General Discussion (dev branch)

[Laqueesha 474]

Infantry units don't display the proper facial animation when they are killed. They should display the SetMimic "dead" or "unconscious" facial animation upon death but at the moment they just have blank and stoic emotionless faces when they die.

[AveryTheKitty 2626]

On 2/16/2018 at 4:27 PM, zukov said:

 

 or a version covered with the same french lizard camo like syndikat pants , the helmet is clearly a french Spectre or something like that

 

It's a PASGT.

[d3nn16 3]

Hello

 

I installed the dev build version of the Linux server using steamcmd and 'app_update 233780 -beta development validate'.

And installed dev build version of the Windows client using steamcmd and 'app_update 107410 -beta development validate.

 

When I start the dev build client on my Windows machine I see correct version 'Development Build version 1.81.144316'. But when I go to the server list I see my server with a red cross and its info card shows 'version: 1.80'. I disabled addon check in server config file (verifySignatures=0, onDifferentData = "", onUnsignedData = "", onHackedData = "", BattlEye = 0). In the server logs I see incorrect version 'Arma 3 Console version 1.80.143869 x86'.

 

Does this mean it is not always possible to test the latest dev build on a Linux server? Do I need to wait until the Linux Server Dev Build 1.81.144136 is released?

 

Thanks for any help

[Beagle 684]

12 hours ago, Night515 said:

 

It's a PASGT.

could as well be a "Schuberth M92", they all look the same.

[fn_Quiksilver 1636]

Low-hanging fruit bug for a bored dev:

 

Easy repro included

https://feedback.bistudio.com/T127737

 

Estimated fix time 15-20 min

[Strike_NOR 898]

Some positive feedback on the Titan top-attack missile "homing" :)

 

 

The missile now finds it's way to the nearest 'homies' much faster. Homing works as expected.

 

image uploader

 

On the flipside.

 

About 0.1 ms after this screenshot was taken, neither of the homies were killed. The tank was damaged, but no crew killed.

 

This is another gripe. Explosive weapons do not seem to damage turned-out crew. Neither do they really damage internal crew, even in cars, where a 40mm grenade to the drivers window "should" kill the guy(s) inside.

Will this be tweaked for tanks DLC? :)

[Beagle 684]

2 hours ago, Strike_NOR said:

Some positive feedback on the Titan top-attack missile "homing" :)

 

 

The missile now finds it's way to the nearest 'homies' much faster. Homing works as expected.

 

image uploader

 

On the flipside.

 

About 0.1 ms after this screenshot was taken, neither of the homies were killed. The tank was damaged, but no crew killed.

 

This is another gripe. Explosive weapons do not seem to damage turned-out crew. Neither do they really damage internal crew, even in cars, where a 40mm grenade to the drivers window "should" kill the guy(s) inside.

Will this be tweaked for tanks DLC? :)

This is a difficult thing. Tzhe problem is, HEAT charges tend to work on a point, not with a hughe explosion and a large kil radius. Also...an An explosively formed penetrator (EFP) does not work very well to kill occupants if it does not hit anything like ammunition or fuel, it just blows a hole and when the hatches are open, no overpressure can form inside. It not easy to make that into a game mechanic....An ATGM s not a bomb. Since WW2 a lot if developemt went into crew protection meinign that after a hot the vehicle is totally out of action, but all or most of the crew are alive. Even the russian developemt prioritized crew proctection now, seeT-14, T-15. Main point is, damage dealt will always be an abstraction inside such a game. For full realtime simulation you would need a whole processor for that task alone.

