong with your result?

[broken pivot]

You ment like tthis?
I gues it ensures a safer variable forewarding (?)

[broken pivot]

The car is able to be loaded. But the inner addAction (green marked "unload" actionn)

[warm hedge]

So you saying the outer is done its job right, then you can select inner, but it's not doing its job?

[broken pivot]

80% correct understood
The car gets loaded in -> LoadIn action removes -> No unload action shows off

Thats the matching stringtable.xml in cases anything is wrong with the title

<?xml version="1.0" encoding="UTF-8"?>
<Project name="AttachTo">
    <Package name="AttachTo">        
        <Key ID="STR_LOADINTOCHINHOOK">
            <English>
            Load %1 into Chinhook
            </English>
            <German>
            Lade %1 in den Chinhook
            </German>
        </Key>
        <KEY ID="STR_UNLOADINTOCHINHOOK">
            <English>
            Unload %1 from Chinhook
            </English>
            <German>
            Entlade %1 aus dem Chinhook
            </German>
        </KEY>    
    </Package>
</Project>

[warm hedge]

So it is about inner addAction doesn't show it?

[broken pivot]

Yes exactly 😄

[warm hedge]

Who is trying to activate the action? From where?

[broken pivot]

Im currenntly runninng the funnctionn inn the innit of the helicopter (_ttarget)

[warm hedge]

And where is the player to activate it?

Is it in the car? Or heli? Or somewhere else?

[broken pivot]

Im currenntly runninng the funnctionn inn the innit of the helicopter (_ttarget)

I also checked if its onn a diffrennt car or at the loaded vehic, cant find it

(Definitly not at a wrong vehicle)

[warm hedge]

Then you might want to assign the inner action to the car not the heli

[broken pivot]

Thats a good thing I actually just noticed atm.
The unload interaction needs to lay at the Chinook helicopter same as the driver and the pilot

But one by one. Why isnt it currently available for people who look at the helicopter?

TThats tthe updated sscript that is currently ussed. If you would like tto research the scripts logics
https://pastebin.com/xbzQUK7p
The screenshot is taken from line 78 and below

I also had a theory that the script ends before it launches tthe inner addAction.
That would be a simple scheduling issue. Is that possible if you think about?

[warm hedge]

A code will not terminate itself without an order to do so

[broken pivot]

Okay, so we can throw my theorry
What else is a possible issue?

[warm hedge]

I already told one, you are assigning the action to the heli, not the car, and you might not actually looking at the heli engine assumes so

[broken pivot]

(Bad screennshot sry)
That one is pointing at the chinooks hull. It is not available on the helicopter

[warm hedge]

Then you want to make sure that the action is actually added or not. Check the actionID of the inner one

[broken pivot]

That is my current main expectionn that doesent happen

[broken pivot]

Thats a great Idea. Maybe its not visible -> Stringtable/ Text

The easiest way I do see is to add a public var to the addAction commands and publicVar

Do you agree? Or am I wrong with my debug?

Thats it.

[harsh vine]

why _this params ["_target", "_caller", "_actionId", "_selectloadingtarget "]; and not params ["_target", "_caller", "_actionId", "_selectloadingtarget "]; i dont knw if _this will work

[warm hedge]

It will

[broken pivot]

So I need to modify my code into "select 2" somewhere in the inner addAction scope

[warm hedge]

Then chances are your text is invisible or you are simply not seeing the action

[broken pivot]

Is it because the removeAction is (maybe) synchrone with the addAction? Im also confused that the _actionID == 2
Its also odd because _this should contain the current actionID inside its scope...

I gues its time to look around to mayybe find somethinng else what could the causer

|| No hint... ||

KEY != Key

I puke

Piece in my soul.

Now I can continue. The next steps will be some logic polishishing. I already saw something else
but at first the more important thing:
A huge thanks to @warm hedge who investigated in and looked about the scheduling, wich always takes time& ennergy ❤️
And also probs to @harsh vine who showed me the alternate syntax and faced me with my scope problems I always have haha

Thats a win:
https://youtu.be/rev-I4L16g8
(Ive pasted the script in the video description its fully OpenSource)

[opal bloom]

Hell yeah! Congrats on getting it there. And thanks for sharing the script

[split ruin]

of course my first thing to do is testing it with tanks and strykers 😆

[eternal ingot]

if anybody has time and some experience with BIS_fnc_showSubtitle; I Would like to know how to get voice lines to work with it, I do know about the subtitles but not the voice lines

[hallow mortar]

BIS_fnc_showSubtitle doesn't have the ability to play sounds. It only does subtitles. If you want to play sounds while the subtitles are on screen, you'll have to do that yourself, using e.g. playSound, say3D, etc.

[eternal ingot]

ya i actually tried that but couldnt get that to work

would it be something like this? ```// 1. Play the sound globally (everyone hears it)
[player, "MyVoiceLine"] remoteExec ["say3D"];

// 2. Show the subtitle
["Unit Name", "This is the subtitle text that matches the voice line."] spawn BIS_fnc_showSubtitle;```

ya i got it, That code did work, But thanks for the help bro

[old owl]

This is just one part (and currently really only part) of an addon I am making to make complex command arguments easier. When I'm done I'll probably toss it up as a repo open source. Not sure if anyone is interested but just a little hobby project I started I figuerd I'd share FWIW :)

[crimson lion]

Oh that is absolutely fantastic, please do release that when you're wrapped up lol

[cyan hill]

trying to create a tp script and it keeps giving me a error this addAction ["Teleport to Helipads", {player setPosASL (getPosASL Pads)}];

[tulip ridge]

You didn’t say what the error was

[cyan hill]

mb

[tulip ridge]

Well, did you read the error?
There's no object called helipad1

[cyan hill]

there is a empty map marker with a varible name as helipad1

[tulip ridge]

Use a game logic if you need a truly dummy object

[tulip ridge]

A marker isn't an object

And also that'd give a different error, since getPosASL doesn't accept a string

[cyan hill]

so how do i fix the get posasl thing?

[tulip ridge]

Place a game Logic object where you want to teleport the player to and give it the name helipad1
You'll also need to remove the variable name from the marker you have

[cyan hill]

ok thank you

[warm hedge]

Alternatively getMarkerPos "YourMarker" if you prefer marker

[old owl]

Okay here is feature number 2.

If I am being quite honest, as much as it works well, I am honestly not too proud of how it is written behind the scenes. I used AI in the production of the UI (which I don't really have shame for) since having made large UIs, it's a tedious sucky process I'd rather avoid and AI actually does it pretty well.

What I am not proud of is that I've been using it to write the SQF. Although it's been effective in completing the tasks I've been giving it, I've had to do a ton of refactoring. Even then there are just fundamental icks with how it made it (excessive uiNamespace variables, gross returns from files which could've been avoided, etc).

Anyway, wanted to be transparent about that, as I'll probably still plan to release it as I do see people finding it super enjoyable and helpful. Just not very happy with myself for the way I did it.

[cosmic lichen]

That left part of the UI could perfectly fit into Display3DENEditAttributes.

Like this for example https://community.bistudio.com/wiki/Spearhead_1944_Single_Player_Mission_Parameters

On another note. Give the colour sliders the actual colour e.g. Red, green, blue

The marker type drop-down should show the marker icon

[old owl]

Ahhh okay I got you now. Sorry I was typing and stopped. I was unfamiliar with that class but just ran a (findDisplay 46) createDisplay "Display3DENEditAttributes"- and along with the wiki you sent, I think I kinda see the vision of what you're going for!

[cosmic lichen]

The drop down list for marker shape can also be triangle etc. Fyi

[old owl]

Also I totally agree with you on it showing the marker icon. Initially I tried to make that a controls group with the actual icons for buttons but had a hell of a time so reverted it back to a combo for the sake of testing.

Not quite sure what I was doing wrong there, but I have a feeling it was with the way I was creating the actual buttons.

Either way totally a great idea I'll plan to give another go before I merge that feature into my main branch 🙂

[fleet sand]

Myb switch to a HTML style for UI then you can create stuff like this much faster IMO.
#arma3_gui message
I beleave dedman has a repo on his github on how to make UI with HTML.

[old owl]

Isn't what you linked not HTML but regular Arma UI just done really well?

The web browser implement rocks, but wouldn't I experience some worse trade-offs in this instance of not being able to interface with the game as much like with map / camera?

[errant jasper]

Presumably you can just draw the map to one side, or draw it on top of the browser.

[coarse needle]

Why does this show Generic error in expression? I'm new to scripting:

publicVariable "RCP_CACHES_PLACED";
if (side Player == west) then {
    deleteMarkerLocal "cache0marker";
    deleteMarkerLocal "cache1marker";
    deleteMarkerLocal "cache2marker";
    deleteMarkerLocal "cache3marker";
    deleteMarkerLocal "cache4marker";
};```

[jade abyss]

Remove "= true"

OR "== true"

[coarse needle]

It doesn't need to check if RCP_CACHES_PLACED returns true?

I'll try that thanks

[jade abyss]

a = true;
waitUntil{ a };
-> Instant go

a = false;
[] spawn {sleep 5; a = true;};
waitUntil{ a };
-> 5s pause -> Then go

I hope you get it 😃

[coarse needle]

so first I make a _cacheplaced = RCP_CACHES_PLACED = true;

then I use that _cacheplaced to check if the condition is true?

Oh no

That won't work

😄

[proven charm]

has it ever happen to anyone that player is not player (isPlayer) in initPlayerServer.sqf?

could be null maybe?

[cosmic lichen]

No

[little eagle]

single = is to set the value of a variable. You can't logically have more than one of those in a single statement.

double == is used to compare two values. It's a command that returns a boolean. You can not compare two booleans though, because they hate us.

[proven charm]

must be those one in a million glitches then, or me just breaking stuff which is more common 😄

does anyone know how to make vehicle stay in formation? im pretty sure this is possible but i forgot how

now the vehicles just takeoff leaving infantry behind

[little eagle]

If RCP_CACHES_PLACED is a boolean, then you can simply write waitUntil {RCP_CACHES_PLACED}, because the return value of the script (what is inside the curl brackets) is expected to be a boolean.

[coarse needle]

oh, ok!

