#dev_rc_branch
1 messages ยท Page 1 of 1 (latest)
release date of 2.12?
Wouldn't you prefer to know 2.14 release date, instead? ๐คจ
yea, always forget that is +0.4 lol
Wait, we didn't even release 2.10 yet 
we are at 2.08 on release, 2.10 on dev, but when 2.10 is released its 2.10 or 2.12?
2.10 is the upcoming release, dev is currently at 2.11 as we have already branched 2.10
๐คฆโโ๏ธ
as we have already branched 2.10
@teal elk there's your answer, I guess ๐ (assuming you were asking about 2.10)
not the best of my mondays
This still holds true for now. Along with #dev_rc_branch message
that what i want to know before the brain fart lol, thank you
Hi I don't find a bug report channel, so I decided to post it here first.
There is a critical crash issue with Hands-Free mode Bluetooth headsets.
Once Hands-Free Bluetooth headsets are set as Arma 3's audio output device, and player tries to get in a helicopter, the game would instantly crash with ACCESS_VIOLATION 0xC0000005.
You can reproduce this with East Wind campaign first chapter.
The "bug report channel" is here: https://feedback.bistudio.com/project/view/1/
However, Bluetooth/USB audio devices crashing A3 is a well-known issue with the Windows 10 USB audio driver. The workaround, unfortunately, seems to be to just not use them. Your headset might have an audio jack you can use instead of wireless mode.
Or, you can probably try the solution that pinned in #arma3_troubleshooting
Okay, however this does not happen on other games. 
As Nikko says, Arma 3 has this issue, regardless other games
IIRC this was some driver issue with Win10, but not sure
mono devices seems to cause crashes, same with steelseries voip mode which is mono only
and if you're using mic on your bluetooth headset it will downgrade the audio profile to mono only
BT bandwidth โค๏ธ
Thats a quite old known issue, started appearing with windows 10
Maybe last M$ LCU fix that problem ๐ค Addresses an issue that affects certain games that use the XAudio API to play sound effects. https://support.microsoft.com/en-us/topic/july-26-2022-kb5015878-os-builds-19042-1865-19043-1865-and-19044-1865-preview-549f5551-fcc5-4fee-8811-c5df12e04d40
The weird thing is that it doesn't seem to be happening for other milsim games with voip...
Well I know 100% that microsoft has a bug in their bitrate conversion code on win10.
Maybe thats in that fix, they said they wanted to fix it.
Both A3 and DayZ were affected because in order to not hit that bug, you had to do something special that wasn't documented.
But we updated both A3 and DayZ to do that special thing in... 2.08 I think.. 2.10 has another audio fixes update maybe that changes anything.
With support for terrain editing in 3den with the Deformer mod with 2.10, would it be possible also add road editing?
but that's a limitation on the engine/game or the mod?
Deformer
Because like, if we can edit terrain, then I suppose we could edit roads?
no idea, that's why I'm asking
Don't know, try if they move
๐ต There's always gonna be another mountain
I'm always gonna wanna make it move๐ถ - Miley Cyrus
Update 2.10 RC-Branch is now available: https://twitter.com/ArmaPlatform/status/1555144511336112130
We've published a Release Candidate build of #Arma3 update 2.10 for players to try. One of the highlights is improved Thermal Imagery.
๐กDetails: https://t.co/eG7c67d0a7
๐ฅSteam access code: Arma3Update210RC
Ded('s) update eh?
hope fully not too many complaints about how bad the new TI is

New thermals look amazing 
In the preview image in the tweet it looks like the smoke column becomes invisible when it's in front of the sky ๐ค
Thats what I meant with how bad it is ๐
It is still there, but for some reason it fades out alot there, its alot harder to see
Yeah, I hope we can revert back to the old thermal rendering. I like how the current one is you have clear hot and cold spots. But the rest of the scene is kind of disorienting, hard to get a feel for space. That is more of a balance.
What do you mean by disorienting? Which part, exactly?
I....wonder if it's depth related.
I've noticed when taking screenshots that since smoke isn't included in the depth buffer, when DoF is on (or at least Reshade's depth-based DoF) it gets affected by the depth of whatever it's in front of - most notably the sky since it has infinite depth. This looks kind of similar and I wonder if some depth-based effect is being tripped up by whatever got changed in the sky's thermal handling.
Harder to get a sense of distance. Not as much immersion as you get from the visual view.
But I haven't tried this RC yet. Just judging by the twitter pic.
Well, I dunno, maybe I'm biased but I usually can't see anything in the old thermal view ๐คทโโ๏ธ
I literally have to switch it on and off to be able to move, because everything is... just gray
Especially now that I started to look at flir imagery on Youtube and it looks much more like the image on the right, so I guess that's how it should be rendered
You do need to avoid looking at hot (or cold depending white-hot/white-cold) objects. They change the dynamic range of your sensore. Looking with thermals and seeing parts of your hot tank, you don't see faint cold things in the distance.
Actually, in connection with this, it might be worth specifically checking the appearance of distant vehicles when viewed against the sky. I've had trouble with staging pictures of flying vehicles because they tend to disappear from the depth buffer when they're a couple of hundred metres away, which inflicts the same problem as I mentioned with smoke. If this is a depth thing as I speculated, this could make distant air vehicles very hard to spot on thermals.
Unfortunately I don't have RC or dev installed for logistical reasons. If someone who does could check the appearance of air vehicles against the sky at ~300m+, I think that could be a useful insight.
I don't know what military grade thermals look like. The current ones do work a bit like thermal cams available in retail market. (in grey-scale, not false color)
I mean... just look at the picture to the left, since you're talking about judging the twitter pic. You can barely see anything, besides that person in the center of the image (and those two vehicles, maybe could be called "visible", too)
You do need to avoid looking at hot (or cold depending white-hot/white-cold) objects. They change the dynamic range of your sensore.
And also I actually disagree with this sentence. I was having trouble fighting infantry in Tanoa's jungle. Barely could see anything, so the "hot engine which skews the dynamic range of the sensor" doesn't apply here
Anyway, looking forward to see how it looks in-game ๐
Maybe because MSAA swap chain is the problem in A3 and DayZ...look in ReShade 5.3 changelog ๐
The one change note related to DayZ doesn't appear to have any connection to the depth buffer, and in any case I'm pretty sure that particle effects not being in the depth buffer is a design choice on Arma's part, not a Reshade bug. I've noticed it because my use of Reshade emphasises it, but it's just how the game does its thing, it would still be the case even if I couldn't see it.
That is the BIS engine problem not from ReShade ๐ฏ
noticed the local object checkbox
I've been looking for a way to swap eden objects with local simple objects since it looks like the most performance efficient way to have decoration in MP.
could a checkbox for that be made aswell?
Hi, would you please give us the information on what ingame time this image was created?
no you can't
Mh that sounds plausible.
But I don't think I can fix that
just chekc the simple object and the local checkboxes
default time on livonia minus a few hours.
24th june about 10:00 with 100% overcast sky and raining
not every object has them
people argued they werent necessary but it absolutely does improve performance
ooohh right... yeah...
Well I don't want to talk about that, I have my opinion on it
ok?
didn't think it was a controversial topic
There is wiki page about simple objects. I donโt know how accurate it is, but seems like everything that not moving does not get so much of a boost from being simple.
been working with replacewithsimpleobject for a while
and I can feel the improvement
only problem is that its a mess for network traffic
hence why I've been looking for a local version of it
Yeah, I figured. Worth a shot though.
Do you have the time and will to check if that M$ update fix problem in A3 and DayZ ๐
I don't have it but for ppl which have it.... but....
well then they have to check
I only asked because you said Both A3 and DayZ were affected because in order to not hit that bug, you had to do something special that wasn't documented. and it's easy for you to determine if it is fix it or not ๐
no its not easy for me
I never had that issue
if I had and could reproduce it, I would've fixed it
Hmmm, I'm not able to get the new "Resolution" PP effect to work while in thermal but it works with nightvision and normal views.
(using the example code provided on the wiki)
["Resolution", 1, [400]] spawn
{
params ["_name", "_priority", "_effect", "_handle"];
while {
_handle = ppEffectCreate [_name, _priority];
_handle < 0;
} do {
_priority = _priority + 1;
};
_handle ppEffectEnable true;
_handle ppEffectAdjust _effect;
_handle ppEffectCommit 5;
waitUntil { ppEffectCommitted _handle };
systemChat "admire effect for a sec";
uiSleep 3;
_handle ppEffectEnable false;
ppEffectDestroy _handle;
};
Thermal has its own, it might take that from vehicle config only
Hey! People here is talking about an player custom action that dig a trench. So after 4 hours of game play the terrain will have lot's of trenchs!
Considering the trenchs are ok (nothing crazy) this can cause any problem?
Editing the terrain already is crazy enough ๐
So it's not to use it like that? ๐
I mean, you can expect some issues like AI doesn't recognize something or so
But it's always free to use
Ah
I will keep trenchs clean
Not crazy over a already crazy thing
Not too deep, not to large
Just for one or two players to hide
in knee position
not too slope
not on reoads, not near objects, will remove grass
๐
it will still be too deep/large
also keep in mind that the terrain is composed of triangles. to "dig a trench" you have to move a triangle vertex lower. this will affect adjacent triangles
now if the cell size is large, say on Altis, which is 7.5 m, you can only dig a trench that is ~15m wide
I understand. Thanks. May be i will keep edition just for missions placement.
But will do some trenchs test.
๐
low hundreds of terrain edits is probably fine.
