[_unit, "TEST", "TEST1", 0,_playAction,time+300]
#arma3_scripting
tough abyss
@queen cargo do i need to consider anything when playing on BE servers after install? like unhook something?
@little eagle ๐ one sec
queen cargo
nope
only do not start ArmA.Studio in MP and you should be fine (nobody is sure if BE already added the DLL to the explicit blacklist for auto whitelisting as it is/was whitelisted)
tough abyss
ok thanks!
could you remove MP functionallity from the exe to protect users from accidently using it in MP?
queen cargo
there is no "functionallity" in there
it is the debugger DLL actually that gets auto-loaded by arma
BE should reject it
you wont get banned though
as the DLL wont do anything on its own
and should not be loaded by BE anymore
tough abyss
@little eagle same error as above
little eagle
copy paste it?
tough abyss
20:21:31 Error in expression <[B Alpha 1-1:1 (Lappihuan),"TEST","TEST1",0>
20:21:31 Error position: <Alpha 1-1:1 (Lappihuan),"TEST","TEST1",0>
20:21:31 Error Missing ]
So it's not the same error
It's the same line, which means at least we found the faulty line.
tough abyss
well yea the TEST thing is new ๐
little eagle
Can you copy paste WW_fnc_isPlayer too?
tough abyss
private "_unit";
_unit = _this getVariable ["bis_fnc_moduleremotecontrol_owner", _this];
(isPlayer _unit);```i'm gonna look at missionEventHandler stuff and see if that works
little eagle
That one won't let you pass arguments
tough abyss
fek
little eagle
["blah", onEachFrame", {}, [player, "TEST"]] call BIS_fnc_addStackedEventHandler;
potential repro for the error?
Since you're using a hard coded ID ("WW_pfh_boltAction") for your EachFrame handler, you might as well use a global variable to pass the arguments to the addMissionEventHandler one
So not really a problem
tough abyss
20:29:34 Error in expression <["blah", onEachFrame", {}, [player, "TEST"]] call BIS_fnc_addStackedEventHand>
20:29:34 Error position: <TEST"]] call BIS_fnc_addStackedEventHand>
20:29:34 Error Missing ]
something seems to be weird with pasing arguments into BIS_fnc_addStackedEventHandler
little eagle
Yeah, no error without the typo
tough abyss
oh ๐
tough abyss
i do get the error without the typo btw
are you on devbranch aswell?
20:38:12 Error in expression <[B Alpha 1-1:1 (Lappihuan),"TEST"] call {}>
20:38:12 Error position: <Alpha 1-1:1 (Lappihuan),"TEST"] call {}>
20:38:12 Error Missing ]
No, stable
tough abyss
The data of the whole game were repacked. Please report any related issues (missing textures and icons, pop-up errors, etc.) to the Feedback Tracker. Thanks!
little eagle
But that looks like what I was looking for. Basically the same pattern.
tough abyss
that was in the last devbranch update
little eagle
Well there you go.
Just to make sure
["blah", "onEachFrame", {systemChat str _this}, [player, "TEST"]] call BIS_fnc_addStackedEventHandler;
I used this in stable without error
If you're on dev and this errors, then it's a bug on dev
To confirm you could try again without mods.
tough abyss
20:44:43 Error in expression <[B Alpha 1-1:1 (Lappihuan),"TEST"] call {sy>
20:44:43 Error position: <Alpha 1-1:1 (Lappihuan),"TEST"] call {sy>
20:44:43 Error Missing ]
Yup, same as the initial error
They b0rked it
And because I'm an asshole: Arma.Studio wouldn't have found this in a million years.
tough abyss
just confirmed without any mods, same error as above
thanks for the help! @little eagle
little eagle
yw
queen cargo
ArmA.Studio can look easily in those functions and check the various contents of variables at runtime when you glance it
it would have found it faster then you stupid
little eagle
than*
queen cargo
commy is the active dude in this chat
tough abyss
https://feedback.bistudio.com/T126078
ticket for the issue before, if someone wants to follow it
pliant stream
that's some crazy shit... what is their implementation like this? ```SQF
_init = "";
{
if (count _init > 0) then
{ _init = _init + ","; };
_init = _init + str _x;
} foreach _args;
_eh = compile format ["[%1] call %2", _init, _code];```
little eagle
Is that actual code or did you made that yourself?
halcyon crypt
judging by the words it's made up
little eagle
Because your loop is a convoluted way of writing:
_init = str _args;
I wouldn't be surprised if it's something dumb like that.
jade abyss
+it makes no real sense, tbh
Why not _init = []
then pushback everything? ยฏ_(ใ)_/ยฏ
little eagle
Because that only works with an array already made out of strings.
rotund cypress
No it works I am pretty sure
jade abyss
@little eagle
["[%1] call %2", _init, _code];
little eagle
["1",2,text "3"] joinString ""; // "123"
Because that only works with an array already made out of strings. - Not correct
little eagle
See my two comments above
rotund cypress
I don't know why, just pointed out that it's inneccessary to count/foreach to add string to string instead of just using joinString ๐
little eagle
Or, str in this case
little eagle
@jade abyss That results in one pair of brackets too much
jade abyss
I quoted from the Code above
_eh = compile format ["[%1] call %2", _init, _code];```
Removing the [] in that above + doing it with pushback in the array -> Makes more sense to me.The whole thing can be reduced to:
_eh = compile format ["%1 call %2", _args, _code];
unless you want all _args stringed (hell knows why)
little eagle
That was the point of the loop.
And it wasn't to achieve anything. It was to explain a specific error on dev branch.
little eagle
_init is a string You cannot use pushBack on a string
little eagle
Because that is just as pointless, because the whole code is a one liner as I've shown.
jade abyss
Not rly, but okay.
jade abyss
w/e. it's okay.
little eagle
Admission of defeat.
jade abyss
Not rly, just not in the mood
little eagle
Then I declare victory and demand land and reparations.
jade abyss
take it, leaving world anyway
*startingRocket*
little eagle
๐ ๐ ๐
jade abyss
๐ -> ๐
vapid frigate
that whole block of code could be replaced with [_args] call _code ?
don't need to convert it into a string and then compile it
err nvm.. thought it was 'call compile'
lean tiger
Awwwww cabbage patch kids
There isn't a generic CfgNotification that accepts parameterized everything :<
Am writing a plugin that I was hoping would simply be (drop into mission folder, add init.sqf reference) and go
IntelAdded is the closest, if I'm not mistaken ๐ฆ
outer scarab
Hey. I'm trying to get a random one of three sounds to play in a trigger, but I'm having trouble. Note, I'll be of limited help since the majority of this is donated, borrowed or otherwise cobbled together. Here's my description.ext: class CfgSFX { class Ion { sound0[] = {"\music\ion1.ogg", db-0, 1.0, 1000, 0.2, 0, 15, 30}; sound1[] = {"\music\ion2.ogg", db-0, 1.0, 1000, 0.2, 0, 15, 30}; sound2[] = {"\music\ion3.ogg", db-0, 1.0, 1000, 0.2, 0, 15, 30}; sounds[] = {sound0, sound1, sound2}; empty[] = {"", 0, 0, 0, 0, 0, 0, 0}; }; class stormloop { name = "Ion Storm Loop"; sound[] = {\music\stormloop.ogg, db+4, 1.0}; }; };
and the trigger:
private _pos = while {true} do {
private _pos = [_this] call BIS_fnc_randomPosTrigger;
if (count (_pos nearEntities [["Man"],200]) == 0) exitWith {_pos};
};
_bolt = createvehicle ["LightningBolt",_pos,[],0,"can collide"];
_bolt setposatl _pos;
_bolt setdamage 1;
private _light = "#lightpoint" createvehiclelocal _pos;
_light setposatl [_pos select 0,_pos select 1,(_pos select 2) + 10];
_light setLightDayLight true;
_light setLightBrightness 95;
_light setLightAmbient [0.6, 0.6, 0.1];
_light setlightcolor [1, 0.95, 0.3];
sleep 0.1;
_light setLightBrightness 0;
sleep 0.1;
private _class = selectRandom ["lightning1_F","lightning2_F"];
_lightning = _class createvehiclelocal [100,100,100];
_lightning setpos _pos; _duration = random 2;
for "_i" from 0 to _duration do {
_time = time + 0.1;
_light setLightBrightness (100 + random 100);
[_lightning,"Ion",500] call CBA_fnc_globalSay3d;
waituntil {time > _time};
};
deletevehicle _lightning;
deletevehicle _light;
};```Main issue i'm having right now is it doesn't seemto recognise "ion"
If I create a trigger to just play the sound, I get something along the lines of "no entry description.ext/sfx/ion1.empty"
vapid frigate
ion1 isn't refrenced anywhere there
you got something else trying to play ion1? or the error is different than what you said?