[little eagle]

If RCP_CACHES_PLACED is not guaranteed to be defined you write: waitUntil {missionNamespace getVariable ["RCP_CACHES_PLACED", false]};

Note the quote marks

[round scroll]

does ropeCreate only work for vehicles? I tried to create one on a building (Nimitz), but the ropeCreate does return nothing.

[little eagle]

@coarse needle You should probably change the variable name of "RCP_CACHES_PLACED" to lower or camel case, because all caps is usually used for static values or macros.

I think the first object of ropeCreate has to be a helicopter (including quadrocopter drones), while the second one has to be movable

[still forum]

But adding a ?? would be a language feature, not a command feature.
And if modifying the language already, anything could be done.

It would not be. ?? doesn't take code block, it wouldn't execute it

[coarse needle]

Thank you @little eagle and @jade abyss ! I got it working now!

Been trying to cut down a tree with a hammer here for the past day. Don't have any experience in actually creating code, so I'm just hacking away at code somebody else made and trying to understand how things work.-

[still forum]

Hello. No don't ping

[radiant lark]

Aight aight

Lmk

Not used to people having such blacklogs

[still forum]

use getOrDefaultCall.
That gets rid of your if, then, isNil. That might make up the difference.

[still forum]

I enabled hashed lookups in all config classes with many entries. That would make at least the early lookups faster. But a hashmap should still win.

[still forum]

That is because / is multiple commands.
It needs to decide which to call based on the types you pass to it. Number or config/string.
It basically has a internal loop over all variants, with type checks for each to find the first matching one
I made type checks much faster though

[still forum]

You would make the right half as webbrowser and the left half as normal UI.
But if you already have working thing anyway, not worth the effort

[hallow mortar]

Just to be clear, the ?? idea I was responding to there was "an equivalent to && that can do lazy evaluation without needing {}". My point was that {} isn't needed because someone arbitrarily decided the && command should do that, it's because that's the only way to tell the game not to evaluate that code right now. Changing the way the game processes code so that {} wouldn't be needed is on a fundamentally different level to just adding a new command, and while it may be technically possible, I have to assume it's such a massive logistical issue (not to mention back compat) that it's not worth considering.

[still forum]

{} is because && is a command-level feature, not a language-level feature.

Changing the way the game processes code so that {} wouldn't be needed is on a fundamentally different level
That is the level I was talking about though.
String interpolation and hashmap syntax are also language level features

[hallow mortar]

I can't tell [genuine] whether this is a semantic thing about the difference between the terms "command-level" and "language-level", or whether you're saying it actually wouldn't be that hard to make {} not needed for lazy eval

[errant jasper]

Without language level changes you cannot make {} not needed. Lazy eval is done by inside && command itself. In fact you can argue it is not even lazy, but eager, since I guess the reference to the code object itself is actually resolved even when the code is not run in a && {..}.

I'm guessing even if-then works like that then (pun).

[still forum]

Both?
On language level, making {} not needed is possible, somewhat easily.

We'd need a jump instruction to jump over the code in case it should not be evaluated.
Which I already implemented in SimpleVM. That actually inlines the code, and jumps over it. Getting rid of the "call" between

[old owl]

I was speaking about GenCoders syntax here when he was suggesting lazy eval- #arma3_scripting message looks like I forgot an &

[still forum]

Yeah still, it wouldn't execute the code block if you give it one, that'd be a syntax error because it needs to get a boolean

[old owl]

Ah okay. My thought was a syntax like GenCoder was suggesting would be possible and that the evaluated code would return the bool required. I imagine there's some sort of order of precedence issue that I must not be understanding of why it wouldn't be possible

Ah okay read a bit more- this I am guessing #arma3_scripting message

[errant jasper]

I wouldn't be sad if only && was changed to return right hand side if left side is false or nil (or perhaps even isNull).

E.g.:

nil && 42   -> 42
nil && {42} -> 42
true && {42} -> true
11 && 42     -> 11
nil && nil && 11 -> 11

// if is null-handling also
player && SomeObj    -> player (on non-dedicated), SomeObj (on isDedicated)
player && {SomeObj}  -> player (on non-dedicated), SomeObj (on isDedicated)

Instead of only being bool-compatible.

[old owl]

I think I kinda misunderstood the syntax initially, that's what I thought it'd be and why I said it would be great for select earlier on. Which tbf thinking about more now, at that point why not use param I guess for the example I gave.

[old owl]

I wanna say players logging in as seagulls was a somewhat reoccurring issue we had a long time ago. No idea when during initialization it happened though but from my understanding it was an Arma issue. I think that should long be patched now though as I haven't heard of it happening in years but maybe something like that maybe? Lol.

[proven charm]

only "weird" thing ive encounter is player being null in mission start on client, maybe this was similar

[old owl]

Now I think of it the instance above I mentioned probably wouldn't be a cause since isPlayer probably checks to see if the object is being controlled by a player; not the object they are?

[proven charm]

isPlayer could return false on objNull as well

[tulip ridge]

It was if you enabled the "Stay on position" flag and the fix was disabling it in MP iirc

[round scroll]

thanks, commy2

[wet shadow]

Hey people, is there a scripting command to execute some existing input action? E.g. in this case I'd like to fire script that among other things would use the gunElevAuto input action to automatically lase range to current target.

[proven charm]

try finding out the input script name and call that

maybe there is file you can run or function to call

[digital hollow]

There is setWeaponZeroing but it doesn't work for vehicle turret, even though currentZeroing does =(
For a lot of other actions you can use the action command https://community.bistudio.com/wiki/Arma_3:_Actions

[wet shadow]

Yeah, I've combed through the list and nothing there seems to be about gun elevation/ranging.

Setting in manually via setWeaponZeroing is not really a solution as that command requires you to set the correct range. You'd have to get the correct range from somewhere first. Rather frustrating as in my use case you'd be staring through a range finder. Thanks anyway.

[digital hollow]

I mean getting the range is doable

[wet shadow]

Can you elaborate?

[digital hollow]

You can lineIntersectsSurfaces down the center of the screen to the first thing and then calculate the distance

[wet shadow]

I don't follow. There is a key action gunElevAuto that you can use to detect when player uses that action, but I've yet to find if you can induce that command via scripting.

[wet shadow]

Oh yea true, I've used that elsewhere. But it seems rather inefficient and complex for a script that would only ever be called when using a rangefinder.

[proven charm]

i thought you might find the source code for the script that does what you want

[radiant lark]

Do you guys think the bug of vehicles flying and glitching due to desync and other factors (including, probably, physx) can be fixed? why hasn't it been fixed? I tested some Leopard20 mods but multiplayer still sends vehicles flying

[proven charm]

usually vehicles fly when they are spawned inside other objects such as rocks

[hallow mortar]

It's not one specific bug. There's lots of different factors that can cause physics problems - network lag, server performance, bad frame rate, ...

[granite sky]

It's fairly fundamental to physX. Large overlap => unstable behaviour. A3 running vehicle sim at multiple localities is not really suited to it.

[radiant lark]

What I wonder is, if a vehicle is inside another object / vehicle "colision" zone, isn't it clear that it's really a bug? In which scenario would a vehicle go through another object's collision or a vehicle?

[granite sky]

Oh also Arma's pre-spawn collision checking is absolute garbage.

[radiant lark]

Seems pretty doable to me that the vehicle could be pushed outwards, from an engine perspective, to do the proper calculations. I do not know if that's the most important factor though, I just got told the game thinks vehicle is inside something and it sends it flying

[granite sky]

PhysX only acts sensibly when there's a small overlap.

[proven charm]

some vanilla arma free space finding scripts are not very reliable in my experience. i have written my own for this

[granite sky]

Arma generates large overlaps.

Yeah I wrote an alternative to findEmptyPosition that's much less suicidal, but you don't have great tools in SQF for it.

[radiant lark]

Mane I think it could be fixed idontknow

[hushed turtle]

Me too and it wasn't exactly easy

[granite sky]

Still doesn't handle the case where a building has columns.

[radiant lark]

I'm just talking about driving a vehicle and flying to space due to a collision. I cannot think of ANY possible way that a vehicle should go thru another vehicle or building like it were some noclip

[granite sky]

That one's typically caused by vehicles being simulated at different localities the network prediction being off.

So it's pretty hard to replicate on localhost but very easy with a couple of players in different vehicles.

[radiant lark]

Oh right, server isnt the one doing the thing right?

Damn

[granite sky]

Each player's client is authoritative about where the driver's vehicle really is in Arma. Everywhere else is guesswork.

[digital hollow]

could limit max force generated from collisions? stuff would deform or disintegrate instead

[hushed turtle]

I use objetcs radiuses by default, which is faster, but less accurate. As option you can enable intersecs, which can place vehicle inside of other object radious, but can be much much slower

[radiant lark]

Arma 3 server could just take into account the two vehicles, un-overlap them, and calculate an average of the two "testimonies"

[granite sky]

Best to do both. Radius checks work better for small objects, intersects work better for large ones.

[hushed turtle]

There needs to be a lot of intersects though. Some hut in SOG stands on some little sticks

Hard to hit them

[granite sky]

It only knows they've overlapped because PhysX says so, and PhysX is what generates the crazy velocities.

I don't know if A3's implementation of PhysX would allow them to intercede between the detection & response.

[radiant lark]

I remember seeing in wiki some mention of vehicles being like simple objects

There was smth that said that on collision you had to set teture manually and all of that, idk where but I remember

[granite sky]

yeah that's the column problem.

But you can still do waaaay better than findEmptyPos without it being too expensive.

[radiant lark]

https://github.com/NVIDIA-Omniverse/PhysX it supposedly got open sourced like a year ago!!

[hushed turtle]

Most fun thing about findEmptyPos is, that it can find empty position inside of other object, like in rock lol

[radiant lark]

Oh, I wonder what Arma really feeds it and what not

[granite sky]

Example of findEmptyPosition doing its work:

[hushed turtle]

That's not even enough space

[radiant lark]

seems to be really empty

[granite sky]

I'm fairly sure it's actually bugged :P

Like there's "find an empty position badly" and then there's whatever the fuck findEmptyPosition is doing.