Noone tested this in large scale so... Congratulations, you're a scientist now
๐ฅธ
Not sure if intended but with the new thermal, the the equipment both Viper and CTRG use stands really far out verses their anti-thermal suits/helmets
With the newest thermal update
The TI improvements specifically target some highlighted criticisms like contrast, the skyline, and noise / resolution (the latter mostly for mods to use if they wish). There are risks to doing this work now, since the TI runs over all data and some data may not be balanced well for the new values. We don't have the resources to tweak the whole game now, so best we can do is try to target edge cases and hope that overall this is an improvement. There's a possibility some scenarios work less well than before, so the question is whether the balance is favorable on the whole. We also realize military-grade TI can do some fancy things by now, let alone in fictional 2035+. But there we have to again be careful about balance within the missions and ORBAT. Best way to help now is to give it a go in the RC and report problematic cases (especially new compared to 2.08 TI), like above with Viper / CTRG. We'll see what's possible 
Thank you, it's fanastic improvement as well.
Is there any documented process on changing towns/places names? Or is that not possible for 2.10?
And what consists a terrain edit? the amont of cells edited? For example changing the height for a cell from 10 to 999 has the same impact as changing it from 10 to 11?
In 2.10, a map-native location can be converted into a [partly] editable location by using the alt syntax of https://community.bistudio.com/wiki/createLocation on it. Then it can be affected by some commands in the same way as a scripted location, including changing the name with setName.
thanks
PSA: Dev-Branch will not see an update this week due to "logistical issues".
Ah, good ol' logistical issues, I missed you ๐
Sun Tzu quote fits right here ๐
The TI updates reminded me of a reddit thread a while back regarding star luminance under night vision - A BI dev jumped in and said that apparently star brightness was turned down accidentally under night vision, and that they fixed it for the next dev branch.
I'll see if I can find the thread, but I was really hoping it'd be part of this RC
here we go - https://www.reddit.com/r/arma/comments/ue1nho/for_what_purpose_doesnt_the_nightvision_show/i6n7edo/
RC was branched after that change; are you sure it's not in (it's expected to be)?
I'm just going off of dev logs, haven't had time to actually download the new steam branch yet, but I'll check if I see any difference when I do
That was fixed in the 2.08 perf/profiling branch months ago, so it's almost certainly in 2.10
Check the comments on Imgur that mentions itโs coming in 2.10. Also that โdeveloperโ dedmen is in here
Ye that's definitely in, whether the change is strong enough for you i can't tell though, we went for a middle ground that seemed good to us
The RC-Branch has received a small update with a few fixes and translations. The actual release window should open next week.
lovely thank you
so what is the right way to initialize missionProfileNamespace ... the file is not created until the first "saveMissionProfileNamespace;" is called
should we just put "saveMissionProfileNamespace;" at the top of the init file
if (!isMissionProfileNamespaceLoaded) then {
saveMissionProfileNamespace;
};```
Why would you want to create file for no reason
how is starVisibilityCoefNV supposed to work? Set to 5 should be 5%, right...?
well we have to create the file initially. biki says we need to execute "savemissionprofilenamespace" to create the file to save the vars
A lot of coef values are based on 1 being 100% so the scale is 0-1, I'd take a look in the config viewer for the inherited starVisibilityCoefNV on some of the terrains to see the current defaults
They don't come up with the config viewer nor scripting commands for some reason
Always returns 0 -- but I'll try using 0.05 instead then
even if you are outside the range I'd imagine the value is clamped to min/max expected values
Yeah as all stars look the same size
I will ask again, what is the point to create a file if you donโt have anything to save in it?
Well, the profile namespace saves even if you don't call saveProfileNamespace, right.
Documentation kinda implies that the mission profile namespace is different.
Either way it seems the sensible approach is to call the save command once after writing your save data, although I've had to rewrite code that had a very different idea.
the sensible approach is to call the save command once after writing your save data
that would be slow
it should be saved periodically. but I think the game also saves the namespace when you end the mission
is the "Local Only" attributes feature a 2.10 feature...?
peak, will come in useful for my excessive amount of props then
Nice, how many props? 200?
More in the range of 6000
And thats only half done
๐ฎ
using platforms to make an entire floor aint it ๐
turns out 1500 cargo platforms bloats the mission size to 10MB
just use the scaling script
Muh collisions
Most of those 6000 objects are already Simple Objects?
im using the local only ting and disabling simulation
local only and simple object has no benefit from what i can tell?
Deponds on their simulation
Objects with simulation="house" don't really benefit from changing to simple objects
It's mostly only effective with PhysX objects
So I can have a non simple object be local only..?
idk about the 3den option, but in general yes (and it would be less network intensive)
Less network intensive than a simple object being local only?
no less than a global object
as long as it's local it has no network cost. doesn't matter whether it's simple or not
gotchu i understand now
for a local object simple/disable simulation only save client performance
so i shouldnt be bothering with calling setsimpleobject and stuff in the init should i
for either client or network performance if its local ^
depends on the simulation kind like I said before
you can find the simulation of an object in config viewer (right click -> find in config viewer)
simulation = "house"; but has no doors either
but seeing as ive got so many even if its a marginal improvement its probably worth doing it
I think I'm at 9k platforms now ๐
and that needs to double but 3den isnt exactly in love with me
@wise dragon how does it perform in multiplayer
anyone going to be playing this 9k platformer or is it for ... personal consumption...
i have no idea
i get over 80fps most of the time though which in surprised about
It only occured to me last night that I can setobjectscale with remoteexec or something and really not need to have so many tbf
-- my maths of how many platforms it is might be wrong too, i cant actually check properly as selecting them all crashes my game
Count it using nearestObjects or something 
I'll do that once I'm home
I just know i have over 2k layers
I think its like 2.4k layers for 200x200 meter tiles made of 4x4 platform objects from a mod
Top layer is about 4km wide so I think my maths isn't too far off seeing as theres another bottom layer too
There's got to be some bigger objects you can use to cover more ground. Using 4x4m objects to cover a 4x4km area is absurd
50x50m objects
But 4x4 of them to make a larger "tile"
Can't find any bigger than 50x50 because collisions don't like it apparently
...but i might try mess with objectscale if it performs awfully in mp
If I've done this correctly then the player is the 16591th object
wait i forgot count exists
15814 objects
will need to mp test
That counts all objects including terrain
its on vr terrain
No further issues on RC branch my dudes ๐ธ?
Im having issues getting the nv star coefficient thingy through debug console but its not a huge issue
I can get more deets in half an hour or so but it always returns 0 no matter the value
You said you made it into a normal config entry instead of whatever weird hack it was before, so it should be possible to retrieve it through the debug console like any other config entry, e.g.
getNumber(configFile >> "CfgWorlds" >> worldname >> "Lighting" >> "starVisibilityCoefNV")```
oh hold on its working when i overwrite it
but when its not overwritten (i.e just the default one) it returns 0
As you can see, not the default variable
But if I go onto stratis (which is one I don't overwrite with the mod I did) and try it
Then it returns 0
Changing worldname to CAWorld returns 0 as well
I know it's not 0 because I tried 0 in the mod and it just hides all stars in nv
ditto with DefaultWorld
i added the option for it to be overwritten by a config entry, but we didn't add any config entry
Not an issue, more a feature request for the next release:
I tried to start a HC with -par parameter file (at least for mod=), but it doesn't work. This would be great if this would work.
Can you show your startup line and the contents of the -par file? As it works fine for me.
Someone reported that footsteps and tire marks (#mark and #track) get buggy and won't delete after some time:
#arma3_scripting message
Tho I don't think it's new
terrain edit will be complatible with multiplayer? 
Do you have tested it on linux 32bit?
nah, 64bit ๐ , let me test it though...
But if its not working on current stable it's not really a problem with RC, but it should be fixed nonetheless.
Yea works fine on 32bit, stable and rc.
mod= or -mod=? ๐
That's why I asked for the content :P.
But should probably move to #server_admins as this isn't a RC issue.
it's even crossposted from #headless_client
ah didn't see it there... ๐ we can continue there.
there is still an issue with the damage model of CSAT tracked APCs (Kamysh and Tigris)
RC
they do not show damage
Another "issue" -- local only objects show up in zeus. unsure if intentional or not but seeing as its purpose on the tooltip is literally to optimise large scenarios im not 100% sure thats a good idea imo
why wouldn't it show if its local to the zeus? ๐ค
should it just disappear instead?
there is a difference between being visible and editable
I think they're talking about it being editable
AFAIK, local only objects are meant to reduce network traffic by not syncing the state of the object. Doesn't indicate anything about it being editable, visible, or any other property.
if you're running zeus mods that makes all objects editable then that is up to you and your chosen mods?
or if the mission maker made the object as editable
I don't think it's a good idea to make local-only objects non-editable. I wouldn't be surprised if there are mods out there which use local-only objects for things Zeus might need to edit, and I'm very dubious that a distinction could be made between Editor-placed scenery objects and script-spawned ones used by those mods.
any object is non editable unless you make
I just tried in eden editor with a local only object, it is not editable unless you make it so
just like any other object
Yeah it's able to be made editable is what I meant
But seeing as the intent for local only is for large scale scenarios I think having that may not be too bueno ๐คท
then why did you make it editable? ๐ค
either the mission maker set it as editable, intended or not
because i might want to edit objects on top of them without having like 50 local only objects coming up as well when i set it as editable ๐
or if its a mod issue, tell the mod author to update the mod into consider local objects before making it editable
also seeing as theyre not synced on the server surely nothing the zeus does would influence other clients..?
or am i being dumb in that regard
if you don't want it as editable, don't make it editable?
It wouldn't affect other clients but that doesn't mean the Zeus will never want to interact with a local-only object for their own ends
Local-only objects aren't only used by the new Editor option. createVehicleLocal has been around for a long time and is used for lots of things, some of which a Zeus may very well be interested in touching. A blanket change to how things already work is not a good idea.
out of the box any eden editor placed local objects won't be editable, so that's inline with what you're requesting?
i cant even remember what the editable objects interface looks like one second
you either add it via sqf or the Add Editable Objects module
ye that module
then you've specifically said to make it editable, that's on you
that's a valid option, it's not like that by default but if mission maker needs it the option is there
one moment let me check what the module ui looks like
you can limit zeus entities and you can select if crew is included
that's it
you sync it to the object you want to be editable
is it the one with the check marks for what you include because if so a local object checkmark would be more in line with what im asking
wait is it not a vanilla zeus feature within zeus? 
you sync objects that should be editable, the blue line
I think you're talking about some zeus mod, if so then you would need to talk to the mod author
serves me right for not playing vanilla for so long then
literally forgot it was a modded feature ๐
ignore me then
Vanilla Zeus is designed with the idea that the missionmaker can and will restrict what Zeus can edit. If they could just place a module to bypass it that would sort of undermine that idea.