outer scarab
I was trying something different when I got that error
I'll try having it play one of the specific sounds, gimme a sec
outer scarab
Just gives me "Sound ion1 not found"
vapid frigate
dunno sorry, guess it's saying it can't find the .ogg file
i thought that would in yourmission\music\ion1.ogg
outer scarab
yeah, that's where I got it
vapid frigate
maybe try taking out the \ at the start
vapid frigate
the error is kinda odd.. thought it would say can't find \music\ion1.ogg, not just ion1
checked your RPT for any hints?
outer scarab
6:25:21 Warning Message: Sound ion1 not found is the only relevant thing as far as I can see
vapid frigate
i guess to confirm that is the problem, you could try changing ion1 to somethingelse1
vapid frigate
i'm not sure about that CBA function, but say3D says it's for CfgSounds
not CfgSFX
outer scarab
Ooooh
that might be it
hmm
desc.ext
{
class Strike
{
sound0[] = {"music\strike1.ogg", db-0, 1.0, 1000, 0.2, 0, 15, 30};
sound1[] = {"music\strike2.ogg", db-0, 1.0, 1000, 0.2, 0, 15, 30};
sound2[] = {"music\strike3.ogg", db-0, 1.0, 1000, 0.2, 0, 15, 30};
sounds[] = {sound0, sound1, sound2};
empty[] = {"", 0, 0, 0, 0, 0, 0, 0};
};
};```not sure if cfgsounds is the same format
outer scarab
Yeah, I was wondering the same
vapid frigate
doesn't seem to have random options available, so you might need to just make 3 cfgsounds and play a random one in the script
outer scarab
Now i'm getting "No entry [...]description.ext/cfgsounds/strike1.title"
{
sounds[] = {};
class strike1
{
name = "strike1";
// start path to sound file in AddOn with @
sound[] = {{\music\strike1.ogg};
titles[] = {0,""};
};
class strike2
{
name = "strike2";
// start path to sound file in AddOn with @
sound[] = {{\music\strike2.ogg};
titles[] = {0,""};
};
class strike3
{
name = "strike3";
// start path to sound file in AddOn with @
sound[] = {{\music\strike3.ogg};
titles[] = {0,""};
};```yeh not sure why it's looking for that.. could try throwing in title = "";, but it's not in the docs
not that that necessarily means anything
your brackets are messed up there
and missing the numbers in the sound array
{
sounds[] = {};
class addonsound1
{
name = "sound from addon";
// start path to sound file in AddOn with @
sound[] = {"@a3\Ui_F_Curator\Data\Sound\CfgSound\visionMode", 0.8, 1, 100};
titles[] = {0,""};
};
};```Yeah, I corrected the brackets, don't know what the numbers are on about
outer scarab
{
sounds[] = {};
class strike1
{
name = "strike1";
// start path to sound file in AddOn with @
sound[] = {"\music\strike3.ogg, 1, 1, 300};
titles[] = {0,"Ion Strike"};
};
class strike2
{
name = "strike2";
// start path to sound file in AddOn with @
sound[] = {"\music\strike3.ogg, 1, 1, 300};
titles[] = {0,"Ion Strike 2"};
};
class strike3
{
name = "strike3";
// start path to sound file in AddOn with @
sound[] = {"\music\strike3.ogg, 1, 1, 300};
titles[] = {0,"Ion Strike 3"};
};```looking good?
vapid frigate
yeah, you probably want blank titles though (but shouldn't hurt). that's subtitles
missing a quote too
after the .oggs
vapid frigate
the error about strike1.title ?
outer scarab
{
sounds[] = {};
class strike1
{
name = "strike1";
// start path to sound file in AddOn with @
sound[] = {"\music\strike3.ogg", 1, 1, 300};
titles[] = {0,"Ion Strike"};
};
class strike2
{
name = "strike2";
// start path to sound file in AddOn with @
sound[] = {"\music\strike3.ogg", 1, 1, 300};
titles[] = {0,"Ion Strike 2"};
};
class strike3
{
name = "strike3";
// start path to sound file in AddOn with @
sound[] = {"\music\strike3.ogg", 1, 1, 300};
titles[] = {0,"Ion Strike 3"};
};
class stormloop
{
name = "Ion Storm Loop";
sound[] = {\music\stormloop.ogg, db+4, 1.0};
};
};```Yep
vapid frigate
i guess try adding it: class strike1 { name = "strike1"; // start path to sound file in AddOn with @ sound[] = {"\music\strike3.ogg", 1, 1, 300}; titles[] = {0,"Ion Strike"}; title = ""; };
that's what the error is saying it's expecting
vapid frigate
it won't be saying No entry [...]description.ext/cfgsounds/strike1.title with that
outer scarab
nah, it's still the same, but it's strike1.titles
vapid frigate
you definitely saving the right file and stuff? it would come up that error if cfgsounds wasn't in description.ext at all
outer scarab
{
sounds[] = {};
class strike1
{
name = "strike1";
// start path to sound file in AddOn with @
sound[] = {"\music\strike1.ogg", 1, 1, 300};
titles[] = {0,"Ion Strike"};
title = "";
};
};```filename strike1.ogg
vapid frigate
and reloading the mission to make sure it picks it up
outer scarab
oh, in the editor?
Sure
I'll try that lol
the trigger as it currently is for posterity nul = thisTrigger spawn { private _pos = while {true} do { private _pos = [_this] call BIS_fnc_randomPosTrigger; if (count (_pos nearEntities [["Man"],200]) == 0) exitWith {_pos}; }; _bolt = createvehicle ["LightningBolt",_pos,[],0,"can collide"]; _bolt setposatl _pos; _bolt setdamage 1; private _light = "#lightpoint" createvehiclelocal _pos; _light setposatl [_pos select 0,_pos select 1,(_pos select 2) + 10]; _light setLightDayLight true; _light setLightBrightness 95; _light setLightAmbient [0.6, 0.6, 0.1]; _light setlightcolor [1, 0.95, 0.3]; sleep 0.1; _light setLightBrightness 0; sleep 0.1; private _class = selectRandom ["lightning1_F","lightning2_F"]; _lightning = _class createvehiclelocal [100,100,100]; _lightning setpos _pos; _duration = random 2; for "_i" from 0 to _duration do { _time = time + 0.1; _light setLightBrightness (100 + random 100); [_lightning,"strike1",500] call CBA_fnc_globalSay3d; waituntil {time > _time}; }; deletevehicle _lightning; deletevehicle _light; };
vapid frigate
don't really need to paste all that.. only relevent line is [_lightning,"strike1",500] call CBA_fnc_globalSay3d;
outer scarab
Fairs
good news and bad news though
the error's gone
but I don't hear the sound
hah
Can volume go over 1?
vapid frigate
you close enough? (within 300m)
vapid frigate
yeah pretty sure volume can be greater than 1
outer scarab
Yeah, it is occuring at the same time as a loud explosion
okay, interesting
I upped the volume to 100 cause why not
I do hear the sound, but it seems it only plays until _lightning is deleted, which is more or less instant
This is where it goes beyond my comprehension
So globalSay3d needs the position for the duration of the sound to play it for its entire length
vapid frigate
it says that will happen here https://community.bistudio.com/wiki/say3D
outer scarab
I think it goes without saying I haven't read too far into this
valid dew
How do I use a parameter value in my Description.ext?