 

[Pyronick 21]

38 minutes ago, Beagle said:

This is a difficult thing. Tzhe problem is, HEAT charges tend to work on a point, not with a hughe explosion and a large kil radius. Also...an An explosively formed penetrator (EFP) does not work very well to kill occupants if it does not hit anything like ammunition or fuel, it just blows a hole and when the hatches are open, no overpressure can form inside. It not easy to make that into a game mechanic....An ATGM s not a bomb. Since WW2 a lot if developemt went into crew protection meinign that after a hot the vehicle is totally out of action, but all or most of the crew are alive. Even the russian developemt prioritized crew proctection now, seeT-14, T-15. Main point is, damage dealt will always be an abstraction inside such a game. For full realtime simulation you would need a whole processor for that task alone.

 

I guess that in this particular case you don't need to make it rocket science. Just the shrapnel of the ATGM casing that will splinter, before or during impact doesn't even matter here, is very likely to kill these two men here.

[AveryTheKitty 2626]

4 hours ago, Beagle said:

could as well be a "Schuberth M92", they all look the same.

 

PASGT is literally in the classname.

[scavenjer 112]

1 hour ago, Beagle said:

This is a difficult thing. Tzhe problem is, HEAT charges tend to work on a point, not with a hughe explosion and a large kil radius. Also...an An explosively formed penetrator (EFP) does not work very well to kill occupants if it does not hit anything like ammunition or fuel, it just blows a hole and when the hatches are open, no overpressure can form inside. It not easy to make that into a game mechanic....An ATGM s not a bomb. Since WW2 a lot if developemt went into crew protection meinign that after a hot the vehicle is totally out of action, but all or most of the crew are alive. Even the russian developemt prioritized crew proctection now, seeT-14, T-15. Main point is, damage dealt will always be an abstraction inside such a game. For full realtime simulation you would need a whole processor for that task alone.

 

Not at all, HEAT is still an explosive and therefore will definitely act as HE (DM12 120mm HEAT-MP was specifically designed to be good against vehicles and infantry).

Even after penetration they are deadly, without spall liner crew inside tanks will be shred to pieces by spalling and fragments of the EFP.

With spall liner the EFP itself will still generate fragments and a blastwave, contrary to popular belief there's no "overpressure" generated that would be harmful, however the shockwave will incapacitate any crew (unconscious).

 

the US did some testing with sheep inside a tank, they put 1lb of explosives inside it with no frag, no sheep were killed however they were all unconscious from the shockwave.

Something very similar is seen with EFPs.

[dragon01 902]

HEAT penetrates with armor with a hot jet of molten metal. On the other side of the armor, you've got a bunch of very hot plasma in a small enclosure. This will likely kill anyone inside a compartment that was hit. Less so with EFPs and sabot, which are much more likely to just knock the crew unconscious from the impact.

 

As for the "HE" part of HEAT doing damage, it does, but it's a purely concussive explosion that is not actually that powerful. Unless you're using a dedicated HEAT-MP round (which tend to have added fragmentation feautres), then the kill radius is gonna be pretty small.

[Strike_NOR 898]

@Beagle I have rigged 2.75'' rocket shaped charge warheads for demonstrational purposes IRL. If that TITAN-AT missile from my screenshot had ANY form of similar warhead, both guys would be dead from the pressure. It's like a HE grenade goes off 1 m from your head. Even without fragmentation, you are dead. Besides, my point is even if you dropped a 500lbs bomb 5 meters from the tank, the turned-out crew will not die in ArmA. Which is inaccurate. I think turned out crew are handled the same way as turned in crew when it comes to explosives. Only projectile hits will "directly" kill them. 

 

@dragon01 The molten metal jet has been debunked here as a myth already. Also my local EOD specialist confirmed. The liner (usually copper) is deformed by the explosive, but not molten. It does not melt any armor. However, the focused blast will force the steel armor to "give way". This deforms the steel so rapidly that the steel armor heats up and sends some steel sparks flying. 

The impact of HEAT looks like it makes the armor melt, but it's just being forced away by a focused jet of extremely high pressure.

Yes, this creates high pressure and very hot steel spalling/fragments, which are likely to kill crew, destroy equipment and ignite ammunition.