Zeus mods on the other hand don't care at all about that (which has caused me problems when I actually do want to restrict things :U )
Yeah thats on me ignore me I'll figure out which mod adds it and go hassle them once 2.10 drops
Just placing down a zeus module and having like 50 platforms come up that are important to the players survival wasnt too useful
no, by design.
yeah its my bad i thought update editable objects was a vanilla module within zeus prior to editor lol
Triggers don't have the option for "Local Only" right? They are either global or server (when placed on the Eden Editor), because scripted triggers do have a command to make it local I believe...
ignore that i was being dumb again
although when local only appears seems to be somewhat inconsistent -- does it come up when simulation is enabled when theres no performance change from it being disabled or something?
for example its coming up on helipad decals while simulation is enabled however not houses etc while simulation is enabled
Are you serious?
Oh in 3DEN
Yes, that is where triggers are used the most I believe ๐
Because houses with enabled simulation have opening doors that are network synced
Okay thought so was just checking
was the RC branch just removed? I was updating my units modpack and switched back to profiling branch, then wanted to continue to do something with 2.10 yet steam does not have the branch anymore and the access code is now invalid
Looks like 2.10 has just been released

๐
i manage to hit the dumbest timings
Nice
Tweaked: Header text of Launcher HTML preset files in order to better explain how to import a preset
What a changelog. Feels like back in Arma 3 Alpha ๐
Let's hope the same for v2.12 
yes 2.10 was enough to get me tinkering again
that was a short RC .. hopefully those bugs got fixed
Yush congrats 
2.10 gets released
ok but what about 2.12?

I only use the dev branch so I don't care about stable releases 
By v2.12 I meant further development
You know what, majority of them don't know what it is
Is the FT for the Taru pods correct? It's the same as the stacked event handler FT
Fixed: Init script error for Taru pods - FT-T164036
Fixed: Error when stacked Event Handler code was too large - FT-T164036
Probably should be this one https://feedback.bistudio.com/T164793
RC time was small, i was about to install it today! ๐ฌ
3 or 4 weeks as every time
Ah ok, i was too late so.
I have an agent of typeOf "C_man_p_shorts_1_F_afro" near me, if i dosqf nearestObjects [player,["C_man_p_shorts_1_F_afro"],10]i receive a []. If i do a:sqf nearestObjects [player,["CaManBase"],10]i receive an array with me and the agent.
Arma 3 2.10
Please make a ticket
I fixed by changing from nearestObjects to nearEntities.
When i tried to subscribe to the ticket system i got an error, long time ago. I will try again.
Also regarding nearestObjects, can #xxx objects be supported as well?
I used nearestObjects a while ago (2.08) to detect #Crater objects and it worked. Not sure about 2.10, still not tested.
Really? 
I'll have to test this again later but iirc it didn't work when I tried it
I even asked for help on #arma3_scripting on how to delete #Crater objects.
After finding them with nearestObjects.
This was yesterday, i believe.
Would it be possible to sync local objects with event handlers? Like barbwire that either has an alive state or a dead state?
if you give it a variable name for example so you can find it, yes
anyone tried the scripted move script so far? the compiled function looks completely different than the plain sqf file, old sqfc file?
nvm i had the old profiling branch companion mod loaded...
I'm using setTerrainHeight in one vertex only. The function is ultra fast in some places but in others it generate a general freeze. Same number of objects above the modified part of the terrain. What's make it slow on some parts of the terrain and fast in others? I want to know that to avoid using it where it will generate a general freeze.
Map is Malden 2035
scripting request: would be great if doArtilleryFire could return whether the command was executed successfully or not... while there are some artillery commands that can check if the weapon solution is in range etc, the AI appears very sensitive so factors such as slope. in the latter case, the command just fails silently and there's no way to understand why it is failing.
But it does fail for fairly minor slopes such as https://imgur.com/DtNn6ZV
(gamelogic on the left is in the foreground, and marks the location from which the artillery is fired)
(map is altis, artillery unit in question is BM-21 from Global Mobilization)
unitReady goes from true to false to true within a split second. When command is executed normally unitReady stays false while artillery is firing.
Artillery fired normally for all other waypoints using the same script so nothing else should be causing this.
The only bottleneck I had was object movement but most of that was fixed ๐ค
I didn't see anything else while testing
Malden in general has some stuttering issues that other maps don't (CDLC Maps) in my experience.
everybody seems to have issues using it on malden
I make many small setTerrainHeight for a digging "fake animation". On some places it works and is a lot fluid, on others i have a massive freeze while the change is running. I removed the animation and did a "one time" terrain change.
Now can't detect any freeze.
Yes, Malden is strange. Many say me "Use Malden for better FPS, Malden is small". But Altis have more fps!
Malden have a 12.5 terrain grid size, Altis have a 7.5 terrain grid size ๐ฎ
In Altis i have no freeze.
Livonia has some issues too
Indeed, that one is the bigger offender but it is not used much for workshop missions unlike malden
huh interesting, i don't use workshop missions
for our community missions we do use livonia a fair bit, which is why we notice, i guess ๐
(malden too, i love that map for urban/semi-urban infantry missions)
Malden is great! I wish it had Altis' performance...
Some time ago there were reports that big stones cause game freeze when you move around them. I donโt remember was it fixed or not. Malden have a lot of stones.
was still a problem in 2.08, another person in our group mentioned it only a week or so ago
I am pretty impressed by the patch but I am even more impressed by its repercussions:
Many of the most prominent modders and mission makers back on their projects, resuming work on their mods, frameworks, missions and content.
The reinvigoration of the whole community is almost palpable
There is still so much life left on Arma 3, it is mind blowing really
Added: Use of setObjectTexture(Global) to override a clan selection texture
does this work in both SP and MP? will it be permanent? (normally in a vehicle the clan selection texture changes depending on the clan of the commander/driver.)
also does this new feature apply to hiddenSelectionsTextures token in configs too?
Here is the ticket: https://feedback.bistudio.com/T167313
I had only one small problem, but on more complex missions i believe this can be a break cause.
Is this broken in 2.10 but was fine in 2.08?
Checked on 2.08 legacy. Broke in 2.10 for all cases, agent and non-agent.
No problem in 2.08.
sounds like a hotfix candidate
The global versions of the command work in MP. It is also "permanent" as in it will not get replaced by the clan texture of any person that enters the vehicle. When setting the clan vehicle texture of the vehicle it will go through the following steps:
- Check for setObjectTexture(global) clan texture
- Check for the clan texture of the driver/commander
- Nothing displayed
You can still revert it by passing an empty string, at which point it acts as before as it doesn't have anything set in the first step.
What do you mean by the "hiddenSelectionsTextures token"?
ok very nice, so this means we can also assign a texture to the clan selection in SP right? for example in nato combat uniform, the left pocket is used for clan patch in MP, but in SP its always empty
Yea, that should work
i mean the initial textures for a unit is set by the config, setobjecttexture basically modifies the config's texture via script yes?
now is it possible to assign a clan selection texture via the config?
example of config from B_Soldier_05_f:
hiddenSelections[] = {"Camo1", "Camo2", "insignia"};
hiddenSelectionsTextures[] = {"\A3\Characters_F\Common\Data\pilot_suit_nato_co.paa", "\A3\Characters_F\Common\Data\pilot_helmet_nato_co.paa"};
can i add a "clan" to the hiddenselections token and assign a permanent texture?
No, that is not supported.

When you is on parachute, in third person, the shores flicker a lot. When you change to first person the flicker stops. This have something to do with Z-buffer, right? But it's a lot ugly. This can be "fixed"?
sounds like config problem @rustic plaza ?
doesn't sound like that. Might be z buffer thing as suggested
I think this would handle most of the complaints (besides it not being OP anymore)
Very nice ๐
I was actually gonna suggest make it a bit grayer but a command is better
But now what about people who don't know scripting? ๐ฌ
New default will be 10-90 instead of 0-100
I'm leaving this open to add more parameters in the future, like maybe foliage brightness range
If start is 1 and end is 0 will it become black hot?!
no
Oh wait it was width not end 
its not end, its.. yea
Negative width then 
but.. could make the width negative 
But I blocked it, 0-1 only. Maybe can get messy later but this is for hotfix
Hmm a hotfix for a thermal vision? That's hawt
mission + server param?
Nice, trunks will glow or not glow based on sum, day time, geo location, etc, or they allways glow?
Let's face it, we were always going to hotfix the TI update, no matter if bugs were found โค๏ธโ๐ฅ
Is that difference between Viper (left) and CTRG (right) backpack intended?
https://cdn.discordapp.com/attachments/807428199284998184/1014222051780866149/unknown.png
The viper is anti-thermal?
The CTRG stealth uniform is definitely anti-thermal, so this must be a WHOT image, which means the Viper backpack shown here is definitely not anti-thermal
Yes viper is anti thermal, just like CTRG. But the viper harness isn't unlike the CTRG backpack which is also anti-thermal
https://feedback.bistudio.com/T160334 will this bug ever be fixed? super annoying bug been around for many months
Sorry, bug? Where is bug?
Am I correct to understand that if you are subscribed to some composition on Steam and the owner removes it you get error?
Let me guess, so players can't unsubscribe yet the error happens?
What is the proposed solution?
Call Gabe so he can chase the offender and make them put compo back?