I'm trying to make respawn time a parameter
outer scarab
I guess I could create a "null" vehicle with the script, and delete that later that _lightning, but I wouldn't know how to go about that
vapid frigate
@outer scarab just create a Logic (or maybe it's Logic_F.. check config viewer)
outer scarab
getting a "generic error in expression"
private _pos = while {true} do {
private _pos = [_this] call BIS_fnc_randomPosTrigger;
if (count (_pos nearEntities [["Man"],200]) == 0) exitWith {_pos};
};
_bolt = createvehicle ["LightningBolt",_pos,[],0,"can collide"];
_bolt setposatl _pos;
_bolt setdamage 1;
_snd = createvehicle ["Logic",_pos];
_snd setposat1 _pos;
private _light = "#lightpoint" createvehiclelocal _pos;
_light setposatl [_pos select 0,_pos select 1,(_pos select 2) + 10];
_light setLightDayLight true;
_light setLightBrightness 95;
_light setLightAmbient [0.6, 0.6, 0.1];
_light setlightcolor [1, 0.95, 0.3];
sleep 0.1;
_light setLightBrightness 0;
sleep 0.1;
private _class = selectRandom ["lightning1_F","lightning2_F"];
_lightning = _class createvehiclelocal [100,100,100];
_lightning setpos _pos; _duration = random 2;
for "_i" from 0 to _duration do {
_time = time + 0.1;
_light setLightBrightness (100 + random 100);
[_lightning,"strike1",500] call CBA_fnc_globalSay3d;
waituntil {time > _time};
};
deletevehicle _lightning;
deletevehicle _light;
sleep 5;
deletevehicle _snd
};```oh, i see it
or not
outer scarab
I managed to fix it within the script
don't know exactly what I did
apparently a missing ; or 2, one of which was at the bottom
problem I'm having now is whether " _snd = createvehicle ["Logic",_pos]; " is legit
outer scarab
Yeah, I'm not the one who made this in the first place, I'm just trying to cobble together a way to play a sound
It works for all intents and purposes, except that
[_lightning,"strike1",500] call CBA_fnc_globalSay3d; works, but the sound stops when _lightning is deleted
dusk sage
This here: random 2, seems silly, for your for loop
outer scarab
so I'm trying to create a logic to get the position to play the sound from
outer scarab
Yeah, I appreciate pointing out the stuff that's wrong - but like I said, I didn't make the original version of this script
dusk sage
I'm not sure what CBA_fnc_globalSay3d is
But if it's what I imagine it to be, I doubt you should use it
remoteExec with say3D would be better
And yes, say3D will cancel when the object is deleted
You need to delay that deletion if you want it to continue
outer scarab
I'm specifically looking for what I'm doing wrong here: _snd = createvehicle ["Logic",_pos,[],0,"can collide"]; _snd setposatl _pos;
``` [_snd,"strike1",500] call CBA_fnc_globalSay3d; ```
};```can collide
And I don't think you can create a logic with createVehicle
As of Arma2, this command cannot be used to create game logics (source). Use createUnit instead.
outer scarab
Ahah
vapid frigate
@valid dew you can use missionConfigFile
same was as configFile for normal configs
so with this in your description.ext:foo = bar; class foobar { bar = foo; }; you could get the first one with missionConfigFile >> 'foo', or for the second, missionConfigFile >> 'foobar' >> 'bar'
(with getText before them to get the text if it's a string; or getNumber, getArray, etc)
warm gorge
Performance warning: SimpleSerialization::Write 'params' is using type of ',TEXT' which is not optimized by simple serialization, falling back to generic serialization, use generic type or ask for optimizations for these types What does this mean exactly? I cant seem to find where its arising from
outer scarab
Okay, I tried something different:
private _pos = while {true} do {
private _pos = [_this] call BIS_fnc_randomPosTrigger;
if (count (_pos nearEntities [["Man"],200]) == 0) exitWith {_pos};
};
_bolt = createvehicle ["LightningBolt",_pos,[],0,"can_collide"];
_bolt setposatl _pos;
_bolt setdamage 1;
_A = getPosASL _bolt;
private _light = "#lightpoint" createvehiclelocal _pos;
_light setposatl [_pos select 0,_pos select 1,(_pos select 2) + 10];
_light setLightDayLight true;
_light setLightBrightness 95;
_light setLightAmbient [0.6, 0.6, 0.1];
_light setlightcolor [1, 0.95, 0.3];
sleep 0.1;
_light setLightBrightness 0;
sleep 0.1;
private _class = selectRandom ["lightning1_F","lightning2_F"];
_lightning = _class createvehiclelocal [100,100,100];
_lightning setpos _pos; _duration = random 2;
for "_i" from 0 to _duration do {
_time = time + 0.1;
_light setLightBrightness (100 + random 100);
playSound3D ["strike1.ogg", _pos, false, _A, 100, 1, 1000];
waituntil {time > _time};
};
deletevehicle _lightning;
deletevehicle _light;
};```Ooh
Nah, doesn't work
Tried putting _A = getPosASL _bolt; before _bolt setdamage 1; no dice
dusk sage
Can you elaborate more on "doesn't work"?
And as before, I'd recommend remoteExec with say3D
outer scarab
I think I've figured it out anyway
playSound3D ["strike1.ogg", _pos, false, (getPosASL _Lightning), 100, 1, 1000];
Buuut now, 8:21:22 Sound: Error: File: strike1.ogg not found !!!
{
sounds[] = {};
class strike1
{
name = "strike1";
// start path to sound file in AddOn with @
sound[] = {"\music\strike2.ogg", 100, 1, 3000};
titles[] = {0,""};
title = "";
};```Was working fine when I was using CBA_fnc_globalSay3d
All that's changed now is the function I'm using in the trigger
also tried playSound3D ["strike1", _pos, false, (getPosASL _Lightning), 100, 1, 1000];
dusk sage
And as before, I'd recommend remoteExec with say3D
And you should read the wiki page: https://community.bistudio.com/wiki/playSound3D
vapid frigate
wasn't your sound in a music folder @outer scarab ?
outer scarab
Yeah
I just tried music\strike2.ogg
since the example on the wiki page gives a full file-path
vapid frigate
read the comments on the wiki too
basically it's painful with a sound in a mission to use playSound3D
outer scarab
This just got 10x more complicated, if you're referring to this comment: You need to get the correct path for custom mission sounds. Use missionConfigFile with BIS_fnc_trimString (to trim off "description.ext"), then add your mission's sound directory and sound file: _soundPath = [(str missionConfigFile), 0, -15] call BIS_fnc_trimString; _soundToPlay = _soundPath + "sounds\some_sound_file.ogg"; playSound3D [_soundToPlay, _sourceObject, false, getPos _sourceObject, 10, 1, 50]; //Volume db+10, volume drops off to 0 at 50 meters from _sourceObject
vapid frigate
yeah, exactly
hence BoGuu: And as before, I'd recommend remoteExec with say3D
you almost had it right before, but you had "can collide" instead of "CAN_COLLIDE"
err nvm missed boguus thing about not using createvehicle
outer scarab
Well the trouble I had with createvehicle logic is as BoGuu pointed out, you gotta use create - yeah
vapid frigate
create a softdrink can instead
outer scarab
I kinda gave up on that, then someone recommended I just use playsound3d
Hmm, I could
Would it have any performance impact? These lightning strikes are probably going to be happening quicker than 1/second
vapid frigate
(or use createUnit)
dusk sage
a strike every second? ๐
outer scarab
Yeah man
vapid frigate
i would guess less overhead than 'createunit', but not certain
outer scarab
~ambience~
vapid frigate
(make sure you delete it afterwards)
dusk sage
I wouldn't advise to spam create lightning and sounds every second
Won't it piss people off aswell ๐คฃ ?
outer scarab
That's kinda the idea
I mean we tested it before I got to implementing sounds, apparently it wasn't perceptibly worse than any other mission
In terms of performance I mean
but a soda can is double the objects
I don't know, it's worth a try anyhow
little eagle
It's always a bad idea to piss your players off.
outer scarab
Question: Does playSound3D delay the sound by distance as with others? As far as I can tell, the CBA function doesn't
little eagle
Someone said yesterday no
little eagle
Couldn't you reuse the same dummy object?
outer scarab
I.. don't know about that
outer scarab
This is pretty much my first attempt at scripting anything
That'd probably be better
little eagle
And the lightest simulation type
The soda can thing was for barriers, because AI reacts to them.
iirc
No reason to do that for say3D
outer scarab
In this instance I'm not using say3D, but CBA's globalSay3d
little eagle
But that function is just a wrapper for
remoteExecCall ["say3D"];
....oh
little eagle
So you're still using the same thing and you don't consider the speed of sounds either.
outer scarab
Yeah, I'm guessing getting speed of sound involved if even possible would take me the rest of the day to figure out
little eagle
vapid frigate
you'll probly get a random amount of delay anyway just from lag
little eagle
Network delay, yeah
outer scarab
I mean it'd be nice to have, because the lightning strike also creates an explosion at the point of impact
and that does take speed of sound into account
little eagle
jokoho is currently doing speed of sound delays for the cookoff stuff in ACE
But it's a bit... advanced?
vapid frigate
sleep (_obj distance player) / 340.29;
it won't be accurate with network lag though
you could make the whole script remote exec instead of just the say3d and do everything locally
then everything will be in sync on all clients.. but might not happen at the identical time for everyone
outer scarab
Well, that's not too much of an issue really
but I think we're getting ahead of ourselves, and way ahead of me ๐
Because, once again, I must've broke something
no script error, nothing in the .rpt, no sound
Oh, I think I know why actually
i have "CAN_COLLIDE" on the Building
And the lightning bolt is destroying it I'm guessing
vapid frigate
outer scarab
I never asked for this
I think I've got it as close to working as I reasonably can
tried disabling simulation on _snd, not sure if it helped, but the sound only plays half the time
tough abyss
@little eagle look what got added to the devbranch changelog...