It doesn't seem to be taking too much notice of the bounding box of the object, which is odd given that you pass in the classname.

Like it finds a position that a unit could fit in, regardless of what you pass.

[hushed turtle]

Is there a way to get size of object based on class?

[radiant lark]

I think it does but doesnt take into account gravity and such

[granite sky]

Nah, you have to pre-spawn. The engine could get it if it wanted to.

The code probably has //TODO: get actual bounding radius

[hushed turtle]

Maybe it just uses papercar

[digital hollow]

How do I tell if a connected uav feed is from the driver or the gunner?

[past wagon]

I'm having an issue with my onKeyDown EH returning true to override the default action. All SystemChats are firing correctly, but the command scroll menu still comes up when I press F3. Same with the escape menu.

(findDisplay 46) displayAddEventHandler ["KeyDown", {
    params ["", "_key", "_shift"];

    if (inputAction "ingamePause" > 0 && player getVariable "processing") exitWith {
        player setVariable ["processing", false];
        SystemChat "RETURNING TRUE"; //FIRES
        true
    };

    if (_key == 61) exitWith {
        private _playerItems = missionNamespace getVariable ["playerItems", createHashmap];
        private _uid = getPlayerUID player;
        private _items = _playerItems getOrDefault [_uid, [0, 0, 0, 0, 0], true];

        private _redgullAmount = _items select 0;
        if (_redgullAmount == 0) exitWith {
            SystemChat "RETURNING TRUE"; //FIRES
            true
        };

        _items set [0, _redgullAmount - 1];

        [[-1, 0, 0, 0, 0]] remoteExec ["TRI_fnc_updateItem", player];
        [player, getPlayerUID player, 0, 180] call TRI_fnc_handleBuffs;

        SystemChat "RETURNING TRUE"; //FIRES
        true
    };
}];

[past wagon]

sorry for the late reply. this code isnt working for me either. here is what I tried:

(findDisplay 46) displayAddEventHandler ["KeyDown", {
    params ["", "_key", "_shift"];

    SystemChat "EH FIRED"; //FIRES

    if (_key in (
        actionKeys "CommandingMenu1"
        + actionKeys "CommandingMenu2"
        + actionKeys "CommandingMenu3"
        + actionKeys "CommandingMenu4"
        + actionKeys "CommandingMenu5"
        + actionKeys "CommandingMenu6"
        + actionKeys "CommandingMenu7"
        + actionKeys "CommandingMenu8"
        + actionKeys "CommandingMenu9"
        + actionKeys "NavigateMenu"
    )) exitWith {
        SystemChat "RETURNING TRUE"; //DOES NOT FIRE
        true
    };
    
    false
}];

The first SystemChat is firing correctly, but the boolean seems to be false for keys F3, F4, F5, etc

[hushed turtle]

F keys are for unit selection. They don't do anything to command menu

Maybe besides opening it when first unit get selected

[past wagon]

im talking about this

maybe my controls are weird... I guess I shouldnt be disabling this with the F keys, but instead with the actionKeys

[hushed turtle]

And that gets operated by mumbers. The line of line numbers bellow F keys

[past wagon]

hmm

[old owl]

Might be as @dusk gust said then here #arma3_scripting message

[past wagon]

ah

I guess I'll just do what he said, showCommandingMenu ""

but im still having the same problem with the escape menu. can i prevent it from coming up when a player presses ESC?

    if (inputAction "ingamePause" > 0 && player getVariable "processing") exitWith {
        player setVariable ["processing", false];
        SystemChat "RETURNING TRUE"; //FIRES
        true
    };

[old owl]

You could potentially disableUserInput if they aren't supposed to be doing anything during that period but that's a little dangerous. Could also probably just have something on frame to close the escape display too and that would probably be better, but I don't have a solution at the current moment since on my phone. I know a wiki page with default idds exist though

[past wagon]

ok

thatll work for me, thanks

[little raptor]

not sure but probably:

private _ai = getConnectedUAVUnit player;
private _isGunner = _ai == gunner vehicle _ai;

[digital hollow]

oo fancy 2.08
unfortunately it returns null unless i'm assuming direct control

[sly cape]

Is there even more than just the gunner view?

[digital hollow]

Right, for now I'm assuming if there is no gunner then it must be driver.

[digital hollow]

But if that isn't the desired behaviour I can make a ticket xD

[restive leaf]

Threatening Leopard20 with a good time I see 😛

[little raptor]

what do the other commands return, like UAVControl or getConnectedUAV?

[digital hollow]

I included in the screenshot that the non-unit returns the uav vehicle when just connected and not taking control

[little raptor]

ah right I see now. but what about UAVControl?

[digital hollow]

Can check in a few :)

[little raptor]

nvm I think I got my answer based on this example:

[radiant lark]

Leopard wow you work at arma

Why would you make a mod to fix the flying glitch instead of fixing it from engine itself?

[tulip ridge]

They didn't work at bohemia then

[zealous solstice]

exitWith dont work within Eventhandler, https://community.bistudio.com/wiki/exitWith look at KKs comment from 2016

[past wagon]

oh

kk thx

[jade abyss]

At least, he is trying 😄

[edgy dune]

with missiles is it possible to have a submunition when deployed be locked onto watever the initial parent munition was?

[radiant lark]

Oh aight

Leopard you have a chance to do something great here!!

I used their mod for the flying fix

And you can use a car to “push” other which is great

Really really great

But only works on local

[tender fossil]

Interesting. My friend made a complete, fully working and non-trivial feature with three prompts with Claude Code. He instructed the AI to use only BIKI as the source of information/reference which probably contributed to the successful end result quite a lot. The actual code is a bit verbose and just a tiny bit confusing in some places (to me at least), but it's definitely readable still

[radiant lark]

He could add RAG to Claude

Claude + RAG will read from a documents folder where you can have the whole wiki scraped / a lot of code / your codebase

Claude is too expensive tho, if friend uses a lot of credits monthly, tell him to consider using openlimits.app which is Claude but without ratelimits and unlimited credits

[tender fossil]

The AI gets automatic move order to player's waypoint upon spawn if the feature is turned on via hotkey. If the waypoint is lacking, too close to player (< 25 m) or the feature is turned off, AI will start following player normally

[tender fossil]

NOTE: Arma 2: CO! I'm having issues with canMove command. It doesn't detect disabled tracks on tanks or disabled tyres on wheeled vehicles. The vehicle must be close to being completely destroyed before it returns false. Is there alternative for it or am I doing something wrong?

[little raptor]

sorry what? I don't have such mod

[winter rose]

perhaps a modded vehicle?

[tender fossil]

Nope, fully vanilla. It's weird

Tested with several tracked and wheeled vehicles

[winter rose]

that's… definitely weird. Maybe you can get hit point values then?

[tender fossil]

Tried to find the config references but apparently some of the parts are in Czech only and couldn't find an exhaustive list, hmm

The lovely legacy 😄

[ivory lake]

this might have most: https://community.bistudio.com/wiki/ArmA:_Armed_Assault:_Selection_Translations

[tender fossil]

Thanks! 👍 Just discovered the track selection names, still missing wheel related ones though

[ivory lake]

kolo

[tender fossil]

What about pneu? It seems to equal to tyre

[ivory lake]

at least what ive seen in selection names kolo is more common

[tender fossil]

Alright, I'll test 👍 Thanks!

[tender fossil]

while {alive _tracked && !(isNull _tracked)} do {

  sleep _refreshRate;

  _markerName setMarkerPosLocal (getPos _tracked);

  _tracksDisabled = false;
  _wheelsDisabled = false;

  if (_tracked isKindOf "Tank") then {
    _tracksDisabled = if (((_tracked getHit "pasL") > 0.3) || ((_tracked getHit "pasP") > 0.3)) then {true} else {false};
  };

  if (_tracked isKindOf "Car") then {
    _wheelsDisabled = if (((_tracked getHit "kolo") > 0.3) || ((_tracked getHit "podkoloL") > 0.3) || ((_tracked getHit "podkoloP") > 0.3)) then {true} else {false};
  };

  if ((_wheelsDisabled || _tracksDisabled) && alive _tracked) then { // Line 49
    _markerName setMarkerTextLocal "   X";
  } else {
    _markerName setMarkerTextLocal _markerText;
  };
};
``` Next question: Why on earth I get the following error of undefined variable 😂:
```if ((_wheelsDisabled || _tracksDisabled) && a>
  Error position: <_wheelsDisabled || _tracksDisabled) && a>
  Error Undefined variable in expression: _wheelsdisabled
File mpmissions\__CUR_MP.chernarus\Common\Common_MarkerUpdate.sqf, line 49
Error in expression <.3)) then {true} else {false};
};

I think I figured it out actually: getHit returns Nothing when there's an invalid selection

[hushed turtle]

That could cause if to return nothing, sounds legit

Btw, you don't need if, when you just want to return true or false. You can just use whatever condition returns directly: _bool = condition. There is no need for entireif condition...