I would assume they're after some way to clean up subscribed-but-removed compositions in order to get rid of the error message. From the comments, it seems like a subscribed-but-removed composition generates an error every time you load the Editor, and since it's removed you can't unsubscribe to remove the error. So they're looking for a way to force Arma to see it as unsubscribed and stop trying to load it.
simply not show the error? just have it show in rpt
or show them in a way that doesnt force players to click close everytime they start the game
simply not show the error? just have it show in rpt
inb4 "BIS Where is my composition!!11!!!!"
just tell them that if they ask
We are going to set hotline
Correct me if i am wrong but there is not all functions from the ai command menu available as script command, right? For example healing an ai by an ai medic is only availble over the command menu. Would it be possible to become these as script command?
@full sonnet I have a question, wasn't that you a few months ago demonstrating some changes to helis that removed their scripted explosion damage?
If it was you I was wondering if those changes were ever going to hit a dev branch at the very least
Hmm the 2.10 group eventhandler "enemydetected" does just fire every 5 sec and loses its point a bit. Maybe the knowsabout one is more reliable when it comes to handling detection of players by ai and for example reporting them to other units.
Any opinion on this?
And how is that eventhandler made actually? I'd like to understand the limitations.
They should be on dev branch
"And how is that eventhandler made actually?"
What do you mean?
When the in-engine enemy detected code runs, the eventhandler fires
I think he is wondering why it is getting triggered every 5 seconds and how reliable it is. And how high does the knows about need to be before the event fires?
Ok so it fires every 5 sec then
Does it trigger when there are no enemies? The code that detects enemies might trigger every 5 seconds and if an enemy is still detected that event will fire. Because we don't really have the opposite of the event.
It does not trigger when there are no enemy, no, but it triggers constantly if there is one. Makes it not really usable, since I would have to check if that target has alrdy been reported and so on. More helpful would be an "enemyDetected" and an "lostTrackOfEnemy" type event. But I can work with what is there. Its just not ideal, its Arma, how would one complain?^^
@winter crescent Found out that you're missing EpeContact updates in rev 149877 in the latest changelog
ref: https://feedback.bistudio.com/T167360
Splendid catch! 4 entries went AWOL, but they've just been rounded up and published to the log.
Does uisleep depend on the fps of the game? I noticed a feature after the update that my uiSleep became slower at low fps
You're limited to 3ms of scripts per frame, so if you're spamming them hard then both uiSleep and sleep will last longer.
As i understand, if your FPS is 5, a uiSleep of 0.001 will wait for 0.2 seconds.
At least 0.2 seconds.
mh..
Direct comparison, but the new version has contrast 0.1-0.3 instead of default 0.1-0.9, to match the actual contrast a bit closer
huge fps improvement
Which one is improved left?
The one that doesnโt say 2.08
Fair enough, looks good
Would the arma engine allow stabilized turrets to be turned relative to the vehicle and not relative to the terrain as it is the case currently?
This is true right? updating a workshop item will make it become invisible if latest EULA isn't accepted
Yes, it's true.
@little granite you tested uploading contributor items <#arma3_tools message>
on A3 and it worked right?
yes you can
but it may trigger above (item to get hidden until uploader confirms latest EULA)
Doesn't work for me in publisher. I get the infamous
ErrorCode 1
ErrorMessage "STEAM_API_DLL TODO"
SteamErrorCode 2
Thank you valve for making your error messages be a TODO...
lol, you can actually just use publisher via command line
And that works?
Why is mine refusing to do it :sad:
did you accept contributor status

that method works with any game with workshop content
but i later discovered publisher lets you enter any workshop id
if u use cmd line
Fun that steam shows me that I am a contributor on an item, even though I never accepted.
But it doesn't let me upload to it :3
Well now the UI will also show it
is that on like "arma 3 tools dev branch"
i dont see it on my publisher i assume thats what you're working on?
lol i feel that
a parameter is incorrect, yes, yes thank you
:)
"werks on my machine" (c) gaben
It's possible I'm losing my mind, but did the "Scenario saved" chime in the Eden editor get noticeably louder for some reason?
Yeah, it it's really loud since 2.10
Sounds like it's playing few sounds at once or smth.
Haven't noticed that, any mods running?
2.10 added playSoundUI which isn't affected by the volume sliders. If the save sound was previously on a channel you had turned down, and is now on the UI channel, that might make it sound louder
we will probably add a slider in options
just needed a bit of baseball bat.
Cannot change Visibility or Preview Image on contributor items, but all else works
Might be that, I have all my sliders around 30% so if it ignores them it might be the reason.
Current volume of the save sound hurts my ears. ;/
Hi Guys. Anybody know about the new start param " SetThreadCharatistics"? I've searched everywhere and can't find any document / introduction of how to use that .
It's intended to convince Windows to use the P-cores instead of the E-cores on hybrid CPUs. I don't know whether it actually helps and I can't find the win32 API reference for it anymore.
Maybe Microsoft killed it :P
there is some mentions in #perf_prof_branch #perf_prof_branch message
in terms of arma you can set this flag in Startup Parameters. As much as I remember if this option is set arma will tell MS Windows that this process is a game and Windows scheduler will try to optimize task schedule to get some performance. But on some Windows releases it can cause hard OS lockups. So it is disabled by default but available as an option.
Not only for Intel 12gen.
Hmm! I can get performance boost in a 10th generation intel processor, with SetThreadCharatistics?
Can use on server, HC and clients?
you can test and see
I am encountering a problem using missionProfileNameSpace: The variables are not being stored persistently, when reloading the mission, the saved vars are gone. Is that intended? ProfileNamespace works though.
Edit: They are stored temporarily, like in missionNameSpace
confused with a few things
isProfileNamespaceLoaded is not listed on the BIKI - could be just a docu thing
more importantly your pastebin shows profileNamespace and not missionNamespace
@dusty scarab
Yes i did a typo there ;/ sry have been tryin hard for too long, ofc I tested with missionProfileNameSpace
it works now
I actually pasebinned the wrong code ^^
there is no isProfileNamespaceLoaded
I typed wrong in the post.
Why the file is not closed after I compiled it via compileScript or preprocessFile ?
Ok, so the problem is still present:
I store variables in missionProfileNameSpace using the missionEventhandler "MPEnded".
The event seems to trigger on #restart but not on #restartServer, also ending a mission with zeus does not trigger it. I do not understand. How does it work?
I read the vars individually by a script that is run from the postInit of an editor placed module. The storing of variables in missionProfileNameSpace seems to be very wonky. Maybe it does not work at all when executed several times within short period?
I think using missionProfileNameSpace to save stuff is not reliable. The command seems to be broken.
Any experience on this?
@dusty scarab make a A3 FT ticket with specific repro steps. will increase chances KK to look into it a great deal
Right. Will do.
I need a 100% repro, there is a little chance you are doing something strange, but also possible that there is a bug
I'm having general missions problems because this don't return no-class objects:sqf nearestObjects [_pos,[],_rad,true]Any way to fix that?
well, there's nearestTerrainObjects
Thanks.

@full sonnet sorry, server was not using 2.10 second hotfix.
I believe now it's fixed, still need to test.
which version?
Oh wait you are on stable. Why would you post on dev channel?
I assumed it was #scripting when I answered :P
didn't see the getTerrainInfo before, big thanks for this one ๐
hey guys, I'm looking for a scripting command to get the player camera position (visual), currently the only command to get true eye position (including head tracking offsets) is with positionCameraToWorld which is simulation and so doesn't work in vehicles.
There also doesn't appear to be any way to pull current 6dof headtracking offsets independently
Simulation vs visual doesn't matter for player, as they are local
Sorry I don't quite follow, I probably butchered the terminology in my message, what I mean is I need to get a render time scope version of positionCameraToWorld that doesn't lag behind the players camera view while flying.
Every other render time scope position I pull that I need for cockpit coding looks smooth to the player, but positionCameraToWorld appears to be updating at the simulation speed
In the hopes of finding a memory point/selection position I could pull in render time scope to grab the head tracking offset I wrote a script that checks for any world position change in every single selection position for every available LOD (and i mean every LOD, even the obscure ones) - the result: not a single player or vehicle player LOD selection position is changed due to headtracking movement (translation). Very sad times.
I'm gonna look into animations and see if maybe there is some animationPhase for the player or vehicle player for headtracking movement but I dont have high hopes ๐ฆ
I see now what you mean, but positionCameratoWorld definately lags waay behind modelToWorldVisual positions also running in Render, which I dont get bc the players view is perfectly in sync with these Render positions
Thats super cool
Thanks! Would be great if I could add 6dof headtracking support but in having struggles
You can use https://github.com/opentrack/opentrack/releases
That's pretty awesome....
Do you think that technology could work for standard NVGs?
So as opposed to blacking out around the side of the display, have an NVG overlay where the tubes are and 'standard' vision outside that?
Thanks but as mentioned before the issue is that I need to be able to find the players head tracking offset in-game in order to position the outter pip camera correctly. Unfortunately the only position I can get that does this is positionCameraToWorld and this doesn't update at the same frequency as render time scope so the camera lags behind in vehicles
I've tried this but as pip rendering is only the external lod this causes issues when used in vehicles, you can see from that clip that the outter pip peripherals show the low resolution cockpit as seen from the outside
When I say offset I don't mean the rotation I mean when the player physically moves their head left/right, up/down, forwards/back
Have you tried taking a look at how ACRE or TFAR handles this? https://github.com/IDI-Systems/acre2/blob/master/addons/sys_core/fnc_updateSelf.sqf
and then you can use CBA_fnc_vect2Polar on the normalized vector between two positionCameraToWorld calls (one at 0,0,0 and the other at 0,0,100) to get the pitch
Thank man, good ideas, I'll take a look at these
I guessed theoretically it would be possible to do some crazy trig off a visual/render position to back locate the origin haha
looks like the just use positionCameraToWorld which has the lagging issue, but I guess doesn't cause a problem for audio alone
They're only using it for the direction aspect, positioning is handled by modelToWorldVisualWorld which should solve the lagging issue.