Tweaked: The BIS_fnc_addStackedEventHandler, BIS_fnc_removeStackedEventHandler and BIS_fnc_executeStackedEventHandler functions are no longer supported
little eagle
It cannot pass arguments, so not necessarily.
Also some events cannot overwrite input, like clicking on the map when using the mission eventhandler version.
modern sigil
Quick question -- rearming an aircraft is easy enough when you're setting the magazineTurret's ammo... but how do we add chaff/flares back to the aircraft? Because they're not part of the PylonLoadout or the [-1] turret path.
modern sigil
Right but I need the turret path, and I don't see where I can find the turretPath for the chaff/flares. It's not the usual [-1] for the main pilot, which is where everything else is stored.
Does the chaff even use a turret path? I don't even know
little eagle
I will try on that A10
tough abyss
so they discontinue support of a function and don't provide a alternative ๐
little eagle
cameraOn magazinesTurret [-1]
["1000Rnd_Gatling_30mm_Plane_CAS_01_F","Laserbatteries","120Rnd_CMFlare_Chaff_Magazine","PylonRack_1Rnd_Missile_AA_04_F","PylonRack_7Rnd_Rocket_04_HE_F","PylonRack_3Rnd_Missile_AGM_02_F","PylonMissile_1Rnd_Bomb_04_F","PylonMissile_1Rnd_Bomb_04_F","PylonMissile_1Rnd_Bomb_04_F","PylonMissile_1Rnd_Bomb_04_F","PylonRack_3Rnd_Missile_AGM_02_F","PylonRack_7Rnd_Rocket_04_AP_F","PylonRack_1Rnd_Missile_AA_04_F"]
looks like it IS in turret [-1]
"120Rnd_CMFlare_Chaff_Magazine"
@modern sigil
cameraOn removeMagazinesTurret ["120Rnd_CMFlare_Chaff_Magazine", [-1]];
cameraOn addMagazineTurret ["120Rnd_CMFlare_Chaff_Magazine", [-1]];
This works perfectly fine for me
modern sigil
Really? [-1] is the turretPath for flares? I need to go back and see what I was doing wrong.
Thanks @little eagle. I'll give it a shot and report back in a bit.
indigo snow
@modern sigil -1 turret is the driver / pilot of the vehicle
pliant stream
_eh = compile format ["(%1) call %2", str _args, _code];``` there i revised my code from earlier... of course the whole the point wasn't some string manip, but rather embedding the context in code as literals, which is obviously horribly broken... the defect report showed the result of `str(<unit>)` being compiled as a literalThe str is superfluous and so are the parentheses, no?
You're not getting off the hook.
But yeah, I agree. That could explain the issue
pliant stream
superfluous eh? have a go without them then
little eagle
I just wonder why they'd change it like that and then remove support for that function entirely one day later.
have a go without them then
Example code?
Because atm it does nothing
pliant stream
fine SQF [/*_args*/ "", /*_code*/ {systemchat str _this}] call setEventHandler
do you see the problem?
little eagle
No, looks fine to me.
vapid frigate
depends what _args is
pliant stream
" call {systemchat str _this}"```whether it can be serialized back and forwards as text
pliant stream
do you see the problem now?
little eagle
Can you give an actual example I can paste into the debug console and have an error when not using str that disappears when using str?
pliant stream
format ["%1 call {}", ""] -> " call {}"
format ["%1 call {}", str ""] -> """ call {}"```surely _args can be anything
pliant stream
yeah, not in the first code i posted, but it can in the second one
little eagle
Because we are talking about a specific example code that turns the array _args into a sting called _init
pliant stream
the one where you said the str was superfluous
little eagle
str is superfluous under the condition that _args is an array as it was in the initial problem
pliant stream
sure sure, but my revised code snippet didn't impose such a requirement
little eagle
So yes, str is superfluous, but you were talking about a different problem where it no longer is
But then you shifted the topic
pliant stream
i did also note that the string manip was never the point in the first place
vapid frigate
do you really need to convert it to a string and back?
pliant stream
no, you must not
vapid frigate
can't just store it as _eh = [_args, _code];
pliant stream
that was the point
little eagle
Your result
"" call {}
makes no sense anyway, because it can no longer explain the issue this was all about
pliant stream
converting the callback context into literals is horribly broken, that was the point
little eagle
If the argument were a string, then the code would error out differently.
sure, but then we're talking about a different error that doesn't exist
vapid frigate
it can't work if, for example, _args was an object
pliant stream
exactly
little eagle
Yes!
pliant stream
or if _args was an array...
little eagle
And that's the problem this was all about
pliant stream
it only works for strings, bools, and scalars
i think... maybe something else but i don't think so
little eagle
No, it also works for arrays!!!
vapid frigate
would work for arrays of those things too
pliant stream
completely and utterly broken
little eagle
Example:
call compile format ["%1 call %2", [0,1], {systemChat str _this}]
works fine
pliant stream
but it loses the array reference
making the context pointless
it can no longer modify the original array
vapid frigate
yeah it creates a new array when you recompile it
little eagle
You're again talking about an issue that did not happen.
And whether or not passed arguments to stacked eventhandlers should be passed by reference or not is an open question imo.
It shouldn't change suddently ideally for bwc
pliant stream
i don't care if someone did it or not, or if it happened or not, i was talking about the fact that converting a callback context to literals is broken, both for types like objects, and also for just simple arrays with scalars, strings, etc...
little eagle
But it's not clear what the expected result should be unless specified.
Maybe the function already copies the input? Idk
pliant stream
and obviously a callback context should be passed by reference... otherwise it's pointless
little eagle
How is it pointless when my code works just fine?
pliant stream
how do you expect to return from an asynchronous operation without the array reference?
little eagle
I disagree that the arrays has to be passed by reference.
pliant stream
depends on the application, but in contextual callbacks they definitely should
little eagle
Idk what asynchronous operation has to do with this. We are talking about eventhandlers that run unscheduled.
vapid frigate
has something changed? i don't understand the problem
if you convert into a string then back to code, it's always gonna recreate arrays?
little eagle
BIS_fnc_addStackedEventHandler is either broken or discontinued on dev branch
pliant stream
yes, that's the problem, it creates a new array
little eagle
That is not the problem this is about, mate
vapid frigate
you're trying to recreate it?
pliant stream
yeah, that may not be the defect that was reported and sparked the whole discussion, but it's still valid
jade abyss
yes, that's the problem, it creates a new array```
Erm, and?Nope, I am trying to explain the shown error. No idea what untelo is about.
jade abyss
ยฏ_(ใ)_/ยฏ
vapid frigate
and that's what's breaking it
little eagle
Lecks, we don't know why it's broken, but that would explain the error message.
pliant stream
@jade abyss copying the array prevents you from modifying the original (obviously), which prevents you from returning the value of an asynchronous operation
jade abyss
uhm, what? oO
little eagle
The goal was to create code sufficient to explain why a specific error messgae is shown.
little eagle
And some people hijacked the conversation to "fix" the code that was never meant to be used
jade abyss
some are bored ยฏ_(ใ)_/ยฏ
little eagle
And are off topic
little eagle
The channel topic is scripting. The conversation topic is about scripting. But that does not mean that the conversation topic should be changed to a different one and that it's fine to do so just because it still complies with the channel topic.