[tender fossil]

Yeah, I added them when trying to figure out why the variable was nil. Will do 👍

[radiant lark]

Yesss look https://steamcommunity.com/sharedfiles/filedetails/?id=2191542091

[tender fossil]

Every selection returns nil 😄

diag_log (_tracked getHit "pasL");
diag_log (_tracked getHit "pasP");

diag_log (_tracked getHit "podkoloL");
diag_log (_tracked getHit "podkoloP");
diag_log (_tracked getHit "kolo");
``` prints nothing ![KEKSad](https://cdn.discordapp.com/emojis/837266099321569291.webp?size=128 "KEKSad")

[still forum]

You need to set the areaRadius correctly.
The vehicle type is only used to determine where the vehicle can be. A boat won't be on land, a car won't be on water.
It then builds the AI pathfinding grid, based on the vehicle type.
And uses the pathfinding grid to find squares that could accomodate it

[still forum]

You could grab the UI and read the text in the control, does it not display different text at top for driver vs gunner view?

[still forum]

getHit doesn't return nil, unless _tracked is nil
getHit always returns a number

[tender fossil]

It's Arma 2: CO, apparently getHit can return nil then. According to BIKI returning a number in all cases applies to Arma 3, but not A2

[still forum]

Ah possible

[winter rose]

you can use… the HitPart event handler to get selection names?
https://community.bistudio.com/wiki/Arma_2:_Event_Handlers#HitPart

(or HandleDamage, or Dammaged(sic))

[hushed turtle]

I was about to suggest getAllHitPointsDamage, but that's new in Arma 3 apparently 😄

[tender fossil]

Thanks for the tip! Gotta go now but I'll try it later today when I'm back home 👍

[ivory lake]

other suggestion would be to do a config dump and look at class hitpoints

[digital hollow]

I didn't figure this out cuz I'm terrible with gui. But I did find this

_allUAVcams = allCameras select {_x # 1 # 0 # 0 == "uavpipsingleview"};
//[[1780505: <no shape> Camera,[["uavpipsingleview","Internal","Normal",0]],false]]

Then I can get the camera position and compare to uav memory points positions

[radiant lark]

Mr dedmen now that physx is open source can we try to unbug it?

[granite sky]

Oh right, apparently I didn't read findEmptyPosition's description for four years and it used to say that the first parameter was the minimum search radius.

[still forum]

The PhysX we use has been open source for a very long time.
And I don't understand how that is related to UAV camera..

[radiant lark]

Oh yeah I wanted to reply to the bounding box thing which derived from a PhysX conversation

Because I was wondering whether the vehicle overlap and all that could be fixed on the engine and why it hadn't been done yet

[hallow mortar]

Did you know that you can choose to turn off the ping when you reply to someone

[flat kayak]

Yeah, took a bit to figure out last I was messing with it. Fwiw, in my testing sizeOf typeOf vehicle _unit was equivalent to (0 boundingBoxReal (vehicle _unit)) select 2 so we use the former in our fast travel position search.

[radiant lark]

Yes but it's not muscle memory

[tulip ridge]

Hold shift when you reply and it doesn't ping

[fair drum]

do we have a way to set a created ammo/shell a side? so that AI will react to it? maybe add a separate createAmmo instead of using createVehicle? or maybe an additional arguement to triggerAmmo to determine the side?

[topaz lion]

Is there a script that would force off all lights for an aircraft? There’s a bug with the dev version of simplex that lots of aircraft get their collision lights forced on which obviously isn’t great when a helo or plane are doing a strafing run with blinking red and green lights.

I’ve tired “_vehicle setCollisionLight false;” which has worked for some aircraft but not all of them so was curious if there’s a blanket solution? Or if I need to find out if modded aircraft have extra light settings or something that wouldnt then be covered by setcollisionlight

[hallow mortar]

In theory setShotParents is how you'd do it, but it's got some serious limitations.
see also: https://feedback.bistudio.com/T180763

[indigo coral]

Hi folks. Was wondering if anyone had some suggestions for settings to use with setAperture for a night op? I've got a group of folks I'm running a D-Day parajump op for and the default Arma darkness is too much for them to function in without night vision.

[errant jasper]

"Default arma darkness"???

IME, unless map config broken or something, night visibility is governed by moon phase, from pitch black to NVG nice, but mostly unnecessary unless attacking from wrong direction in hilly terrain.

And varies by cloud cover of course

D-Day was done at full moon

[indigo coral]

Trust me I think it's ridiculous too, it's just a complaint I hear a lot

[fair drum]

most of us probably just change the date time to match a more full moon kind of night

[indigo coral]

Right on. My group was insistent that the aperture should be adjusted so I figured 🤷‍♂️

[fair drum]

unless you are doing some building intersection checks, you're gonna toast everyone's eyes when they enter a lighted building with setAperature

[indigo coral]

That's kinda what I figured

I appreciate the guidance!

[fading gorge]

@lone glade VS Code is absolutely great apart from this one big problem - http://i.imgur.com/CiG8ZCd.png - the highlight on closing brackets is barely visible and i can't figure out how to change it

can import textmate themes and hack seemingly the whole UI of VS Code just using html/css/js but change the highlight of closing brackets? nope

[royal quartz]

https://community.bistudio.com/wiki/enableVehicleSensor

This command is GA LE, being LE... does that mean this needs to be remoteExec 2 for it to take effect for all people?

If its run just on the computer where the vehicle is local could other people still see it as active and be locked by that vehicle? haha

[fair drum]

https://community.bistudio.com/wiki/Multiplayer_Scripting#Locality

do you still think the same thing after re-reading this?

[granite sky]

I don't think it makes sense for it to actually be LE?

Unless I'm misunderstanding the relation between sensors and radar. Which I probably am.

[royal quartz]

I agree, I feel it should be LE GA at best, trying to test in a dedi server with people now

[granite sky]

Maybe sensors are just for UI logic and AI spotting checks and therefore they could be local, while active radar state is separate and global.

[royal quartz]

I think you might be right there. I couldn't test because I couldnt get the other players RWR to pick up I was locking it even fired a missile, I'll test mor when I have a guy free

[fair drum]

So what's the scam on this? Clicking the picture?

[exotic mesa]

If I was to guess the intent is to bait people to going to the website in the first picture then from there either phishing or session token cloning to steal an account

[bleak valley]

hi
just to be sure, the script

[myObject, true] remoteExec ["hideObjectGlobal", 2];

will show myObject ?

my bad, after re-reading it, the "true" at the start means it will hide it

I was confused by the "2"

[hallow mortar]

2 is the locality target. It's part of remoteExec, not hideObjectGlobal. It means the command will be sent only to the machine with that network ID; 2 is always the server. That's correct for use with hideObjectGlobal because it only works when executed on the server.

[proven charm]

and if you run the hideObjectGlobal on server you dont need remoteExec

[bleak valley]

I run it in a trigger

I had bad situations in the past

[split ruin]

any scripted way to add external target camera to object ? 🤓

[faint burrow]

The trigger's "On act." code can also be launched on client and/or server sides.

[split ruin]

what is the "20Rnd_12Gauge_AA40_Pellets_Snake_lxWS" mag round classname ? 🤔

[hushed turtle]

Look for the magazine in config

[warm hedge]

getText (configFile >> "CfgMagazines" >> "20Rnd_12Gauge_AA40_Pellets_Snake_lxWS" >> "ammo")

[split ruin]

@hushed turtle nice idea, found it -> "B_12gauge_Pellets_Cartridge_lxWS"

[split ruin]

I am trying to make "antidrone" rounds for AK-12 but no luck for now, any suggestions?

player addEventHandler ["Fired", {
    params ["_unit", "_weapon", "_muzzle", "_mode", "_ammo", "_magazine", "_projectile", "_gunner"];
    deleteVehicle _projectile;
    private _newProjectile = createVehicle ["B_12gauge_Pellets_Cartridge_lxWS", getPos _projectile, [], 0, "CAN_COLLIDE"];
    _newProjectile setVelocity (velocity _projectile);
    _newProjectile setPos (getPos _projectile);
}];

I just want to shoot AA12 rounds, because I don't want to avoid using a mod that enables two primary weapons
I have the feeling it just adds one pellet lol

[hallow mortar]

Well you're deleting _projectile and then trying to do getPos etc. on it. You might consider doing it in the opposite order.

Also, use getPosASL/setPosASL or getPosATL/setPosATL instead of getPos/setPos. It's faster and more reliable.

[split ruin]

@hallow mortar I tried wihtout deleting it, its the same story, just the old round is flying (or the normal round plus one pellet )

[hallow mortar]

I didn't say "without deleting it", I said delete it after you get its position and velocity instead of before

[split ruin]

I tried

player addEventHandler ["Fired", {
    params ["_unit", "_weapon", "_muzzle", "_mode", "_ammo", "_magazine", "_projectile", "_gunner"];
    private _newProjectile = createVehicle ["B_12gauge_Pellets_Cartridge_lxWS", getPosATL _projectile, [], 0, "CAN_COLLIDE"];
    _newProjectile setVelocity (velocity _projectile);
    _newProjectile setPos (getPosATL _projectile);
    deleteVehicle _projectile;
}];

but no joy, the only thing is it deletes the old projectile, or pellet is too small to see it ...

[hallow mortar]

Use setPosATL instead of setPos

[split ruin]

I swear it worked for the first round then nothing ...

player addEventHandler ["Fired", {
    params ["_unit", "_weapon", "_muzzle", "_mode", "_ammo", "_magazine", "_projectile", "_gunner"];
    private _newProjectile = createVehicle ["B_12gauge_Pellets_Cartridge_lxWS", getPosATL _projectile, [], 0, "CAN_COLLIDE"];
    _newProjectile setVelocity (velocity _projectile);
    _newProjectile setPosATL (getPosATL _projectile);
    deleteVehicle _projectile;
}];

I am shooting blancs ...

😔

[faint burrow]

Adapt this script:

player addEventHandler [
    "Fired",
    {
        params ["_unit", "", "", "", "_ammo"];

        if (!(_ammo isKindOf ["BulletBase", configFile >> "CfgAmmo"])) exitWith { };

        _bullet = nearestObject [_unit, _ammo];

        _bulletPosition = getPosVisual _bullet;
        _bulletVectorDirection = vectorDir _bullet;
        _bulletVectorUp = vectorUp _bullet;
        _bulletVelocity = velocity _bullet;

        deleteVehicle _bullet;

        _rocket = "R_PG7_F" createVehicle _bulletPosition;

        _rocket setVectorDirAndUp [_bulletVectorDirection, _bulletVectorUp];
        _rocket setVelocity _bulletVelocity;
    }
];

[hallow mortar]

It seems to be because of that ammo specifically, I tried with a .50 round and it works fine

Probably because of how it works with submunitions I guess

[hushed turtle]

Maybe it needs triggerAmmo?

[hallow mortar]

Tried that with no effect

[hallow mortar]

If you're up for making a mod at all, you can make a mod that adds shotgun shell magazines to the AK-12 - no need for separate weapons.
If it has to be scripted without mods then it seems like there's going to be an issue.