It would be nice to have a local command for addMagazine just like addItemCargo. So I can add Magazines with a ammo count to a local storage.
Just another idea, Put Eventhandler should also give the container object back, if he drops one. Like for example Vest into cargo, would give back the ContainerObject from the Vest.
you can probably find a local position and then use modelToWorldVisual
and then run it inside a draw eventhandler
that's how we had to solve the cockpit interaction checks
I ran code to check but Unfortunately not a single memory point on the player or vehicle player takes into account head tracking translational movement
Though now I'm curious did you find a render scope position in cockpit that includes headtracking for pressing buttons?
I think this is probably the wrong channel to go in to depth for this discussion, but we can continue in DMs
Hiya, we've had other priorities these past 2 weeks, so there was no Dev-Branch update. There will also not be one today for lack of changes. Don't worry, they will return 
It is great to hear that. I hope our dear devs return soon, stronger and infused with renewed BI motivation ๐
You mean ๐ค ๐ค ๐ค motivation? ๐คญ
Well, that too... but what is it worth some extra ๐ฐ without a lease for devs to do their own bidding without restrictions?
There are always going to be restrictions I guess... but being given creative freedom is always welcome as long as the pros surpass the cons ๐
I'm more interested in knowing what other priorities are these
Just pure expeculations but I imagine it could be due to several factors like: 1) the DayZ Halloween patch/event/discount requiring extra manpower. 2) a Reforger emergency rally... 3) A company wide restructuring which involved some reshuffling/brainstorming on all corporate levels. 4) Massive hiring and capacitation of new staff that required all "veteran devs" as tutors/teachers. 5) A combination all the previously mentioned plus some additional spanking/commendations/promotions ๐
Ohhh and Tencent's international shares dropping like flies could have sparked some red flags and some changes too, thou not likely due to how isolated BI seems to be from Tencents affairs: Tencent Holdings Ltd -69.20 (-24.64%) (Last 30 days shares value)
Did not know about that... but just found the related official post: https://www.bohemia.net/blog/bohemia-interactive-minority-investment-tencet
So many other options that you haven't even considered. Then again, this speculation is not needed and does no one any good.
Indeed Sir, my apologies and I will refrain from dwelling further into speculations in the future
Ohhh God please no! Meta is so bad! Can we keep that "new device" a Steam Deck perhaps or something cooler like that ๐
If you weren't speculating you would have said that the priority was "Reforger Workshop" related without giving any further detail into the matter
hope: Discussing PBR for Arma 3
Reality: Some reshuffling of priorities for reforger/Arma 4
There's not even a need to speculate. We've outlined what we're doing in the latest SITREP (Operations section) ๐
https://dev.arma3.com/post/sitrep-00241
It's a combination of our new Amsterdam Enfusion project + ongoing Arma Reforger support.
I completely forgot it was mentioned in there.
An incredibly insightful read. Recommended to all arma community members who dwell here. My appreciation for the whole BI team was already renewed by the time I was halfway though that report
Is there anything actually going on right now in the RC branch
Since no RC branch is on right now, no
Wondering again about the possibility of a positionCameraToWorldVisual command; from further testing it would indeed seem as though PositionCameraToWorld is run in simulation and not render scope. PositionCameraToWorld is the only command in the game engine currently that takes into account headtracking movement in addition to player's position, and so currently there is no way to get an accurate player camera position that doesn't lag behind moving vehicles
Direction is the important part here though I think?
But for local/non-remote player, simulation and render should be the same I'd think?
Just checked, it pulls from camera which has no visual/render scope thing.
Maybe you could try to pull the position on Draw3D instead of EachFrame?
and so currently there is no way to get an accurate player camera position that doesn't lag behind moving vehicles
But there is, https://github.com/michail-nikolaev/task-force-arma-3-radio/blob/master/addons/core/functions/fnc_defaultPositionCoordinates.sqf#L33
@winter crescent Sorry for the ping - the latest Dev says ["Arma 3","Arma3",211,149975,"Development",true,"Windows","x64"] despite the new changelog says 149990, cannot use the text texture
It is showing 149990 in the launcher after updating, interesting. I'll look a bit further.
Using productVersion returns 149990 for me.
In your Steam A3?
Yea, just updated to the Dev steam branch
Okay, let me check and verify again
Gee, now I wonder why my first installation says so even Date modified says it's updated. Looks it's a false positive
Sorry! ๐โโ๏ธ
No worries, Steam can be weird sometimes with downloads.
Some feedbacks:
-How do I change the color of text?
-I think background color format should be RGBA, which is more standard, no?
Ah I knew I forgot something.
yeah probably? Not sure actually. I reused the one function we had to parse these hex strings, and for some reason it did ARGB so I just left it like that
@full sonnet thanks for getting back, unfortunately selection position pilot is not influenced by headtracking position, I have tried every memory point available and none respond to 6DOF headtracking
Ah because it also translates the head/camera and not just rotation
Now I understand....
I think that code for that really just moves the camera
I spent last night trying to run draw3D for positionCameraToWorld with no luck, I also have tried subtracting/adding the position offset between render and simulation to null out the lag of the camera position call but it still isn't smooth
Either a positionCameraToWorldVisual, or make one if the memory points factor in 6DOF headtracking so we can pull that position in render scope. I've looked at anims and there's nothing there either for head tracking
"positionCameraToWorldVisual" doesn't exist even in engine
"or make one if the memory points factor in 6DOF headtracking" if it doesn't actually trigger animations (which it doesn't) that won't work
maybe we can give you the inputs
have you tried inputAction?
That may return the value of the input coming from headtracking?
But then you'll be missing sensitivity stuff
Hmmm interesting, yeah I could then at least approximate the offset from the headtracking input and apply this to a memory position for smooth positioning this would definately be the best solution so far
Good point about the sensativity though, wondering if there would be a way to get the final headtracking output value in-game (like in terms of the confic values for min and max XYZ (I can't remember the exact nomenclature of the top of my head) but there's a place where the config defines max headtracking translation movement for a given vehicle)
This could then easily be scaled to a distance value and applied to a memory point position with vector add
regarding the new dev branch addition:
ENGINE
Added: "text" Procedural texture source - FT-T161708
How can this be changed on the fly? Or is it simply defined once in the object model's surface before compiling into pbo
ah I see, thank you
would it be possible to determine which LOD pip cameras render? current it's locked in the game world/external LOD causing issues for in-vehicle pip applications
Another text feedback: shadow style and its color?
I think I'll leave these both out
What would be the use cases of procedural textures with text?
I'm thinking: nametags on uniforms hiddenSelection, ACE tagging with actual text, adding player/group specific markers to vehicles based on who is crewing it
HUD information?
I can actually see the reasonable way to rotate texts
B-but... is there anything else?
The same with every Arma 3 fonts
https://community.bistudio.com/wiki/FXY_File_Format#Arma_3
โฌ
there's a bunch of direction tablessigns in the game resources, for example
set the rotation, set the texture to red "ARMORY" on transparent background, ???, profit
One of my ideas was the PUBG scoretables that were just textures I think, with this you could make it procedural
Maybe the FT ticket lists some usecases? Don't remember actually
It has become time to tease what I'm working on. Hopefully arriving on dev branch next week or the week after
Electronic whiteboards?
post the other thing ๐
control/ui to texture huh?
Energy drinks are unhealthy.
I'll leave it to you to guess what the feature is ๐
...are you making Arma inside Arma?
Kinda surprised that nobody made a DOOM in Arma yet
We need another Super Miller Land, just that you play it on an actual ingame laptop ๐คฃ
Does that mean that I can finally (dis)play "Bad Apple" in-game? ๐ค
You're the second one saying that today ๐พ
I don't know what it means, but rendering video on surfaces was already a thing previously
but this might make it easier... actually I should probably test that
You should brace up for the upcoming 20 comments that will say the same thing ยฏ_(ใ)_/ยฏ
Although, after seeing Bad Apple on Disassembly Flow Graph , in [REDACTED] Pro, I think nothing will surprise me, at this point ๐
I am already wondering who will create playable video games on laptops already, and what that game gonna be.
Maybe someone will go that far and write even their own mini OS in it. 
Or how a person will push the "Hack laptop" mission beyond a simple "Hold action".
There is already..... "ArmaOS" its called?
https://youtu.be/_nEdm9vPIhQ?t=603 ah thats german.
A laptop with terminal but using display overlay for its functionality
#arma3 #arma3mod #arma3german
00:00 Intro
01:22 Modvorstellung
09:30 ArmaOS
19:20 Outro
Dies ist die Vorstellung meines neuen Mod-Projekts fรผr Arma 3 "Advanced Equipment". Ich stehe noch ganz am Anfang und suche mit diesem Video einige Mitstreiter, denen die Idee auch gefรคllt. Es geht darum, bisher tote und rein dekorative Assets aus Arma 3, v...
just to confirm can these in world UI screens be attached to model coords
Would it be possible to increase the bone limit in model.cfg skeletonBones[] beyond 255 please? Naively I might hope it's a simple question of changing the data type?
I'm really curious about what kind of skeleton you're making that has more bones than a human body
Vehicles. The limit is very easily exceeded.
Hey, Ded, whatever you're making, I'm very thankful. Can't wait to see what my brain can output, I already see some usecases. Thank, You.
you're a legend
I guess I can finally stop bitching about userMFDValues
shit this pretty much makes the whole MFD system obsolete
If setObjectTexture can do it, this can do it (well and weapons/model textures..)