Obviously there is no point in stringifying an array when you immediately recompile it.
But it can explain why an object without object name would error out exactly like what was found in an RPT file that happened on dev branch.
pliant stream
yes that was the point, and to back it up i gave two examples of why it's broken, 1) because not all types have literal representations and 2) because array references are lost
little eagle
And how does that explain the error on dev branch?
pliant stream
the code i posted did explain it, converting the callback context into literals
outer scarab
How would I go about finding out how many lightpoints are in my mission, or making sure they're working? I've placed 34 with ```var = [] spawn {light = "#lightpoint" createvehicle getpos L7;
light setlightbrightness 5;
light setlightcolor [0.1,0.8,0.1];
light setlightambient [0,0.8,0];};````
in init.sqf, but for whatever reason none of them are showing up now.
little eagle
the code i posted did explain it, converting the callback context into literals
So you don't even know what the error on dev branch is. K thx.
pliant stream
well i was talking about this one https://feedback.bistudio.com/T126078
it tries to compile SQF [B Alpha 1-1:1 (Lappihuan),"TEST"]
which looks like someone tried to convert the context into a literal, but it fails because objects don't have a literal representation
little eagle
well, they do when you give them object names in the ediotr "vehicle var names"
So that is important when trying to repro the issue.
pliant stream
ok, let's say it worked for objects, my point was that even if it worked for objects, the whole idea is still broken for other reasons
little eagle
Sure.
It's also pointless.
Yet my derived one liner code is identical to what was posted as explanation.
And the explanation had the same issue.
So if my one liner keeps all the issues of the original, it's even better suited as simplification
Because as I said, it's literally the same, but as a one liner...
pliant stream
i'm not sure what one liner code you're referring to here
little eagle
Wait
original:
_init = "";
{
if (count _init > 0) then
{ _init = _init + ","; };
_init = _init + str _x;
} foreach _args;
_eh = compile format ["[%1] call %2", _init, _code];
simplification:
_eh = compile format ["%1 call %2", _args, _code];
i had a quick glance at the functions yesterday late night and they seem to store the arguments and the code in a missionNamespace array but i don't remember if they stringify them or not
little eagle
On dev branch?
tough abyss
yes
outer scarab
var = [] spawn {light = "#lightpoint" createvehicle getpos L7;
light setlightbrightness 5;
light setlightcolor [0.1,0.8,0.1];
light setlightambient [0,0.8,0];};
`
in init.sqf, but for whatever reason none of them are showing up now.```i can paste the functions on my lunchbreak
little eagle
original and simplification are equivalent, unless you want to deliberately change the _init variable, which would be horrible design.
jade abyss
Wich one exactly @tough abyss ?
little eagle
BIS_fnc_addStackedEventHandler
tough abyss
The BIS_fnc_addStackedEventHandler, BIS_fnc_removeStackedEventHandler and BIS_fnc_executeStackedEventHandler
tough abyss
exectue is the one that reads from missionNamespace and add is the one that saves it there
pliant stream
@little eagle what do you mean by deliberately changing _init?
jade abyss
little eagle
Changing the value of a predefined variable _init
jade abyss
read from MissionNameSpace
little eagle
Assuming the lack of privatewas intentional.
pliant stream
_init is magic?
little eagle
Nope
pliant stream
i just named it that as short for initializer
because it was the array initializer without []
little eagle
Think of both codes as a black box
pliant stream
or are you talking about local var corruption due to lack of local?
or private in A3 i guess...
tough abyss
@jade abyss could you paste the execute function aswell? ^^
jade abyss
@tough abyss The one i pasted is the one, how it seems. But i can't find that Codepart, that is reffered to in here oO
little eagle
_init = 0;
ORIGINAL
typeName _init // STRING
_init = 0;
SIMPLIFICATION
typeName _init // NUMBER
or are you talking about local var corruption due to lack of local?
No
jade abyss
Paste = fn_addStackedEventHandler.sqf
tough abyss
BIS_fnc_executeStackedEventHandler
pliant stream
ok then i have no clue what you're talking about
tough abyss
few files below
jade abyss
fn_executeStackedEventHandler.sqf
/*
Author: Nelson Duarte
Description:
Deprecated
Returns:
BOOLEAN - FALSE
*/
false;```๐
tough abyss
wait wat?
tough abyss
that wasn't the case yesterday evening 100%
little eagle
ok then i have no clue what you're talking about
How so? The example is easy to understand.
tough abyss
and i'm 100% sure they didn't had that deprecated entry in the devbranch changelog...
pliant stream
@little eagle so you are talking about corrupting _init due to lack of local
or... private whatever...
little eagle
@outer scarab
allMissionObjects "#lightpoint"
That reports an array with all light source objects.
Yes. Corrupting, or changing it deliberately and the simplification no longer doing that.
pliant stream
well obviously my code was just a silly example and was never intended to be valid for any purpose
little eagle
Do you agree that my simplification is equivalent to your original aside from modifying potentially predefined variables in a different way?
outer scarab
Hmm, Commy I don't imagine there's a way to count only editor, well, init-placed lightpoints?
pliant stream
well no i don't, because it behaves differently when _args is string, but otherwise, yes of course
little eagle
I don't think allMissionObjects lists anything else.
pliant stream
assuming _args is always array, it's equivalent and much better, of course
little eagle
Untelo
_args is defined as array by the original
} foreach _args;
That doesn't work with strings.
pliant stream
i just had a brain fart about array to string, but once again the string manip was not important
ok i had no clue i might as well have called _args _floop and _code _blargh
jade abyss
So again: Where is that codepart from?
that's some crazy shit... what is their implementation like this? ```
```sqf
_init = "";
{
if (count _init > 0) then
{ _init = _init + ","; };
_init = _init + str _x;
} foreach _args;
_eh = compile format ["[%1] call %2", _init, _code];```it was just meant to demonstrate turning the callback context into a code literal
jade abyss
So it was nonsense you posted/wrote?
little eagle
ok i had no clue i might as well have called _args _floop and _code _blargh
Sure, but that doesn't change the fact that forEach makes the variable _floop to be an array if we assume that the code works.
tardy wagon
SQF seems weird as hell
little eagle
And I demand reparations
jade abyss
@tardy wagon Just forget everything you know about other languages (even your mothertounge)
little eagle
jk, it was a fun and pointless discussion
pliant stream
@tardy wagon i see you're a connoisseur of the finest of programming languages known to man, thanks be to Bjarne
now... to keep the flame war going but with a new topic... how to get the return value when i call remote functions using remoteExec etc?
tough abyss
seems like a troll dickmove to drop support of a function and in the same time just break it intentionally by doing a return false;
little eagle
now... to keep the flame war going but with a new topic... how to get the return value when i call remote functions using remoteExec etc?
Another remoteExec I guess.
Yeah, Lappi. No idea what's going on.
pliant stream
but how do i associate it with the original caller?
little eagle
clientOwner I guess. Hoping that the code is not meant to work in local hosted MP
pliant stream
_blah = [_code, _args] call some_remoteExec_wrapper_that_has_the_ability_to_return_values```how do i derive the return value from _blah
pliant stream
well the way i would do it is by returning a future
now... how would you implement futures
peak plover
What? remotexec that returns a value? Do you want a value from another client?
pliant stream
i want to call a function on a remote machine, and get back the value from the function
pliant stream
what the remote machine is or what the function is are irrelevant
what is relevant is how to approach returning values from asychronous operations
such as a remote call
jade abyss
RemoteExec it back?
call a func on the Target, that returns the data. Can't think of anything else
peak plover
run script_a on machine_1 >> remotexec script_B on remote machine_2 >> script_B remoteExecs script_C on machine_1
little eagle
some_remoteExec_wrapper_that_has_the_ability_to_return_values = {
params ["_code", "_args"];
My_RE_Return = nil;
[[_code, _args, 0], {
params ["_code", "_args"];
private _return = _args call _code;
[_return, { My_RE_Return = _this; }] remoteExec ["BIS_fnc_call"];
}] remoteExec ["BIS_fnc_call"];
waitUntil {!isNil "My_RE_Return"};
My_RE_Return
};
Something like this?