[split ruin]

@faint burrow holy moly its working ! 🥳 thanks to all

[hallow mortar]

🤔

[split ruin]

here it is the working version if somebody want to have some chance against drones

player addEventHandler 
[ 
 "Fired", 
 { 
  params ["_unit", "_weapon", "_muzzle", "_mode", "_ammo", "_magazine", "_projectile", "_gunner"]; 
 
  if (!(_ammo isKindOf ["BulletBase", configFile >> "CfgAmmo"])) exitWith {}; 
  _bullet = nearestObject [_unit, _ammo]; 
  _bulletPosition = getPosVisual _bullet; 
  _bulletVectorDirection = vectorDir _bullet; 
  _bulletVectorUp = vectorUp _bullet; 
  _bulletVelocity = velocity _bullet; 
  deleteVehicle _bullet; 
  _pellet = "B_12gauge_Pellets_Cartridge_lxWS" createVehicle _bulletPosition; 
  _pellet setVectorDirAndUp [_bulletVectorDirection, _bulletVectorUp]; 
  _pellet setVelocity _bulletVelocity; 
 } 
];

[exotic gyro]

... Just use submunitions?

[hallow mortar]

_bullet = nearestObject [_unit, _ammo]; ```
So why are we doing this instead of using the provided `_projectile` reference?
And what's different about this script that makes _that specific ammo_ work, but not the other way?

[hallow mortar]

Presumably because they are trying to do this without a mod.
*with a mod you wouldn't even have to use submunitions beyond what the shotgun round already does, just let the gun fire the shotgun rounds

[exotic gyro]

that'll do it

[exotic gyro]

I should make a salt shot round

hit = 0.01

[split ruin]

good question 😆

[hallow mortar]

I still want to figure out what the hell is different between these two methods.
With the original method, a normal .50 bullet is created and correctly replaces the original round, but a shotgun shell does nothing.
With Schatten's method, a shotgun shell works. What causes this difference? The difference in positioning logic shouldn't matter; it was working out fine for the .50 round. Is it the other createVehicle syntax?

It's the setVectorDirAndUp

Normally direction doesn't matter for bullets, but for this one it does because of course that's used for submunitions

_bullet = nearestObject [_unit, _ammo]; ```
This part is still pointless though. Just use `_projectile`, it's provided by the EH so it's faster than doing `nearestObject`.

[hushed turtle]

There is a chance you could catch differnet projectile with nearestObject 😆

[split ruin]

yeah, if players fires RPG with this evenhandler on
but this prevent it, right ?

if (!(_ammo isKindOf ["BulletBase", configFile >> "CfgAmmo"])) exitWith {};

but i can easily filter it with "_weapon" param anyway

this will make it work with only one type of magazine

if ((typeOf _weapon != "arifle_AK12_GL_lush_F")) exitWith {};
if ((typeOf _magazine != "30rnd_762x39_AK12_Lush_Mag_Tracer_F")) exitWith {};

etc ...

[hallow mortar]

typeOf is not needed, _weapon and _magazine are already classname strings

[hallow mortar]

That isKindOf is used to check whether the projectile being fired is a bullet. The point of that is to exclude other types of weapons like rocket launchers, hand grenades, etc. If you have a more specific check against the weapon or magazine, you don't need it.
That isKindOf does not prevent against possible misidentification with nearestObject. In theory (though it's a pretty small chance), another projectile of the same type could be passing closer to the player than the fired ammo is, which would cause nearestObject to target that other projectile instead of the one you just fired.

That's part of why you should just use the provided _projectile reference instead of doing the whole nearestObject thing.

[split ruin]

@hallow mortar actually the chance is very big because all team members will use AK-12 and they will be shooting close to each other

player addEventHandler
[
    "Fired",
{
    params ["_unit", "_weapon", "_muzzle", "_mode", "_ammo", "_magazine", "_projectile", "_gunner"];
    if (_magazine != "75rnd_762x39_AK12_Lush_Mag_F") exitWith {}; //makes this mag "antidrone mag" :)
    
    _bulletPosition = getPosVisual _projectile;
    _bulletVectorDirection = vectorDir _projectile;
    _bulletVectorUp = vectorUp _projectile;
    _bulletVelocity = velocity _projectile;
    deleteVehicle _projectile;
    _pellet = "B_12gauge_Pellets_Cartridge_lxWS" createVehicle _bulletPosition;
    _pellet setVectorDirAndUp [_bulletVectorDirection, _bulletVectorUp];
    _pellet setVelocity _bulletVelocity;
}
];

sript is working fine with "_projectile" param as you suggested 👌

[hallow mortar]

You should still have your checks for weapon and magazine, so this EH isn't firing for other weapons they might have. I said typeOf wasn't needed, not that the entire check wasn't needed.

[split ruin]

@hallow mortar just added it 🙂

[hallow mortar]

If you have a specific magazine check, you don't need the isKindOf bulletBase check, because you know that magazine only contains bullets

[warped thicket]

seems like a theme issue

[queen cargo]

vs code is a piece of shit when it comes to defining your own lang
its even more PITA then arma dialogs

[digital hollow]

Is there a way to set the zoom of turrets with continuous zoom optics (as oppose to discrete optics modes) like the rcws ugv?

[old owl]

https://steamcommunity.com/sharedfiles/filedetails/?id=3704064685
https://github.com/milorules1012/SQF-Lab

And for anyone else who will find it useful 🙂

Haven't added a CONTRIBUTING.md but if anyone wants to add anything or whatever, feel free!
-# Very early stages, but I was eager to create a release :P

[fading gorge]

no matter what language or theme, the closing bracket highlight is barely visible on my machine. i know it's off-topic but if anyone knows a solution i'm all ears

[royal quartz]

checked on this further and you are correct. It just disables the component for that person not the vehicle. I dont think there is a way to hard turn off and keep off a radar unless you do that command with remoteExec 0

[royal quartz]

Oh this will be very useful!

[proven charm]

is flyInHeight broken because i can make heli go down to 40m but not below. the wiki says 20m is the min

[cosmic lichen]

Did you try the alt syntax?

[proven charm]

yes

[cosmic lichen]

and getPosATL returns > 20?

[proven charm]

yeah

[cosmic lichen]

Idk then

[zealous heath]

Trying to have a hint display after a countdown, but I am getting an invalid number error. Any help greatly appreciated:

with uiNamespace do {
    [
        [
            ["TESTING...", "align = 'left' shadow = '1' size = '0.7' font='PuristaBold'"]
        ],
        0.256
    ] spawn BIS_fnc_typeText2;

    playSound ["zoomIn", true];
    uiSleep 0.3;
    playSound ["assemble_target", true];
    uiSleep 0.5;
    playSound ["surrender_fall", true];

    [10] call BIS_fnc_countdown;
    call BIS_fnc_showMissionStatus;
    waitUntil {
        (missionNamespace getVariable ["BIS_fnc_countdown_time", 1]) <= 0
    };
    hint "Done";
    sleep 5;
    hintSilent "";
};

[proven charm]

is 0.256 valid there? pls post the error msg

[zealous heath]

0.256 is valid, that number has to do with the scrolling text. The invalid number, it says, is at the hint

tried only having hint "Done";, which works, but the hint won't display after the countdown

Is there maybe a better way to have code be initialized after a countdown?

[hushed turtle]

What is initialisation of code why after countdown?

[faint burrow]

pls post the error msg

[zealous heath]

wierd... error isn't popping up anymore with the exact same code... hint still won't display tho

[zealous heath]

Because I want a hint to be displayed after the countdown ends

[faint burrow]

waitUntil { !([true] call BIS_fnc_countdown) };

[zealous heath]

holy hell, that worked, thank you ^^

[little raptor]

if you have an object above that height they won't go lower iirc

ah wait 20m. right. yeah between 10 and 20m doesn't work. lower than 10 does work. idk why

[proven charm]

just verified that. weird

[proven charm]

flyInHeight command needs fixing scotty

the command it self is buggy

in game

see our convo above with leo

[versed belfry]

Heyo, quick question.

Is sending large hasmaps over network a bad thing?

Context:
I am storing values on the server from each client once per second.
Then I am sending those values to the Zeus client(s)

If there are 30 players, that's 30 messages from the server to a Zeus (because the values are stored on the player object).

So I'm wondering if instead using a hashmap containing player objects and their values being sent once per second to the Zeus is better than 30 messages per second of singular values.

[mortal folio]

the main concern here is how big/many the values are - if they're numbers or small strings it's fine (as long as you're sure theres no other way of doing it, like measuring whatever you're measuring through zeus instead of the server to avoid networking etc)

the hashmap should in theory be less expensive because it's less individual calls

[versed belfry]

The format would be [UID, [data_1, data_2, data_3, data_4]]

Where the player UID is the key, and the data is just numbers iirc

[mortal folio]

I see. Well, i dont know the exact use case but it boils down to only sending them when theyre needed or updated instead of per second, or measuring locally instead of through server (if possible) - other than that if you need the data then you need the data, just gotta find ways to make it optimized, tho honestly given the example its not a big deal

[versed belfry]

The use case is for the FPS monitor mod that's on the workshop, so chances are it will be updated every second 😅

[mortal folio]

Gotcha - youre fine then, far as im aware numbers are cheapest besides bools in network anyway, plus youd be surprised how much data mods send over network that is at larger scale than this and you dont notice a thing

[versed belfry]

Makes sense

Thanks

[mortal folio]

the only thing i'd optimize in your case then is avoid sending if the fps monitor is turned off - and maybe add margin of error (i.e if fps is within 3 frames of the previous networked framerate you discard it, etc)

[versed belfry]

Noted, thanks again.

[errant jasper]

Why bother with a hashmap?
Why not just setVariable ["CalculatedMeanFPS", _fps, true] locally on each player machine and let the zeus clients read it ?

[granite sky]

Because that's worse? :P

[mortal folio]

more command calls and more network queue

[errant jasper]

The intended goal is to debug performance issues for clients? The entire cost is dominated by how data is analyzed and emitted not how often it is sent.

[mortal folio]

well his question was explicitly about network traffic

[granite sky]

Before 2.20 I'd have said the vast majority of the cost there would be the network traffic. Even from a server CPU perspective.

[mortal folio]

the hashmap method would in theory be less network traffic so it's ideal to go for that (i could very well be wrong based on my assumptions but 🤷)

[errant jasper]

How are hashmap's less traffic by nature of being hashmap?`
Surely the benefit here is you only forward from server to zeus clients.

[mortal folio]

and heck even CPU time - calling setVariable once per player is gonna take longer than sending one single variable in mission namespace that contains a hashmap with all players

[granite sky]

It's not. It's because it's N times fewer packets sent.

[errant jasper]

Are you saying arma does not compact messages into fewer packets?