Most likely data type change, across dozens or hundreds of places all over the engine, probably new p3d format version... in short no
This is quite fun already
fps hit for updating that map every frame isn't even that bad, thats a 4k texture
Also the updating only happens when texture is actually visible on screen, quite efficient 
In real life there are those military fold-out smartphone mounts that attach to the chest of your plate carrier. I wonder if someone could use this tech to simulate that, so you can freelook down to see your GPS or whatever.
you could
If your smartphone screen has a selection, right?
and with the bone attach to, you could even get it to look somewhat right
Huh, I thought attachTo can't show something in FPP
I don't know why not
Oh well I realized what I've said is broken enough
I think if it was attachTo your body would render in front of it. I was thinking of making it part of a vest or something like that
[hypothetically making it a vest, not me personally actually doing it]
displayUpdate command that needs to be called manually.
Will set a flag in the display, and next time if its on screen, and the flag is set, it will re-render.
I don't think I want to have some option to set each frame re-render, can just be done with EachFrame script. And performance wise its not a good idea generally.
There are few UI's that change every frame
so I'd imagine MFDs are much more performant than doing something like that with RTT
I don't know how MFD's work, so can't tell
RTT is not expensive though
It renders directly on the GPU, and the texture stays on the GPU, and you can decide when to update it
The main problem is VRAM size of the textures
it's fine with 24 gigs 
You want better clan logo's on vehicles?
No problem, combined with NeilZar's work on setObjectTexture to set the clan selection
I thought Harry Potter is not a thing in Arma
hah! it indeed works just fine
Oh boi, can't wait to see the creations people come up with.
I can bet first will be anime girls
Please also make clan selection override available for configs ๐๐๐
You can override it in config, via init EH ๐
Need an option to turn this off for those extra 2fps on client :P
I'll still need to test fps impact, but its just UI rendering it should be very low
That only works for spawned units tho
Will it work for when selecting the uniform from the arsenal?
And if you drop the uniform and pick it back up
Having an engine solution would be better than scripted ๐
i see a scripted predator camo incomming
with the new UI to texture features being added allowing for detailed pip mfd/displays for weapon control seats in vehicles, a key remaining requirement for this use case will be correctly controlling aim sensativity for the vehicles turrets when viewing the pip screen in first person. We'll need a way to control the turrets aim sensativity based on the fov of the pip screen/camera and not the players view.
I made a ticket for this here: https://feedback.bistudio.com/T168865
This would also allow us to finally create high powered dual render scopes, currently it's almost impossible to aim accurately with dual render scopes over 4x
@full sonnet I also wanna thank you for working in this new feature for us, it's dawning on me all the new features we'll be able to add with this!!
Also wanted to say thanks for this feature as with this we can stop asking for more mfd values ever year
Looking good so far, unless something unexpected happens (which happens quite often :U) it will be in this weeks dev branch.
Docs so far <#community_wiki message>
Will definitely need quite a bit of testing, so if you have a use case please try it out, we still have plenty of time to adjust things
omg it's up?!
time to call off work sick ๐ฌ ๐
16-11-2022
EXE rev. 150030 (game)
EXE rev. 150030 (Launcher)
Size: ~80 MB
DATA
Fixed: Minor clipping of fingers on the right hand when using the Promet rifle
Added: Black variant of Carryall backpack - FT-T167324
ENGINE
Tweaked: Added textColor to text procedural textures and changed the color order to RGB(A)
Added: ctrlAt script command
Added: displayUpdate script command (part of UI to Texture)
Added: displayUniqueName script command (part of UI to Texture)
Added: UI to Texture support
Fixed: allMissionObjects returning an empty array when running on Proton
Added: groups <side> script command - FT-T164521
Added: targets script command syntax for groups - FT-T168408
Here we go, quick UV Checker
Yeah I've expected something more complicated but it is way easier to handle
Pretty nice something yeah
I think the hardest part might be when you have multiple instances of same display, and trying to associate them with the onLoad handler. But the unique name thing should solve that
Can the Procedural texture syntax e.g #(rgb,1024,1024,1)ui("RscDisplayRenderTest","abc"), be applied to a surface directly in an objects p3d?
or must it be applied as a hidden selection texture through script?
for vanilla communities we would at least need scripted, more options always good
Can confirm, possible to use this as a part of p3d texture
https://i.ytimg.com/vi/56mp3fdf6Ys/maxresdefault.jpg
My brain yells for this
p3d textures work just fine, but you may need to be careful with initialization time.
onLoad handler may be called earlier than expected.
https://i.pinimg.com/736x/b5/d3/c6/b5d3c6c73ea683720b1d60a5591a0e20.jpg
Perhaps this as well
Now, I wonder why you've added this ๐
Used it alot in Enfusion and its just nice to have.
And someone requested text to texture rendering, and from there UI wasn't far to go
Okay here is the freaking proof of concept.
dedmen you're my favourite person in the world
how do you do the displayUpdate calls?
from eventhandler?, and at the end the reload/empty thing with eachframe or spawned script?
Draw3D this time
got an rscmapdisplay loaded and working on the blackhawk's cockpit in just a couple minutes and everything worked pretty much like I hoped it would
no more meticulous budgeting of userMFDValues or creating 50 invisible pylons from now on
The scaling of the grid numbers is weird. I also noticed that on my tests ๐ค
That might be dependent on resolution?
could be, though for in the cockpit having the numbers very big like that actually makes them super readable
Looks readable though
ok, then its a feature ๐
vaguely related note but does anybody know if it's possible to rotate a map in a dialog? so I can align the map's direction with the helicopter's heading
Something I didn't test was if UI scaling option in video settings affects the displays 
IIRC it is Big for me?
Maybe it's 0..1 being the whole screen instead of default ui region in the center
but yeah I was messing around with this yesterday and I'm just blown away by how seemlessly this works
I was expecting weird glitches and performance issues
Yeah, this is freaking good feature
I'm looking at the ctrlMap commands and mapAnim and I cannot actually find anything about rotation ๐ค
I don't know of any modded maps that can rotate, I only know of the vanilla info panel map being able to do it
The RscMap config, has a flag that switches between north orientation, and rotating with player
it does?
I didn't notice that, what's the name of the flag
because then I can just switch between two maps, one north and one forward
Yeah as far as I know it is possible, a GPS screen can do that
mapOrientation
0 == north at top
1 == current vehicle heading (that may not work, UIOnTexture doesn't pass the vehicle)
2 = current weapons heading
Ah no nevermind, current vehicle should work.
It uses the focusOn (is that a script command? I think it is?) vehicle, or if not present then cameraOn
and weapon does the same
awesome
thanks!
god why do I have to have a job, I wish I could try this right now ๐
guess it's time to install arma on my work computer and try this on my lunch break lol
Two notes I've found:
- The syntax is case sensitive
- The display name and instance name doesn't require to be quoted
- The syntax is case sensitive
That is intentional, it had to be for "text" because you want to render uppercase letters too
What part specifically is case sensitive that you stumbled on? For classname it wouldn't matter
- The display name and instance name doesn't require to be quoted
hmpf. I guess thats fine but I think I'm not trimming whitespace, which would mess up class name โ I'll fix that later (the trimming)
Mh actually it might.. There might be more stuff wrong there, I may make it require quotes I'll test it
For setting it into a p3d might struggle a bit for its sensitivity
Why? I can't see why
OB tries to lowercase it as far as I tried. Maybe I'll try again later
Welp... well for UI that can be handled.
But for procedural text thats kinda bad ๐
Yeah tried once again. It can save it with mixed cases, but once I open it in OB again it tries to lowercase it
It would seem as though ui2texture, like hidden selections, doesn't work on some weapon attachments in that they can't be applied to scopes/accessory gear? Wondering if anyone else can verify
Disregard my last, just verified that ui to texture works for scope accessories and we can now incorporate full UI into dual render scopes!
Interestingly I found that if you have a UI2Texture object selection applied in pilot LOD and not also in LOD0 it caused a CTD, copying the pilotLOD UI2Texture selection into LOD0 fixed this
@full sonnet ๐ซต After months of searching, I've found the Arma 3 wizard
He always has been ๐
I'm about 9 years too late for the party, but I'm so so glad this game is still being updated and active. Even something that can seem trivial makes a huge impact in this game, and I'm loving every bit.
Wow that something i haven't thought about, animated reticles ๐ amazing
Yeah it is something I can imagine
Can you send me the crash mdmp? Or a mod that has that bug that I can trigger the crash with?
I can consider your update this time is the best ๐๐ 4 me 
Mh indeed it works fine without quotes.
The classname just reads till the comma, and strips off the quotes, but if there is whitespace before the comma its bad.
And the unique name just reads till end of the args portion and also strips quotes.. so yeah indeed both work just fine (except whitespace) without quotes ๐ค
I'll leave it like that but also cut off whitespace
I'm exited to try out some of the possibilities! I was wondering though, "3d" scopes use the collimator rvmat to move the scope shadow elements with weapon sway, I tried this on my new UI2texture surface but it didn't seem to move the UI, wondering if anyone's experimented with materials and UI textures yet
Will do, I'll have a go tomorrow
Even this mess works
#(rgb,1024,1024,1)ui(RscMain " , abc")
But please stay with proper quotes and stuff, everything else may randomly break in a future update and then I won't accept any complaints
IMO the current way that doesn't require quotes is better, since is easier to define in a config
Esp, is an array (like hiddenSelectionsTextures)
I'll also get ' quotes working for config use.
Well without quotes works for now, but if I find a reason to need to not allow it, it can go away. So far I haven't found a reason
https://twitter.com/polpoxEN/status/1592914689603694593
BTW Ded, is it possible to have a proper color shades to use it in an object? As you can see the Neo's texture that used UI to tex is darkened for whatever reason
I don't know why it would do that, its just plain texture 
Also uh, how do we call this mechanic? โDisplay to Textureโ? โUI to Textureโ?
I call it UIOnTexture internally
Mhm
Huh, it probably does so because it lost rvmat shading? The close one is just a setObjectTexture #(argb,1,1,1)color(1,0,0,1,co) and further is the same texture but in UIOnTex
but setObjectTexture shouldn't get rid of the material, no matter what texture ๐ค
i bet POLPOX meant that shading makes a texture look brighter than you see by just looking on it in preview
Well, shading I mean, aka rvmat
#(argb,1,1,1)color(0.5,0,0,1,co) this time. UiOnTex doesn't respect lighting it seems
is your ui also argb vs rgb?