jade abyss
Or let it change a setVar Pub. no clue
pliant stream
well i can think of publicvariables, which is what i actually do because i live in A2 land
pliant stream
yes this is all very obvious... what i'm getting at though is that using any more global state than absolutely necessary is very bad
which is why, as i said, i would personally implement this using futures and promises
now... implementing futures and promises without global state is possible using array references
jade abyss
call a func on the Target, that returns the data.```some_remoteExec_wrapper_that_has_the_ability_to_return_values = {
params ["_code", "_args"];
My_RE_Return = nil;
[[_code, _args, clientOwner], {
params ["_code", "_args", "_owner"];
private _return = _args call _code;
[_return, { My_RE_Return = _this; }] remoteExec ["BIS_fnc_call", _owner];
}] remoteExec ["BIS_fnc_call"];
waitUntil {!isNil "My_RE_Return"};
My_RE_Return
};
forgot the owner part
pliant stream
and this gets us to my point about why callback contexts should never copy arrays
jade abyss
BIS_fnc_call? Wouldn't a simple "call" be enough?
little eagle
remoteExec ["call"] would execute the code in unscheduled environment
peak plover
Ohh yeah, if it's more than 1 targets, then yeah. You would have to do some kind of a wait OR you could save the amount of targets you've sent out to on the original machine and then everytime the return function runs, check if it has got information from all the target machines yet
jade abyss
_Vargs RemoteExecCall ["call"]
things i did for debug stuff:
[_Veh,{player moveInDriver _this}] remoteExecCall ["call",_Player];```remoteExec ["call"] // unscheduled
remoteExecCall ["call"] // unscheduled
remoteExec ["BIS_fnc_call"] // scheduled
remoteExecCall ["BIS_fnc_call"] // unscheduled
peak plover
mind blown
little eagle
That is how it is.
pliant stream
don't get stuck on the implementation details for remoteExec
peak plover
Can that be added to the biki?
pliant stream
instead let's change the problem from remoteExec to a remote database load via extension on a server
little eagle
outer scarab
I'm really stuck on the details
little eagle
Scripted function - scheduled environment (suspension is allowed, i.e. spawn, execVM).
Script command - unscheduled environment (suspension is NOT allowed).
outer scarab
So much for having the finishing touches of my mission done within an hour or two
I've been at this since 4am
little eagle
call is a "Script command"
outer scarab
var = [] spawn {light = "#lightpoint" createvehicle getpos L34;
light setlightbrightness 5;
light setlightcolor [0.1,0.8,0.1];
light setlightambient [0,0.8,0];};
where L# is a trigger with the corresponding variablename
[11:05 AM] Morbo: All I remember having done between them working and when I noticed they weren't was 1: change the "Lightning" weather condition, and 2 started/reloaded the mission a hundred times messing with a script, that as far as I can tell shouldn't interface(edited)```So
remoteExec ["call"] -> unscheduled
jade abyss
Commy should start reading.
little eagle
Reading what?
jade abyss
[_Veh,{player moveInDriver _this}] remoteExecCall ["call",_Player];```remoteExec ["call"] would execute the code in unscheduled environment```Yeah. You can replace remoteExecCall with remoteExec here and it makes no diff
jade abyss
Are you drunk?
little eagle
No
jade abyss
So where did you mentioned that it's unscheduled? (fk, i hate that word)
little eagle
BIS_fnc_call? Wouldn't a simple "call" be enough?
It's not enough, because
{hint str canSuspend} remoteExec ["call"];
-> false
{hint str canSuspend} remoteExec ["BIS_fnc_call"];
-> true
little eagle
That is why it wouldn't be enough to use it. My code uses waitUntil therefore only works in scheduled environment
I think it's a good idea to keep that even when continuing on another machine
jade abyss
( i wonder why waitUntil anyway, but okay)
little eagle
Because that is the only way to do the initial task
vapid frigate
to wait for the var to be set
jade abyss
Send it back via rec -> Done O.O
little eagle
_blah = [_code, _args] call some_remoteExec_wrapper_that_has_the_ability_to_return_values
This was the initial task
And you have to use waitUntil to get this to work afaik
jade abyss
Wich is imho pretty nonsense ยฏ_(ใ)_/ยฏ
No need to waste stuff in a waitUntil ยฏ_(ใ)_/ยฏ
little eagle
Well, that is kinda how the new remote debug console in CBA works, so it's not "nonsense"
pliant stream
you absolutely DO NOT have to use waituntil
little eagle
Not waitUntil, but it's a similar idea
pliant stream
you can do it without the scheduled environment
jade abyss
Doesn't make it better. Split it up to 2 Funcs -> Done.
little eagle
you absolutely DO NOT have to use waituntil
Then show me code that would make:
_blah = [_code, _args] call some_remoteExec_wrapper_that_has_the_ability_to_return_values
working without waitUntil
pliant stream
ok
pliant stream
_future = [_code, _args, _client] call my_remoteExec;
[_future, _code2, _ctx] call my_future_then;```untelo, I expected some_remoteExec_wrapper_that_has_the_ability_to_return_values to be something
Not to change the question
pliant stream
renamed to my_remoteExec in version 2
little eagle
Ok, then solve
_blah = [_code, _args] call my_remoteExec;
how does my_future_ten work without waituntil?
little eagle
^
pliant stream
it attaches a callback to the future
which is executed when the future becomes ready
jade abyss
wut? no
little eagle
The right way to go about this is, that this cannot be solved without waitUntil and THEREFORE it's better to split it into two functions.
vapid frigate
so u 'wait until' it becomes ready?
outer scarab
Can someone help me here, please?
pliant stream
no
outer scarab
lol
jade abyss
Since when does it wait? oO
halcyon crypt
you need something that loops until whatever is doing something is done, period
little eagle
Yep
jade abyss
โ x2
pliant stream
when the client receives a response via some mechanism such as remoteExec, it sets the value on the promise, which executes all callbacks attached to the future
there is no need for any waituntil or sleep looping etc
that would be horrible
vapid frigate
show the functions?
jade abyss
Is this wishthinking or currently possible?
pliant stream
it's absolutely possible
little eagle
_blah = [_code, _args] call my_remoteExec;
Just so it doesn't get lost. I claim that this cannot work without my_remoteExec using waitUntil or sleep
jade abyss
cause i can't imagine, that it works.
pliant stream
i've implemented it already as a single class Task, but redoing it right now to separate into Promise and Future to make things easier
vapid frigate
it could if my_future_then uses sleep/waituntil
jade abyss
have you done this on a Dedi machine?
vapid frigate
but 1 of them has to
jade abyss
Yeah, Lecks
jade abyss
func pls.
little eagle
โ
jade abyss
I rly doubt, that RE(C) returns values. I would be pretty surprised if.
little eagle
It doesn't. The JIP version reports a ticket for the JIP stack
pliant stream
i'll have to finish this future/promise implementation first
jade abyss
REC
Anything - Nil in case of error. String otherwise. If JIP is not requested this is an empty string, if JIP is requested, it is the JIP ID. See the topic Function for more information.```i'll give you the code when it's presentable ok
jade abyss
Nah, just a showcase is enough
pliant stream
ok give me a minute....
little eagle
Poor @outer scarab is ignored ๐ฆ
jade abyss
Haven't you answered him already?
little eagle
Morbo - heute um 12:23 Uhr
Can someone help me here, please?