[granite sky]

It does, but it's not very good at doing that very quickly.

Consider that it's also N times as many JIP queue entries to handle.

And that one used to be the limiting factor.

[mortal folio]

tbf ideally you wouldnt use the JIP queue anyway and should just send directly to the zeus client ID instead

[granite sky]

setVariable true is JIP

[errant jasper]

^ yeah true.

[granite sky]

With the bulk array/hashmap then you could skip JIP entirely (and you should).

[errant jasper]

True JIP affect was stupid of me. But JIP issue is completely orthogonal here and has nothing do with hashmap.

[granite sky]

Hashmap or array is irrelevant, but there isn't any other way to send bulk data that I can think of.

Oh, I guess you can make it into a string.

Might actually be cheaper but whatever :P

[errant jasper]

Anyway, without knowing the intended usage of the data, I still suspect it does no way require 1s resolution.
Would be nice if we had way to send unreliable communication in Arma.

[mortal folio]

this existed https://community.bistudio.com/wiki/sendUDPMessage

but no worky

[errant jasper]

That never "existed" in customer version as the note says.
And of course requires proper network setup for both, but especially receiver, port forwarding etc.

[mortal folio]

yeah, but i imagine it means the game at least has the sqf integration for it and whatever underlying system is needed

or at least parts of it 🤷

[mortal folio]

good thing extensions exist, just make your own

[errant jasper]

I am not going to ask people to run an extension unless it is truly necessary for something serious.

Already surprised data extraction/ransom attacks haven't happened already in certain arma communities

[mortal folio]

well the extension page does make it very clear on multiple occassions that they're dangerous - plus your game wont even launch it if you have battleeye and it hasn't been whitelisted, so i'd say there's enough measures in place already to prevent that

[errant jasper]

That is not the level users operate at. Some server runners use extension, people complain they can't join, they tell people disable BattlEye and join. The typical user does not even know they are running "arbitratry code".

[mortal folio]

fair

[fair drum]

I never really messed with hashmaps over network because it sends all of it. Say I have a 100 element hashmap, or in my main use, hashmap objects, and I only want to change a couple of things at various times. Do we have any metrics on that kind of scenario?

[mortal folio]

maybe there should be a warning if you start the game with battleye disabled and mods enabled, something along the lines of "If you've been asked to disable battleye for a mod, beware that this might be used for malicious purposes"...

a popup with a dont ask me again but then at least the users would be warned

[versed belfry]

The goal is for a mod that displays the client's FPS live to the Zeus when he's Zeusing. ALlowing you to notice if your player's FPS is dropping in anyway.

[errant jasper]

So if there is a heavy 15s drop, but no zeus was around to see, did it matter or not ?

[mortal folio]

i'd imagine it's intended to be used with zeus only lol

[errant jasper]

If there is a 2s heavy drop, the zeus saw, but it went away did it matter.

[mortal folio]

there's always a zeus in some groups

[versed belfry]

Aren't strings the WORST thing to send over network? >_>

[granite sky]

Nah, depends on the data.

[errant jasper]

I mean don't get me wrong. Your Arma will probably work fine almost all the time without issue if 100 players, send each second to 4 zeus.
But that one guy with bad connection is gonna slow down other traffic(message processing) when his "reliable" messages drop and must be resent

[versed belfry]

From my experience Zeusing it does matter. It specially helps if the drop remains until you start cleaning or handling it.

[mortal folio]

tbf that's a risk that exists even without that mod, setvariables over network are everywhere in arma, but also are we sure arma doesn't work around that? i've never had a slow-network client impact other clients

[versed belfry]

I mean in all honesty? The original mod who I've heard people complain about its code never caused issues for me

[mortal folio]

the original mod just used lots of lazy scheduling iirc, also didnt display the fps smoothly

[versed belfry]

and for ~30 players it sends near 1000 network messages between them overall, (each player receives ~30, and sends ~30.

[errant jasper]

Well can only speak to my own experience. But my traffic 98% originate because some events that must be synchronized (reliably acknowledged) anyway. I can't of the top of my mind think of a case where I wanted regular interval-based time traffic.

[versed belfry]

Yea but I've had people complain about the number of messages sent, mainly on the theory it is bad, but in practice I've used the mod without any noticeable impact with heavy performance scenarios.

None the less I'm trying to spend the time to improve it where I can

Which is why my original question came up

Does anyone here have a resource about network trafic in Arma 3? Or is it a black box?

[mortal folio]

mm... if you use tons of other mods though and not just your own, you can be sure plenty do a setvariable frequently

[errant jasper]

Yeah, and they need a really good reason or they are bad...

Sturgeon's law et cetera

[mortal folio]

true

[granite sky]

That's the thing. You can get away with one mod doing lazy shit with one variable, but it adds up.

[errant jasper]

I am impressed with the things people "get away" with network wise in Arma 3, but that might be shaded by my old Arma 2 experience of how wrong it can go.

[mortal folio]

well also people today have generally good enough internet to brute force through it

which isnt ideal, but its true 😄

[granite sky]

Acre sending 16MB lumps on init was pretty funny

[versed belfry]

Another kind of related question, is there any benefit of just sending an array of arrays instead of a hashmap?

[errant jasper]

I think what you find on the BIKI is the most documented. From Arma perspective all you really know that it is: Reliable ordered traffic: (so on dropped packets the sender must resend, and receiver does not process certain remote execs, publicvariables, certain events until the missing dropped ones have been properly received) .

[versed belfry]

Any specific BIKI entries?

[errant jasper]

Not sure in your array of arrays example. But by sending only an array with specific position containing certain values, you save on sending the "keys" part.

No, most is covered here and in the specific commands https://community.bistudio.com/wiki/Multiplayer_Scripting

[versed belfry]

Well the arrays would contain the "keys" in any way since the keys are player UIDs

[versed belfry]

Chheers

[versed belfry]

Unless hashmaps have a higher cost then I'd just go with them, since look-up is faster on the Zeus side iirc

[errant jasper]

Unless you have a lot of constant keys I don't think the difference will matter. Since you pretty much only has one I doubt you measure any difference.

[granite sky]

Best you could maybe do is maintain an ordered list of players separately, because that would change much less frequently than the frame rates.

probably not worth it though, assuming that you're only sending to zeuses.

[versed belfry]

Yea, chances are hashmaps will be just fine, hopefully

[granite sky]

It's a lot of extra complexity for not a lot of saving.

[errant jasper]

I would just do more collation on client side. Like gather 10s of data into min fps, max, median, average, 95 percentike, 25% percentile and send only every 10s instead.

And zeus clients could keep the data, so visualize data for a player over longer periods.

Though you can do that with per-second fps too of course.

[versed belfry]

Then it wouldn't be accurate / you'd assume whatever you're seeing is 10 seconds old on the Zeus side.

What I can also do if I got the time is make it so instead of just a hardcoded every 1 second, make it a CBA setting, that each server can set.

[errant jasper]

I guess I just don't understand the problem to be solved is?

So Zeus spawns something, and monitors player whether FPS drops?
Suppose it do for 3s but then get back up what now ?

[mortal folio]

having used an fps monitor before, partly yes it's about knowing whether the thing or things you just spawned are causing issues, but also generally just gauging performance - if your players are at 20 fps and nobody is saying anything, but they're stil suffering, you'll be more inclined to start deleting things or disabling stuff etc

[versed belfry]

The mod it self is already on the workshop, I am more or less trying to optimize it.

How you use the data of a player's FPS dropping is up to you.

Personally I use it to tell if only a certain player is having FPS issues, or if everyone suffered from a certain spawn, or certain event, or effect happening.

[old owl]

This was my understanding as well.

[errant jasper]

Sure, but how is it a problem with 10s old data?

[mortal folio]

in theory it isnt, 10s should be fine, but you can apply that same logic to a lot of things - 1s is already optimized enough, you're sending like... a couple hundred bytes if you're at like 20-30 players per second?

[errant jasper]

Not about bandwidth, but packet reliability

[versed belfry]

The issue with sending a packet every 10 seconds of the last 10 seconds of data is that I need to rewrite that part of code to do it

[mortal folio]

i guess... but even then, it's once per second, lots of mods already do a lot of traffic on a sub-1-second interval and there are no issues 🤷 arma seems to gracefully handle it

[versed belfry]

Put plainly, I am only willing to put so much time and effort to solve a problem that in my experience does not exist.

I am more or less doing it so others stop complaining about it because the code looks bad.

Like Kharos said, chances are other mods do A LOT more than this mod does, but ey what can you do,.

I can fix and improve systems. I can't fix people.

[mortal folio]

that's why i recommended something like an fps error margin (<5 fps difference since last networked message is ignored), and not sending data if zeus doesnt have the monitor active, etc

[versed belfry]

Yep that is actually a real neat idea

[versed belfry]

The base for that already exists, right now it does not send it if the FPS is the same, but let's be honest, how often is FPS the same over an entire second?

So adding a margin of 3 or 5 or so one would be a great way to handle it.

[mortal folio]

this is also a "do as i say, not as i do" since for my fps monitor i didnt bother with that 😂

[versed belfry]

Could even make it a CBA setting as well, so that people can configure the mod to their liking

[mortal folio]

i mean fps is average, if you're frequently jumping 5-10 frames average delta per second then you've got a problem 😄

[errant jasper]

Glance more or less at the ground/sky ?

[versed belfry]

Welp, thank you all for the help so far

[mortal folio]

true, tho wouldnt happen often - granted this solution would maybe reduce the traffic by 10-20% becuase.. yeah, as you said, looking in different directions, but we are trying to optimize 😄

[versed belfry]

This cleared up a lot, and helped solidify my understandings a bit more so I can do it.