Mh the normal color would be _co, so Diffuse texture type.
UIOnTex is R2T texture type, maybe that makes a rendering difference
Does R2T have similar issue/feature?
oh there is a bug where it doesn't set srgb mode
Nice, so will be fixed?
can you send me UI to test that?
Mind if DM?
nu
I don't need mission, I already built one, just need the UI class
If you want full screen UI on texture you can just do x/y=0 w/h=1
yu
Noic
the safezones are also from corner to corner
Mh all the normals also go away..
But your UI is #(rgb,1,1,1)color(0.5,0,0,1,ca)
instead of #(argb,1,1,1)color(0.5,0,0,1,co)
But that makes no difference, and the srgb flag also doesn't ๐ค
Maybe I can make it diffuse instead of R2T
or I can let you select the type, just like color thing does
displayUniqueName doesn't work at very first frame of the texture (aka onLoad), intentional or nu?
not intentional, I thought I did set it early enough ๐ค
So ya, smth like onLoad = "uiNamespace setVariable [format ['PLP_foobar%1',displayUniqueName (_this#0)],_this#0] ;"; doesn't work
IKR there would be a workaround already, but headsup
_this#1 is the name
O, I might missed your doc?
Yep that was it. But it breaks colors in some other cases
Okey added optional parameter to select texture type
#(rgb,Width,Height,MipCount)ui("DisplayClassName","UniqueName","co")
same as color procedural. So that can be used where needed, and we'll hope that doesn't break anything now ๐
(also with quotes if wanted :u )
in my test it works ๐ค
onLoad = "uiNamespace setVariable ['GMyDataN', displayUniqueName (_this#0)];";
And uiNamespace getVariable "GMyDataN" returns "abc"
And i can't remember fixing that after dev-branch
Hmm
God wait until the Operation Trebuchet devs get their hands on this
That was more trouble than expected 
Wots dis?
I had to rewrite part of our R2T handling to make it R2T non-R2T texture types ๐คฃ
But now you can also make UI normal maps and thermal maps and... specular and macro and stuff
and is the miscoloration from the second screenshot here fixed? Or it needs, say, overriding the rvmat with setObjectMaterial?
as seen in the video, left variant is the "fixed" version with _co texture type
and "unfixed" in this context is what? _ca? Or just skipped "co" in UI texture definition?
I'm curious on how are features decided to be implemented. The subscribes count on a ticket matters?
I'd guess it's a balance of what's in demand (and would open up a lot of opportunities), what's practical and feasible, and what the devs are interested in. A lot of subscribes doesn't really matter if the request can't be done using the available resources or would take forever; a request with few subscribes may be done if it's something quick and easy with a tangible benefit.
I don't care about subscribers.
I judge whats useful and how much work it'd take
But for example, two features that are possible and would take the same amount of effort, but one has more subscribers
But I get it, I'm a dev too... I'm just asking because the A3TI team made some feature requests that are needed to finish their vision for their mod
phabricator is just too crappy to track that
with dev heaven using Redmine, it worked quite well
however we had also a very strict review and approval process for reports
and BI programmers largely only checked those reports
as programmer a reliable and quick repo, is usually most important
feature requests are more like Dedmen said
popularity cant reflect complexity and effort involved
finally had the opportunity to try this, works great! thanks!
On the subject of UI to texture and materials, do you know why the UV transformative collomitor rvmat won't work with UI to texture surfaces when applied as the material?
no
Afaik this rvmat moves the UV map to make scope shadow elements move with weapon sway
The thingy shouldn't care about UV's
maybe its related to the material issues polpox found, and it needs to be a _co type texture

Ok cool, yeah was following that, so I'll try defining the texture type for the texture being used with the collomitor rvmat, hopefully that's the trick, thanks for the info
decided to toy around a bit with the new UI RTT stuff https://www.youtube.com/watch?v=LQYLGDQGqyY&ab_channel=Yax
amazing
Having done more digging, I think this might just be why the scope sway rvmat isn't working
Sorry I know this is a super specific use-case and so I feel bad to bug about this, the new feature is amazing and has opened up a ton of cool new opportunities, but it would be epic if it could be made to work with this rvmat stage, as we could then have animated 3D and Dual render Scopes.
Here's a short vid that hopefully helps to show what I'm referring to:
https://gfycat.com/gracefulscaryazurevasesponge
The center crosshair is a an ordinary texture with an rvmat as follows:
class Stage0
{
uvSource = "texCollimator";
};
The center cross correctly moves with the weapon sway mimicking an infinity position/scope optic paralax - showing the true aim point of the weapon
The UiOnTexture text line is also using the same rvmat, however, it doesn't have its position modified correctly, and is static on the surface of the scope glass.
The way these 3D scope optics are setup is having all the elements at the rear surface of the scope, the texCollimator uvSource then moves the texture's UV to mimic the optical elements like crosshairs and give the illusion of the optics being 3D
Yeah, who needs Discord for streaming, anyway? ๐
Can anyone recommend or help me with the callExtension ability?
I'm looking to get the data from vehicles such as; offroad detection, speed, RPM, machine gun fire etc
Then translate that into SimHub for a buttkicker to turn it into vibration.
At the moment it just detects noise which doesn't work very well when 'quietly' going over rough terrain as you can't feel it
@sudden vapor this is for #arma3_scripting (check https://community.bistudio.com/wiki/Arma_3:_Sound:_SoundControllers and other https://community.bistudio.com/wiki/Category:Scripting_Commands_Arma_3)
Maybe you just wait till next update and see if setting to _co is fixing it
Ok for sure! I just wanted to share my thoughts in case it helped in any way, I'll keep an eye out! ๐
Can someone help write a simple extension for me?
Holy moly, that's incredible! I can't wait to see that soon ๐ฅต
It's still got some minor bugs in there around rendering diary entries, and the code isn't incredibly optimized, but I'll probably push it out on the workshop when the ui-rtt feature goes live
Confirming:
class Stage0
{
uvSource = "texCollimator";
};
now works when used on UiOnTexture ingame!
Thanks again @full sonnet ๐
Does UIOnTexture support mask texture type?
Thank you for proper rvmat support, my UVChecker does work finely with proper shading โบ๏ธ
Also BTW, guess it is reported already but does seem to have the proper text aspect ratio when the screen is only 4:3. Or I'm crazy I guess
it supports all types now ye
not reported? but I thought I had dealt with that
I... think I couldn't make multimat happened, maybe I've done something wrong. Maybe I'll touch tomorrow
https://twitter.com/polpoxEN/status/1596050647505203200
Gabage cod does mean nuthin when it does work!!!!!
is anyone else having issues RTTing a dialog with a map on it? since the last update my game seems to CTD whenever I load any map controls to texture
I was getting CTDs while running on diag_exe, can you try running in the normal dev branch exe?
oh yeah that's it, it runs fine on the normal exe
that's what changed, nothing to do with the update
need probably add note about that on the wiki page, I wanted to go over the wiki thing but no time yet
Does anyone know if I spawn a UiOnTexture (RTT PIP feed) as the text line in a UI class, if I have to call the update in a loop in game to keep it a live feed? Setting up a rtt surface in the object model doesn't require constantly calling an update and the pip feed just runs after spawning
if you want it to update you have to update it
so you're saying in theory we could use this tech to render screenshots to texture by rendering an RTT camera to a RTT-UI and never calling displayUpdate again for that display
no idea what I would use this for, but at some point I will probably come up with some awful reason to use this
unfortunately it wouldn't work for JIP players
uh... yes ๐ค
Functional polaroid camera
youll probably have to update twice. Because of one-frame delay both for RTT render and the UI on texture render
and also I think RTT (as UIOnTexture) only actuall renders to the texture if its visible on screen
god it's tempting
also dedmen I just want to say again that this feature is amazing, we've started using it on some new features in tandem with the MFD system to get the best of both worlds and it integrates effortlessly so far, and performance is also awesome
running a map with displayUpdate every frame seems to be completely fine, before I tested it I was afraid it would pretty much ruin my life doing that
on tanoa rendering a map is definitely noticable on the fps counter, for me
But so is just pressing M and looking at the map ๐
that makes sense, silly me is mainly using stratis as a test environment - do you happen to know if map rendering becomes less intensive for controls like RscMapControlEmpty? Or does the same code still run in the background
they all use the same map control, which is the expensive part
That sadly is what I expected, thanks for the info, I'll be sure to do some testing on Tanoa so I can judge performance better
But in either case that's nothing for you to worry about, I'll just have to add in a potential rate limit on the updates when the FPS gets too low
By the way, do you have any idea why using the map control in diag exe causes a CTD? If so, is that anything you think you'll be fixing?
I know why, no won't fix it.
Its a safety check that's failing, correctly so
๐ good to know, I'll try to just write some check in to my code that stops it from creating the map when you're on diag exe
do we have any simple way of finding out if you're on diag? maybe a preprocessor define
__A3_DEBUG__ is always set on diag
productVersion script command might indicate it
Can confirm that PIP RTT UIonTexture takes two refreshes to load each frame, I think this essentially means the framerate is cut in half unless im missing something.
It occured to me last night that we can now overcome the PIP aspect ratio being broken if the players monitor is other than 1:1 - Just get the horizontal and vertical game resolution at game start and apply the ratio difference to the RTT PIP displays Width
so to ghet a correctly proportioned PIP image for a 1080p monitor just do the following:
1920/1080 = 1.777
PIP RTT class Height = 1
PIP RTT class Width = 1.777
Tested and produces perfect proportions in PIP
oh right i wanted to fix that, forgot about that 
ah no that's pip, so doesn't matter?