Untelo - heute um 12:23 Uhr
no
Morbo - heute um 12:23 Uhr
lol
We need #arma3_scripting and #arma3_scripting-questions with the Topic if(_Question == true)then{systemchat ask};
pliant stream
what no
halcyon crypt
he asked his question earlier
jade abyss
Check the current topic
simple solstice
i know, but it needs correction
jade abyss
No shit sherlock
simple solstice
lol
little eagle
== reports false when both sides are null (except for configNull, because)
That could be a reason to use == over isEqualTo
pliant stream
local _channel = (if (_topic call my_topic_isQuestion) then { "scripting-questions" } else { "scripting" }) call my_getChannel;
[_channel, (_topic call my_getMessage)] call my_channel_post;```But...but it's slightly slightly slower...
little eagle
It isn't anymore when you have to add isNull checks.
jade abyss
+Now with remoteExec func.
halcyon crypt
@outer scarab try to figure out if that bit of code is running at all with hint or diag_log
tough abyss
Notepad++ > Find and replace in files > == to isEqualTo > huge fps gain
peak plover
Yes
pliant stream
yes
halcyon crypt
no ๐
vapid frigate
no
pliant stream
but we don't have ternary op so
little eagle
Contrarians.
pliant stream
ยฏ_(ใ)_/ยฏ
halcyon crypt
poor man's ternary expression
outer scarab
Don't worry, I got it sorted @Marcel @little eagle
tough abyss
๐ฅ
outer scarab
had this too far up in my init // ======================================= // Ambient Sounds //enableEnvironment false; while {true} do { [player, "stormloop",500] call CBA_fnc_globalSay3d; sleep 15; //"0" represents what length the track is, so replace the 0 with your soundtrack's length };
vapid frigate
yeah if we could use _var = x ? true : false; i would
outer scarab
put it at the bottom, boom
peak plover
haha, it never get's past the loop ๐
jade abyss
I am still waiting for the remoteExec return ยฏ_(ใ)_/ยฏ
pliant stream
@vapid frigate: it wouldn't work, it would have to be SQF cond ? { true } : { false }
tough abyss
I would give it to you if I had my computer at the moment @jade abyss
vapid frigate
yeh probably
jade abyss
@tough abyss You working with Untelo?
pliant stream
@jade abyss you can read https://en.wikipedia.org/wiki/Futures_and_promises while you wait :D
pliant stream
and you can implement futures and promises in sqf
little eagle
local _channel = (if (_topic call my_topic_isQuestion) then { "scripting-questions" } else { "scripting" }) call my_getChannel;
local _channel = (["scripting", "scripting-questions"] select (_topic call my_topic_isQuestion)) call my_getChannel;
Where is your god now?
pliant stream
i think i would make the array static
only... the performance implications of having thousands of static arrays for shit like this worries me
simple solstice
local is same as private?
pliant stream
because global vars are not statically linked
little eagle
You do know that else also creates an array?
pliant stream
local is private in A2
jade abyss
little eagle
{ "scripting-questions" } else { "scripting" }
->
[{ "scripting-questions" }, { "scripting" }]
*StillWaitsForRemoteExecExampleForReturnValues*
vapid frigate
still prefer the if/then
little eagle
So no, there are no performance implications aside from it gets faster
Lecks, that is called irrationality, but Ok
pliant stream
man i'm implementing this future/promise right now... my old shittier implementation didn't support callbacks
jade abyss
I call it CommyNism
pliant stream
although it fairly easily could have...
little eagle
You calling me irrational, Dscha?
jade abyss
calling you a CommyNist
pliant stream
@jade abyss https://pastebin.com/raw/VQJV7P8m https://pastebin.com/raw/pjVih0kQ for the future/promise, it's got some missing dependencies but they're pretty self-explanatory, e.g. delegate is just code + context. also it's just an untested first version so might have typos or even design flaws, but it shows you the basic idea
little eagle
I can already tell that Dscha will like this.
jade abyss
I do, i rly do. (sarcasm)
halcyon crypt
I see 0 remoteExecs
jade abyss
Still searching for RemoteExec
pliant stream
that's obviously separate from the future/promise, which is just an abstraction for the result of an asynchronous operation
pliant stream
no
little eagle
๐ฟ
jade abyss
For me, it's just macroGibberish and i still can't see the return of (a) RemoteExec
pliant stream
_future = [...] call my_remoteExec;
_handler = { diag_log format ["ctx: %1 args: %2", Delegate_Context, Delegate_Args]; };
Future_Then(_future, Delegate_New(_handler, "hello"))```like i said, future/promise is an abstraction. next i'm writing the remoteExec code
tardy wagon
https://www.youtube.com/watch?v=BpT6MkCeP7Y
Arma3 would be beyond words if they just slapped in a near-photoreal raytrace engine
jade abyss
Has nothing to do with Scripting.
tardy wagon
vaguely
jade abyss
pliant stream
not even close lol
jade abyss
next i'm writing the remoteExec code
Haven't you said, it was already working?
halcyon crypt
@pliant stream are you currently somewhere in the academic world either as student or something else?
little eagle
I'm intrigued
pliant stream
@jade abyss it was working, but it didn't use the future/promise that i just wrote. also i've since redone the server backend (implemented in C++) so the SQF side is now incompatible, which is why i have to redo it anyway, and i decided to return futures instead of a future-like RPC object which i was returning previously
@halcyon crypt why do you ask?
little eagle
I like simple answers to simple questions.
pliant stream
you know what's simple to write? waitUntil on a global variable
you know what it isn't though? scalable... maintainable...
halcyon crypt
^ that's why I asked ๐
little eagle
_blah = [_code, _args] call my_remoteExec;
How fancy can you get solving this ^
?
pliant stream
i can get very fancy, and i will
little eagle
And I doubt that you're still working on something we would accept as solution to the given question.
pliant stream
_requests = [];
for "_i" from 1 to 1000 do
{
_requests pushBack ([_code, _i] call my_remoteExec);
};```this is a bit harder to do with just global vars and wait until isn't it
jade abyss
Mostly because of: remoteExec just doesn't return anything else, then Nil/JIP-ID or empty String
halcyon crypt
I still don't see a way around something looping somewhere ๐ค
jade abyss
ยฏ_(ใ)_/ยฏ
little eagle
There is none.
jade abyss
It might work in another Language, but in sqf.... I doubt it.
pliant stream
of course remoteExec doesn't return, and it can't, but it's possible to wrap around it an API which can
little eagle
Proof it.
jade abyss
Wich has some kind of suspension in it
little eagle
^
vapid frigate
yeah, like commy did earlier
pliant stream
no there is still no need for scheduled environment :D
vapid frigate
but not without waiting/polling/looping somewhere
pliant stream
yes there is suspension, in the same way that coroutines have suspension
but there is no need for the scheduled environment
little eagle
suspension outside the scheduled environment. ๐ค
pliant stream
do you know what a coroutine is
jade abyss
๐ฟ
little eagle
do you know what a coroutine is
No.
jade abyss
Once again: It might work in another Language, but in sqf.... I doubt it.
pliant stream
sqf is turing complete isn't it
little eagle
Idk what that means either.
pliant stream
ok
well soon as i've got my remoteExec code working i'll be sure to showcase it to you
little eagle
Did I mention that I like simple answers for simple questions?
pliant stream
it doesn't need macros, macros are just faster since we don't have a compiler that can inline
jade abyss
Don't wanna waste x minutes just to switch around and see what macro does what.
little eagle
Q: "How to get this code working?"
A?: "Do you know what 'Turing complete' is?"
halcyon crypt
from the Unity docs:
A coroutine is like a function that has the ability to pause execution and return control to Unity but then to continue where it left off on the following frame.
So it's basically just the scheduled environment in Arma ..... ๐คฆ
little eagle
Oh for fucks sake.
pliant stream
as for the macros, i just prefer to define public APIs using macros, it could just as well be functions
halcyon crypt
as if the underlying threading system matters at all for the implementation
pliant stream
like Future_GetValue(_future) -> _future call my_future_getValue
same thing
just written differently
arma sqf scheduler is a threading implementation
jade abyss
https://community.bistudio.com/wiki/Code_Optimisation#Make_it_work.
"Excessive cleverness is doing something in a really clever way when actually you could have done it in a much more straightforward but slightly less optimal manner. You've probably seen examples of people who construct amazing chains of macros (in C) or bizarre overloading patterns (in C++) which work fine but which you look at an go "wtf"? EC is a variation of premature-optimisation. It's also an act of hubris - programmers doing things because they want to show how clever they are rather than getting the job done." - sbsmac```I meant the implementation of whatever you're trying to proof here
jade abyss
My Fav is the "wtf"-part.
tardy wagon
brb implemented templated metaprogramming lambda thread schizz foo
jade abyss
This is what i see, at about 95% of all Macros i've seen so far.
little eagle
I think I'm laughing.
pliant stream
alright, then in those terms i shall prove that returning from remoteExec is possible using only one thread
jade abyss
without sleep/waituntil/similar stuff.
tardy wagon
If you like threads just write some compute shaders
vapid frigate
pls do
little eagle
_blah = [_code, _args] call my_remoteExec;
Reminder ^
halcyon crypt
lol
tardy wagon
for your insane programming benders
jade abyss
Yep. And pls remember: There were > 4 ppl who knows their shit in here, telling you: Not possible.