[old owl]

We've done a lot of network optimizations over the years, the largest thing that was hurting us was JIP queue. Although we also had tens of thousands of messages.

If you're broadcasting data on a regular interval, and you're only broadcasting it to server; I don't think you'll really have any network issues. Packet loss would be more your concern as Muzzle Flash mentioned but obviously there's ways you can case for that.

My advice would be not to set the data public though and instead just send to server.

[mortal folio]

yeah i think even the original mod is essentially just sending to server, which then sends only to zeus

nothing is being public-ed to all clients

granted you could optimize further by sending directly from each client to zeus (yes the game does send traffic through server anyway but it's technically one less step) - just public the current zeus' client ID and have the sending happen that way

[versed belfry]

Well the current version I am working on is setting public

[mortal folio]

Oh no

[versed belfry]

So it's each client updating the variables for EVERY OTHER client

[errant jasper]

Ignoring the collection details and visualization, this is basically how I would do it:

// Player side Send every X second (_toTime - _fromTime) basically.
private _data = [player, [_avgFps, _maxFps, _minFps, _medianFps, _fromTime, _toTime]];
_data remoteExecCall ["FPS_Monitor_Ingest", G_ZeusClients];

// Zeus side: Receive (FPS_Monitor_Ingest):
params ["_owner", "_data"];
_owner setVariable ["FPS_Monitor_Latest", _data];
// Add to like 30 minute history buffer that Zeus could also access. Maybe a graph in GUI?
[_owner, _data] call FPS_Monitor_AddData;

[mortal folio]

Yeah focus on fixing that incredibly wasteful

[versed belfry]

Huh having the data to be accessed later would be interesting

That'd be a down the line thing though but it's a real nice idea to visualize things as well

[mortal folio]

Requires working with arma gui

I have nightmares

[errant jasper]

That was just optional. The data over time part is more important. Then you could even compare if a drop short while ago happened for all or just some.

[mortal folio]

Jokes aside yeah its good MP diagnostics to measure through a mission lifetime, never thought of that

[old owl]

I think also one thing that wouldn't hurt if not discussed already is only having it listen if a Zeuse client is actually awaiting results. No point in sending updates unless the Zeuse client toggles the feature maybe

Also having control over which clients the Zeuse client wants to monitor, then it doesn't have to inherently be everyone if diagnostics is only needed for one person

Kinda just thinking about if I was going to do it, ultimately constraints of how you wanna make the project is up to you; but FWIW :)

[errant jasper]

If still want history, one could do hybrid by coalescing, say 10 updates, when not actively monitored to ensure history is transmitted around (so it would not have to be done in one large bulk later), but upon active monitoring set the update rate back to the default.

[versed belfry]

Hence the "down the line" a.k.a.when I got the will and energy xD

[versed belfry]

So my plan is add the Zeus UID to a list on the server whenever the Zeus interface is open

and remove him whenever it is closed

Technically doable so no updates are being sent unless also the UI for the mod is activated

But that's a can of worms I don't wanna get into sincce it is not my mod

[digital hollow]

That's true. Validate the fps by camera angle and face-to-wall-ness, maybe by distance to nearest thing in center of screen from cursorObjectParams

[versed belfry]

At that point I wonder if you're using more resources than you're saving xD

[digital hollow]

No network* resources though ;)

[mortal folio]

Issue with that is most objects arent occluders so fps isnt affected by looking at a wall

[versed belfry]

Valid xD

[hallow mortar]

setName syntax 2 allows you to set a first and last name for a unit, as well as the "all-in-one" name set by syntax 1.
What command allows you to get the first and last name?

[restive leaf]

BIS_fnc_getName?

[hallow mortar]

That just does name internally

[restive leaf]

Hmm, name does return the name, like Nicolai Dvostok or something though. So that, split returned string for first and last. I use a roaming AI script that does a name on the leader of an AI group and it sends a radio message with that AI's name...

[hallow mortar]

name returns the "all-in-one" name, which for default names probably includes both first and last names. But if you've done setName syntax 2 with some different name, it does not return the separate first and last names.

[restive leaf]

Oh, Sounds like a job for feedback tracker...

[hallow mortar]

Source: did _unit setName ["test", "aaa", "bbb"] and name _unit returned "test"

[restive leaf]

Is that because it would be expecting sqf player setName ["Ben Kerry","Ben","Kerry"];like in the third example and yours should be sqf player setName ["aaa bbb","aaa","bbb"];?

Though I have no idea what the addition of the last two parameters is used by...🤷

[hallow mortar]

That seems like it would be a little bit pointless

[restive leaf]

Thus my last comment 🤣 Maybe it was used... but doesn't really make sense since that syntax was added in ARMA 3 1.02?

[hallow mortar]

The game has some kind of capacity to distinguish because it makes the distinction in e.g the commanding HUD

[restive leaf]

Maybe the last name to clearly identify it for this: Only last name will appear in command bar i.e. this setName _myNameArray will display _myNameArray select 2. If setName is used with a string e.g this setName "blah", nothing occurs in the command bar and the default randomized name is displayed? But just guessing

Like if you had a character named John W Ick?

[hallow mortar]

It could be more useful if there was a way to actually get it :U

[restive leaf]

Hopefully Dedmen will pop in...

[real tartan]

is there a way to limit max free camera distance from player for BIS_fnc_EGSpectator (vanilla spectator)

*zoom

[hushed turtle]

I don't think that answers the question.

[mortal folio]

He was looking for max free camera distance, not being locked into one view

Yeah but its not free camera anymore

Anyway the way to do it would be to take the spectator camera object, define max distance, and then perframe compare the distance from player to that camera object and when its exceeded you clamp the camera position using the _cameraPosition vectorFromTo _playerPosition to get the "backward" direction, vector multiplied by absolute of _currentDistance - _maxDistance, and set the new camera position based on that returned vector being vectorAdded to current position

(pseudo-mathing this, busy currently but that's the idea i can think of)

that'd be in the free camera

i guess? i dont use spectator often, all i know is he wants to limit how far the camera can go

all you need for that is to have the camera object and the anchor object

[hallow mortar]

Why do you think this isn't about EGSpectator?

[hallow mortar]

Free camera just means a camera that can be moved freely rather than being attached to something. Zeus has a free camera, the old spectator has a free camera, and the Endgame spectator system (EGSpectator) has a free camera.

[mortal folio]

is there any way to get whether the player is aiming down sights or not?

[faint burrow]

https://community.bistudio.com/wiki/getOpticsMode ?

[mortal folio]

hm.... can probably make that work, thanks

(thats if it remains empty if not aiming down sights 🤷)

[hallow mortar]

cameraView will return "GUNNER" when ADS

[mortal folio]

Oh, great!

[jade acorn]

i'm trying to figure out a checker to detect if player is moving. animationState works for mostly everything but I can't find anything that would let me check if the player is "freelooking" around. Any ideas?

inputMouse is only for pressing buttons, I'd need something to detect mouse movement I guess

[digital hollow]

https://community.bistudio.com/wiki/freeLook

[split ruin]

I removed the annoying marker dot from AH-9 with

heli setObjectTextureGlobal [1,""];

but I wonder can I load some other targeting dot? are there any texture like this in the game already - dot, crosshair, etc ?
Update: I ended making one pixel red dot as reticle, it does the job but it could be much better looking ...

[mortal folio]

inputAction "AimLeft" or Right or Up or Down

Gives you a number, based on what exact metric im not sure, but not 0 will be movement input

[jade acorn]

why did google not show me this 💀

[prisma kraken]

Is there anywhere to find how the vanilla Arma 3 rearm functions behave in terms of scripting it?

Trying to make it so AI vehicles can rearm like players can when near an ammo sources through some scripts or a mod.

[tulip ridge]

https://community.bistudio.com/wiki/Arma_3:_Actions#Rearm
Repair / refuel are just below that as well

[prisma kraken]

thank you

[stark spade]

is that only me who is facing this problem while searching on forum ? or search feature is turned off by website itself

[restive leaf]

It's pretty well indexed by google at this point, could do a site specific search for what you want there

[versed belfry]

Hello :D

Quick question, what's the easiest to detect when Zeus was opened / entered into Zeus mod?

[versed belfry]

Uhhhh if it works do pass it please

I am mainly looking for something like an eventHandler rather than a constant check with a waitUntil

[hallow mortar]

You could use a user action event handler to detect when the Zeus button is pressed, then check for the Zeus UI to figure out whether it was opened or closed

[versed belfry]

How would I got about doing that? Also would that account for say a user being forced into Zeus UI?

The only reliable way I've found so far is using a waitUntil but I'm not fan of a constant check that'd run until a Zeus interface is open

[hallow mortar]

It would not account for being forced.
How you would [begin to] do it is https://community.bistudio.com/wiki/addUserActionEventHandler

Unfortunately BI doesn't really offer a lot of EHs for Zeus

There is the curatorUnitAssigned scripted EH (https://community.bistudio.com/wiki/Arma_3:_Scripted_Event_Handlers#Module_Game_Master_(Zeus)) which might be helpful

In theory you could insert an EH into the Zeus interface UI somewhere. Might need a mod for that though, not sure if mission description.ext UI stuff can modify existing UI like that

I'm a little surprised CBA doesn't seem to offer any Zeus UI events

[versed belfry]

I'm doing it for a mod anyways so technically possible.

Honestly the waitUntil and findDisplay combo might be the most reliable here whether I like it or not

[tulip ridge]

It does, but only through a displayLoad eh
It has some "feature camera" stuff to catch special uis like arsenal, zeus, eden, splendid camera, etc. but needs some extra logic since if you go from not in zeus -> editing a unit you get three values. "" -> "zeus" (might be curator) -> "arsenal". Then exiting arsenal gives "zeus"

What I've done is hook into it and then just check if zeus display exists

[versed belfry]

I assume you mean CBA?

[tulip ridge]

Yes

["featureCamera", {
    private _inZeus = !isNull findDisplay 312;
    // ...
}] call CBA_fnc_addPlayerEventHandler;

[versed belfry]

Oh nice, thanks.

["featureCamera", {