But something was wrong with text stretching
There's been an aspect ratio variable in the RTT PIP text line since forever but it never worked
Maybe it's the same thing affected procedural text too im not sure, havent played with that text much yet
but by default the RTT PIP feed assumes the player's monitor is 1:1 and squashes the horizontal UV unless manually counter squashed, I used to do this in the object models RTT/PIP selections UV
PIP yes, but is it a problem for UI on texture too
Is there a way to fix PIP not working if a Zeus is active?
The PIP height/width resolution doesn't currently do anything for PIP spawned as UiOnTexture, just verified ingame
I don't know how you mean that, if its a bug I should fix then I need repro
My b,
-Have an object ingame with a UiOnTexture defined for a surface like: #(rgb,1024,1024,1)ui("RscDisplayRenderTest","abc")
-Have a UI class structure setup in config that references said UiOnTexture surface
-Make one of the UI controls have a text line like : #(argb,64,64,1)r2t(rtt,1.0)
..or..
-Set one of the UI controls text line to #(argb,64,64,1)r2t(rtt,1.0) via script in game
--Result is a pip feed that always has a resolution of 1024x1024
(my guess is that the pip rtt resolution always inherits the UiOnTexture resolution - but I haven't tested this theory yet)
Ah I see ๐ค
hmpf
playing .ogv video in ui on object is also fun
I really love that you can just do any insane stuff imaginable with the UI-RTT feature
Well now I need an aminated start up logo inside my scope mod ๐คท
Yeah I agree, it's like almost limitless possibilities. I'm working on a mod that basically showcases every feature of this so will be sure to share it here once finished
Scope menu system I'm working on
https://gfycat.com/keyfatherlyandeancondor
Outstanding! Can't wait to see more things the community will make, yeah?
Let me guess, it is inspired by Escape from Tarkov?
Thanks! Not originally, but then I came across that games scopes and took some ideas from it like adding a pixel grid, screen border etc
It's been in the works for over a year but the new UiOnTexture added a whole new level of potential depth to it
Nice. Would be the best thing for you for a while
I needed a new coding endeavor :p
Is it still using the โstandardโ PIP and then adding addition layers with the new UiOnTexture feature?
(Sorry- on mobile so small images!)
So I actually tried both RTT as its own surface and also RTT being drawn onto the UiOnTexture, both work, and I have more control with it on the UI, but that 1frame lag while calling an update does become noticeable, so I'm looking into a way around that currently
Quick feedback about colored laser, when Alt+Tab the game will make the laser IR-only until you turn it on again
๐ค wat
So when you tab out, the visible laser becomes IR only, until you turn it off and on again?
Ye as far as I can see
Maybe something for Neil to fiddle with ๐ซ
@carmine juniper Do you have an example code for when it happens, or does it always happen when the final argument (isIR) in drawLaser is set to false?
Sorry if isn't clear enough, I meant laser pointer accessory with isIR = 0 config
Ah, okay. I didn't think of the config version.
There is a test addon in the ticket
I think I may have already found the issue, if this comment is anything to go by.
// only scripted(Temporary) lasers can be non-IR, and we've just deleted them all. Thus there can be no daylight lasers anymore
That comment probably isn't updated right, but the code under it is. But that flag getting broken on tab out could be it
I'm not able to reproduce it with the test PBO.
Nevermind, I am able to reproduce it now. You need to have the pause menu open when alt-tabbing
I also noticed that the bullet traces in the arsenal also update their rendering when I turn on the laser again.
Seems that is unrelated though
Through some similar work in Apache mod we've noticed it is dependent on screen resolution
I can't send pictures here, but it looks aligned at a 4:3 monitor resolution but stretched horizontally at a 16:9. Can't put evidence pictures here but happy to send it via another method
(not a problem we can't work around though if required, the rest of this system has been absolutely amazing to play with)
Has anyone run into text being stretched when used as UiOnTexture?
I've checked the UV mapping and it isn't streched, all my picture elements spawn with the correct aspect ratio too (provided I use Keep Aspect), but all the text is stretched sideways, and I can't find a way to adjust the aspect ratio for the letters, adjusting W/H just moves the borders for the text string
Like this?
ah.. yes exactly like that, my b ๐คฆโโ๏ธ
I tried to repro that stretching but couldn't
What is the trick to make it as bad as possible
My use case might not be the best as an example but I can describe my setup, So I have a perfectly circular flat surface (1:1) in my object with UiOnTexture setup as 1024x1024, with a circular unstretched (1:1) UV map for it too. I reference this in my config to get a handle for code. I try spawning my UI text elements onto that display with their width and height set to normal width and heigh values in config equal. (Changing width and height in config and in script only adjust the text string's borders, not the aspect ratio of the text itself (not an issue just a game thing))
Result - text is always stretched to be longer sideways than it is tall.
I'm running the game on a 16:9 monitor which I think is the cause of the problem as @carmine juniper had experimented
The thing is BC we have control for the width and heigh of picture UI elements in game we can counter for this in config and scripts, but afaik, there's no way to adjust the aspect ratio (width and height) for text elements
I would send picture but I'm at work ATM
That sounds similar to what I have seen. The difference between 16:9 and 4:3(12:9) is small enough that it is easy to miss unless you have a reference like on the attached imgur. If it wasn't for the exact same font in a different method being directly below, I likely would never have noticed
I would guess that I could fix the stretched UI text elements by squashing the uv map for the UiOnTexture surface, as this works for the same pip aspect ratio issue, but this would only solve the problem for people using 16:9 monitors, for anyone with a different monitor aspect ratio they have the opposite issue
In our position, we are lucky enough to be able to have control of the vehicle, so UVAnimations can work for us
Hiya, there will be no more branch updates this year. We'll return in 2023 on the road to 2.12. Wishing all of you splendid holidays ๐
I didn't get a chance to test this so let me ask a couple of questions:
- Does it create a new display (as in adds a new entry in allDisplays)?
- How does the clean up work if not?
also how do you get the display object to modify it? 
is it thru the "uniqueName"?
#(rgb,Width,Height,MipCount)ui("DisplayClassName","UniqueName","texType")
like uiNamespace getVariable "uniqueName"?
ok after reading some comments it seems like you can only show a pre created display
but not sure how you do the main menu and other displays in this video, because as far as I see you haven't "created" them? 
- Yes (No)
- What u mean with how? If texture goes away the display is removed. But you got no control over texture going away as it goes to a cache
Init handler in display class
The "seems" is wrong
Oh I see now. Thanks a lot!
hmm. but how is this one done? is it somehow drawing a part of the display as texture?
and I could ask the same question about the DUI radar in your video
or are you drawing the whole display?
Display created twice, once on texture and once normal
oh I see 
Spawned the DUI display on texture, and then some script tricks to make it use that one in script
Next dev:
needService
allowService
allowedService
"Service" mission EH
documentation incoming
DEVELOPMENT and RELEASE CANDIDATE branch discusssion goes here
Also fixed:
Arma allocating up 20GB of data into the pagefile.
Dev branch or malloc mod?
I presume is just a nonsense joke
Is it possible to have steam have 2 branches for different folders? It's almost impossible to upgrade and downgrade 1.5gigs on Aussie internet all the time
HAHA @steady sundial and after that you want GabeN to make peace in the entire world
((PRAISE OUR LORD AND SAVIOUR))
You can use the game updater in the tools for having different branches of arma3 at the same time
Ahh awesome cheers @fading ether and @gilded basin all hail master GabeN
i've got all steam branches separately, using steamCMD to download each, not steam client
Tweaked: Added support for CT_OBJECT creation using ctrlCreate - FT-T165245
โผ๏ธ
Very, very nice!
please test thoroughly and let me know is anything is missing
yeah 21:9 vs 4:3 
The width of fonts depends on the aspect ratio at game start 
This seems not fixable
Welp fixed it. Question is just what things I broke with that now 
this is something i do so often, run into the issuie, try to fix it, fail, complain about it, and fix it before anyone notices i was struggling
Amazing dedmen, thanks again for looking into this, this will fix a ton of stuff on my end ๐๐
That's good news that it's not completely doomed and unfixable
can't wait for the fix to drop
So for that keybinding issue from #perf_prof_branch
I'm thinking, we make keys block by default.
Meaning for SHIFT+W keybind. If you hold W in editor you fly around, if you then press shift you stop. When pressing shift, the shift+w action wins and the W action deactivates.
Then we have config entry, like modifierBlocking=0 where you can turn it off, so that your action won't block others
I think thats the best solution?
@versed abyss @halcyon moth ?
Ref to the thing that started this https://feedback.bistudio.com/T164496
Makes sense to have actions define if they're blocking or not.
#community_wiki message
I hope that works for your usecases, mainly for Darkbelg as you want the non-default behaviour
I'd say it should be the opposite (non blocking by default)
For backward compatibility
Backward compatibility would mean to have it blocking by default
That's how its been

"So that your action won't block others" does that mean that it runs parallel ? If so then yes this is fine.
Could another possibility be to look at the onActivate config code and when it returns false the next onActivate is triggered? I mean the first thing i tried was to return false in the onActivate.
So the CBA implementation of keybinding also has this problem?
Iโm guessing thereโs something fundamental in the code that means the irdotcolor array only works for visible lasers in this statement about new goodies for 2.12?
If it were available for IR lasers as well, I suspect using large values in the array would be a way to control the dot brightness so would be a solution to the issue I was asking about in the modelling channel?
Would extending this to IR be possible?
(Colour would be irrelevant - just want a way of changing brightness)
CBA works completely differently, it doesn't block keys
IR has no color, or rather, the color is IR, thats why IR
brightness, no
you can use dotSize to control size/brightness?
That just makes it bigger not brighter
its just a 2D image, not a light
The effect Iโm after is a much less bright beam (which Iโve managed) but much brighter impact dot and itโs that bit that Iโm struggling with
But does it not have an rvmat with an emissivity value?

I think/hope that I can get the effect I want by upping that emissivity.
Although like the laser beam, would probably need to make a new model to assign a new value


โผ๏ธ