But if it works -> Kudos.
pliant stream
@little eagle i upgraded that to 1000 concurrent remoteExecs
little eagle
That was not the question.
vapid frigate
1 will do
pliant stream
fine
jade abyss
(and is bad anyway)
halcyon crypt
using PVEHs is cheating btw
vapid frigate
you don't need globals
little eagle
Well, I don't see how using PVEH could solve the issue that is:
_blah = [_code, _args] call my_remoteExec;
So it's not cheating, but unlikely to be the solution.
jade abyss
@halcyon crypt It still has to "wait" until the Var is set, wich can be delayed ๐
jade abyss
๐
little eagle
PVEH could be used as alternative to remoteExec, but that doesn't get around the need for suspension and therefore scheduled environment
pliant stream
A2 doesn't have remoteExec so i have to emulate it using PVs anyway...
halcyon crypt
I know that, what I meant with cheating was that PVEHs still rely on looping/polling/suspending etc behind the scenes
pliant stream
But it makes no difference
halcyon crypt
if he proves us wrong then I think he deserves a place in the veterans group ๐
you know, as a price
pliant stream
Lol
vapid frigate
or in biminions group
pliant stream
Whats the biminions group for btw
vapid frigate
employees
jade abyss
Slaves
pliant stream
Ok i thought so
halcyon crypt
if you lose though, you get X39's old group ๐
jade abyss
lmao, thats mean
pliant stream
Whats that
jade abyss
Don't ask.
pliant stream
Ok
halcyon crypt
๐
little eagle
Stop chatting, Untelo and post your code.
jade abyss
*CollectingSomePitchforks*
pliant stream
I'm walking my dog. I'll continue coding when i get back. Have some patience
little eagle
You will NOT get away with this.
vapid frigate
you don't need to waitUntil you get back
jade abyss
badumm tss
vapid frigate
can do it at the same time.. that's the whole point
pliant stream
I have attached a continuation to the asynchronous dog walking operation
jade abyss
So, a new Thread.
pliant stream
No the dog walking executes on the same thread. The continuation gets synchronized to it using a dispatcher.
little eagle
_blah = [_code, _args] call my_remoteExec;
@pliant stream so you create a proper async system for arma?
How would you realize the "reattach" to the previous scope? Unscheduled could break your neck here when you use a await function
At least with the method that would come in my mind right now
pliant stream
simple future/promise with continuations
of course coroutines would require compiler support
queen cargo
But how you want to solve that?
pliant stream
similar to C# Task<T> if you're more familiar with that than e.g. C++ std::future, std::promise
future/promise, or coroutines?
queen cargo
Coroutines
little eagle
Still waiting for the code.
pliant stream
oh i don't intend to, like i said that would require compiler support
in essence parsing something like await ( [...] call my_remoteExec ) and transforming the function into a coroutine
stackless coroutines, obviously
the locals would have to be extracted into a dynamic coroutine frame
queen cargo
So... You did not found the holy grale of sqf?
pliant stream
if you mean pretty coroutines, then no. said i said that would require compiler support
queen cargo
Y
dusk sage
Not sure if I regret reading the last few hours of chat or not
halcyon crypt
looking at std::future and std::promise it still needs to suspend/block using std::future::wait. In Arma that would either mean blocking the engine in unscheduled or using some kind of polling in scheduled ๐
dusk sage
TIL commy says 'mate' when getting triggered That is not the problem this is about, mate
pliant stream
you can however do ugly coroutines like i showed above, i.e. SQF Future_Then([...] call my_remoteExec, DelegateNew({ /*continuation*/ }, /*context*/));
little eagle
oh i don't intend to
๐
Was I being bullshited all along?
pliant stream
@halcyon crypt look instead at std::experimental::future, or boost::future
std::future::then is a C++17 feature
dusk sage
Finally too
queen cargo
Extremely ugly
Due to macro usage that also should be problematic for anything using newlines or commas
little eagle
TIL commy says 'mate' when getting triggered That is not the problem this is about, mate
Yo, mate
pliant stream
@little eagle did you misunderstood? in _blah = ... _blah was never the result, it was a future which will at some indeterminate time provide the result
queen cargo
Ugly in the sense of: you loose your current scope completely
dusk sage
๐คฃ
pliant stream
there's no way around that without stackful coroutines, which would require SQF support and that's obviously not happening
queen cargo
Leave him in his little world untelo
little eagle
wtf
pliant stream
and stackful coroutines aren't desirable anyway
actually i guess it wouldn't matter in sqf since it has a variable sized stack afaik
ยฏ_(ใ)_/ยฏ
queen cargo
Extension could do it
pliant stream
yeah, but it would be unstable
not unstable in that it would crash necessarily, but it would be broken by binary patches to the game
i think if you wanted pretty coroutines, the best way would be to just transform the extended sqf into valid ugly sqf with explicit continuations
but that requires tooling that doesn't exist
little eagle
So, Untelo, you're not gonna post a solution to:
_blah = [_code, _args] call my_remoteExec;
that does not require scheduled environment?
pliant stream
yes, but it's not my fault if you misunderstood what _blah is
little eagle
_blah is a local variable
little eagle
containing the return value of _code called with _args on a different machine
pliant stream
it is a future object that will at some indeterminate point provide the result of (_args call _code) executed on the server
dusk sage
The good old, at some point this may, or may not, have completed
pliant stream
that one
but all boilerplate is abstracted away, making it easier and safer to work with
dusk sage
I promise something might happen ๐
queen cargo
You abuse the reference "problem" of arrays?
little eagle
wtf is the point of the "future object" when you still have to suspend your code to read it's value?
queen cargo
Still on mobile and did not want to trigger commy2
halcyon crypt
I don't agree with your proposed solution to @little eagle's problem
pliant stream
@little eagle yes you have to suspend, but not by sleeping or using waitUntil
little eagle
Then how
pliant stream
you simply return, and then when the value is returned the continuation will execute
little eagle
how
vapid frigate
don't think you can, but if you could i'm pretty sure the game would freeze
little eagle
You can't, which was my point.
pliant stream
the client executes an eventhandler when the value is returned via PV or RE, the eventhandler finds the relevant promise and fulfills it, executing the continuation
pliant stream
yes it suspends, but it does not sleep or waitUntil
there is no polling
or any shit like that
vapid frigate
how does it suspend then?
pliant stream
it's completely event based
little eagle
_blah = [_code, _args] call my_remoteExec;
But it does not allow for this ^ then.
halcyon crypt
event systems still use polling at some point
pliant stream
@little eagle yes it does, but you misunderstand what _blah is
little eagle
No, _blah is the return value of _args call _code on a remote machine.
pliant stream
it is not. what you describe is impossible without freezing the client for an indeterminate amount of time
_blah is a future
queen cargo
_blah is a reference to the value that will be returned
It is not yet available but in a future point it will be
little eagle
Yes it is:
some_remoteExec_wrapper_that_has_the_ability_to_return_values = {
params ["_code", "_args"];
My_RE_Return = nil;
[[_code, _args, clientOwner], {
params ["_code", "_args", "_owner"];
private _return = _args call _code;
[_return, { My_RE_Return = _this; }] remoteExec ["BIS_fnc_call", _owner];
}] remoteExec ["BIS_fnc_call"];
waitUntil {!isNil "My_RE_Return"};
My_RE_Return
};
See?
You said this can be done in scheduled / without waitUntil
pliant stream
right, now you use waitUntil
little eagle
And no you can't.
queen cargo
You can get the actual value by using a yield function
little eagle
Shut up X39
pliant stream
like i said, if you want pretty coroutines you need compiler support
it's that simple
but you can have ugly coroutines without it
like i posted above
pliant stream
/*continuation*/
}, /*context*/));```I think we should try bargain for something better than SQF first
pliant stream
@little eagle by compiler support i don't mean changing SQF, the SQF "compiler" is essentially f(x) = x
little eagle
The question was what some_remoteExec_wrapper_that_has_the_ability_to_return_values should be to get the return value of a function called on another machine with given arguments.
I said this can only be done with waitUntil / in scheduled environment
And you said it can be done in unscheduled.
And you're wrong, because you still have not demonstrated that it can be.