    private _inZeus = !isNull findDisplay 312;
    if (_inZeus) then {systemChat "something happened!"};
    if (!_inZeus) then {systemChat "something ELSE happened!"};
    }] call CBA_fnc_addPlayerEventHandler;

Just a quick example, I think this works perfectly

[tulip ridge]

Can just use if/else but yeah

[versed belfry]

Yep yep

Lack of sleep moment

Thanks again!

Yea this should work for now :D

Thanks though

[broken pivot]

Yoo, what a nice day to ask a question :D
Im trying to paint my systemChat that I load in with localize "STR_STRINGTABLE" inside of an addAction scope...
I feel like beeing close to a solution but also dont know if its really possible

Thats my execute:

    _SPrice_Normal = 10;
    _APrice_Normal = 10;








//StandartFleck
_mausradBundeswehrStandartFleck addAction 
    [
        format [localize "STR_MR_BW_TRUPPFUEHRER_FLECK", _SPrice_Normal, _APrice_Normal],
        {
            execVM "EBER\MausradLoadouts\Bundeswehr\StandartFleck\Truppfuehrer.sqf";
            if ((!isNil "KP_liberation_supplies") && (!isNil "KP_liberation_ammo")
            ) then
            {
                _Price_Text = format [localize "STR_MR_COST_INFO", (_this select 3 select 0), (_this select 3 select 1)];
                [_Price_Text] remoteExec ["systemChat", _this select 1];
                KP_liberation_supplies = KP_liberation_supplies - (_this select 3 select 0);
                publicVariable "KP_liberation_supplies";
                KP_liberation_ammo = KP_liberation_ammo - (_this select 3 select 1);
                publicVariable "KP_liberation_ammo";
            };
            
        },
        [_SPrice_Normal, _APrice_Normal],
        1.5,
        true,
        true,
        "",
        "true",
        3,
        false,
        "",
        ""
    ];

Thats my stringtable.xml snippit

        <Key ID="STR_MR_COST_INFO">
            <Original>Your loadout outlay &lt;t size='0.8' color='#BD0000'&gt; Supplies: %1 &lt;t size='0.8' color='#00BD06'&gt; Ammunition%2</Original>
            <German>Deine Ausrüstung kostete dich &lt;t size='0.8' color='#BD0000'&gt; Nachschub: %1 &lt;t size='0.8' color='#00BD06'&gt; Munition%2</German>
        </Key>

The problem appears in the chat. The message fires a little to direct and doesent get structured.
Does anyone know how to do it, because I could bet its possible..
I read about structuredText but doesent know if the enviroment fits

[old owl]

I wanna say it may be because of your remoteExec usage with it although I don't remember why so maybe I'm mistaken

[broken pivot]

Yeah Im not sure if thee KP supplies are lowered right, Im writing a function for that to get rid of that

[faint burrow]

[_Price_Text, { systemChat (parseText _this); }] remoteExec ["call", _this select 1];

[broken pivot]

ouuu
https://community.bistudio.com/wiki/parseText is my solution.
I just didnt cared anymore because other things are displayed the same way

**But this is good to know. I really try to keep it in mind 😄 **

[wet shadow]

So uh, anybody know whats up with SQF in -1 * 0 returning -0. Rather bizarre to have that popping up in the UI for a return.😵‍💫

[faint burrow]

abs -1 * 0 😎

[wet shadow]

Well yeah, but annoying to have to change the formula to account for it since I want negatives otherwise. What a weird notation.

[winter rose]

isn't -0 == 0 ?

[wet shadow]

Yeah, checks true. This is just an UI annoyance for me.

[errant jasper]

It is not an SQF specific thing. That is just how floating point numbers are in the almost universal spec IEEE 754, so you get the same in Python, JS, et cetera.

[charred monolith]

Hi ! I've been looking at playSound3D to play a voice line on a AI, but the weird thing is when I set the distance I need to set it to like 100 to be able to hear it. If I set it to 10 or 5, I don't here the audio file. Even if I breath down the neck of the AI.

[hallow mortar]

The distance is the absolute maximum range. The volume will fade out until it reaches zero at that distance, and because of the shape of the volume:distance curve, it can be pretty quiet in the outer half of the radius.

Consider how quiet someone would have to be talking in order to be completely inaudible past 10 metres.

You can try increasing the volume of the sound, either with the parameter in playSound3D or in the file itself, but 5-10 metres is still going to be a very close limit. It's not very far in real life and it's less than you think in Arma.

[charred monolith]

Ok thank you very much ^^

[grand moss]

Whats the script for importing an object?

[sly cape]

What do you mean by importing an object?

[grand moss]

Trying to get an object from Blender to arma 3

[sly cape]

Ask in #arma3_model.

[tulip ridge]

There isn't one

[crimson lion]

It's a whole process. In a nutshell, you have to export your model to a .p3d with LODs (Arma Toolbox for Blender is good for this), then make a config.cpp for the mod, and then pack that mod into a .pbo to be loaded into arma.

[tulip ridge]

There's Clock's Tools to export as a p3d, but there's no script to export it

[grand moss]

I've gotten to the export part but don't know how to import

[crimson lion]

You're gonna need to make it into a mod in order to be able to see/use it in-game basically. Now that you have your .p3d model file, you need to make a config.cpp and define that model in a way Arma 3 knows what to do with it.

[grand moss]

How would i do that?

[mortal folio]

There are youtube tutorials, google for arma 3 weapon import tutorial and youll find a few

Asking for each step here would take ages

Also wrong channel

[cobalt path]

Not sure whwre to ask this, but its kinda scriot related.
Advanced Repelling mod gets kinda broken by WMO (walkable moving objects) mod. Because I assume WMO recognises ropes as objects u can walk on.
WMO has CBA option to blacklist some objects from being walkable.

Does anyone know "rope" objects item class names for this purpose?

[mortal folio]

check if IsKindOf "Rope"

it's literally just Rope 😄

that's the base class that ideally the rope used by that mod should be inheriting from

[twilit scarab]

When I try to use the command to force vehicles to stay on road they get really jittery is that normal?

[faint burrow]

https://forums.bohemia.net/forums/topic/238654-forcefollowroad-script-makes-ai-move-weird/

[restive leaf]

Anyone know offhand of a reliable and performative method to get the class of house that a player/ai is located in? Would it just be an insideBuilding and lineIntersects combo?

[real tartan]

typeOf nearestBuilding player

[restive leaf]

That can be unreliable in large buildings, since the position calculation is based on the building model's centre...

[real tartan]

alternative getEnvSoundController, look for "houses"

[restive leaf]

Though for my particular usage case (cooking in houses at stoves and fireplaces generating smoke from the chimney) it might be okay... (only problem could be barracks next to sheds)

(Also wish all Enoch houses had the fireplace output memory point rather than just some of them, sad sigh)

[mortal folio]

Thats how id do it tbh, reliable, as long as you use insidebuilding to avoid unnecessary intersects

Just intersect from aglToAsl unitAimPosition downwards to the posASL - 0.1 on Z

[real tartan]

nearestObjects also works, ordered by proximity

[restive leaf]

Yeah, just need to ignore the occassional rug spawn 😉

[mortal folio]

Not for large buildings it wont, could give you the neighbouring building thats technically closer

[granite sky]

If you're looking for "completely reliable" then I'd say no.

With a few rays you could probably get 99.9999% though :P

[mortal folio]

there's also the boundingbox method that would require vector math and may or may not be cheaper/more expensive than line intersects (never tested myself) - just check if bounds of unit intersect/are inside the bounds of the building using vector math magic 😄

but also unreliable with unevenly shaped buildings

[granite sky]

Bounding box extends outside the building so I'm not sure how useful it is.

[mortal folio]

yeah

unless

you do bounding box checks against each individual component (geo or fire geo) in which case you'd have more reliablity, but im 90% sure that'd be more expensive than a single line intersect

unless you do it in C as part of an extension or something

scope creep

[granite sky]

Other issue with ground raycasts is that sometimes the terrain sticks through the floor.

[mortal folio]

eh terrain returns objNull so you can just filter objNulls from the intersection and extend it downwards a bit for safety

since im pretty sure the line will still go under terrain if you ask it to

[granite sky]

ah, if you let it do multiple hits, yeah

[restive leaf]

Yeah, looks like a downward lineIntersects is probably gonna be the best (avoids most furniture). The position of the stoves/fireplaces should avoid terrain intersect problems. But also will modify per last couple of messages. Ah, the joys of implementing a proper cooking system with visual clues for other players 😉

[mortal folio]

man i really wish we had an optimized way of detecting collision or "nearest collidable face"

plenty of ways around it ofc as we just discussed, still 😂

[restive leaf]

Would be nice if insideBuilding had an alternative syntax that returned the building type

[mortal folio]

that'd help here yeah, eliminates the need for any other checks

[errant jasper]

If the building must have a chimney do the intersection towards the chimney positions

[real tartan]

can somebody share a function, that will find random empty position for spawning unit ( soldier or vehicle ) that I can use multiple times for n-numbers of units to spawn, and avoiding spawning units in rocks or positions already with spawned units ( e.g. not spawning vehicles on vehicles )
findEmptyPosition, BIS_fnc_findSafePos

[errant jasper]

Do you want to spawn them together, or spread out?

Easier if together, just find one larger area, and manually spawn inside.

[real tartan]

spread out, my usecase is to spawn multiple units/groups in marker, goal is to spread them out, avoiding spawning them on one place, or spawning vehicles on vehicles

or spawning vehicles on already spawned units

[errant jasper]

In that case, I don't think there is much better solution than findEmptyPosition with extra check after to see if actually clear.

You could try selectBestPlaces, but I don't recall the filtering being suitable for this.

[restive leaf]

Most of the chimneys on Enoch buildings are not placed above the fireplace 😉 Some weird smoke physics must exist in the ARMAverse

[digital hollow]

I'm a big fan of ray along camera direction, at least for logic/mechanics

[restive leaf]

Downward directly works best for my use case, because the only furniture that would intersect would be a rug, which I can ignore. Any other direction could be a variety of furniture

[digital hollow]

Yeah true. I'm thinking like "look at stove and action to start cooking" >
"That's not a stove you donut" or
"Now cooking at this stove"

[restive leaf]

Basically that's sort of how it's gonna work. Wood stoves, Land_dkamna_uhlis are placed automagically as appropriate and I have another small "wood pile" object (not the vanilla one) in fireplaces... For the smoke I use the fireplace output memory point if it exists in the model, and a config that has the offset for the smoke point per building model if not