 

As for the HE potential, it's still enough to kill you at close range, but in no way a "general purpose" weapon. In my screenshot, the two tankers should be dead.

[Tankbuster 1744]

18 minutes ago, Strike_NOR said:

@Beagle I have rigged 2.75'' rocket shaped charge warheads for demonstrational purposes IRL. If that TITAN-AT missile from my screenshot had ANY form of similar warhead, both guys would be dead from the pressure. It's like a HE grenade goes off 1 m from your head. Even without fragmentation, you are dead. Besides, my point is even if you dropped a 500lbs bomb 5 meters from the tank, the turned-out crew will not die in ArmA. Which is inaccurate. I think turned out crew are handled the same way as turned in crew when it comes to explosives. Only projectile hits will "directly" kill them. 

 

@dragon01 The molten metal jet has been debunked here as a myth already. Also my local EOD specialist confirmed. The liner (usually copper) is deformed by the explosive, but not molten. It does not melt any armor. However, the focused blast will force the steel armor to "give way". This deforms the steel so rapidly that the steel armor heats up and sends some steel sparks flying. 

The impact of HEAT looks like it makes the armor melt, but it's just being forced away by a focused jet of extremely high pressure.

Yes, this creates high pressure and very hot steel spalling/fragments, which are likely to kill crew, destroy equipment and ignite ammunition.

 

As for the HE potential, it's still enough to kill you at close range, but in no way a "general purpose" weapon. In my screenshot, the two tankers should be dead.

Am I the only one who can't understand why this debate is going on? These two guys are fucked. End of.

[lex__1 422]

Tell me at what stage the decision of this ticket

https://feedback.bistudio.com/T82570

If the ticket is difficult to solve, it is possible to open this ticket

https://feedback.bistudio.com/T120265

The problem in the first ticket is to stop the rotation of the UAV - it forces you to use offline mode. But the offline mode turns off the camera lock and there is no target tracking in the PIP window.
As a result, the problem with UAV is wider than could be.

[goko-- 147]

7 hours ago, Beagle said:

This is a difficult thing. Tzhe problem is, HEAT charges tend to work on a point, not with a hughe explosion and a large kil radius. Also...an An explosively formed penetrator (EFP) does not work very well to kill occupants if it does not hit anything like ammunition or fuel, it just blows a hole and when the hatches are open, no overpressure can form inside. It not easy to make that into a game mechanic....An ATGM s not a bomb. Since WW2 a lot if developemt went into crew protection meinign that after a hot the vehicle is totally out of action, but all or most of the crew are alive. Even the russian developemt prioritized crew proctection now, seeT-14, T-15. Main point is, damage dealt will always be an abstraction inside such a game. For full realtime simulation you would need a whole processor for that task alone.

 

It would be great if crew catch fire though. Catch fire and jump out of the vehicle.

[x3kj 1247]

Passengers or crew do not exist for explosive damage calculation. Explosive damage to passengers (of any kind) is the value of explosive damage to the vehicle multiplied with crewExplosionProtection config property - no matter if turned in or out. And it doesnt matter if it's a battleship or a motorcycle - every crewmember will receive that damage. Has been like this for eons.

[Strike_NOR 898]

6 hours ago, x3kj said:

Passengers or crew do not exist for explosive damage calculation.

Thanks for verifying. I had this suspicion for ages.

[bars91 956]

8 hours ago, x3kj said:

Passengers or crew do not exist for explosive damage calculation.

 

Well, that's surprisingly shitty. For 2013... it's 2018 now.

[Grumpy Old Man 3545]

20 hours ago, Strike_NOR said:

About 0.1 ms after this screenshot was taken, neither of the homies were killed. The tank was damaged, but no crew killed.

 

This is another gripe. Explosive weapons do not seem to damage turned-out crew. Neither do they really damage internal crew, even in cars, where a 40mm grenade to the drivers window "should" kill the guy(s) inside.

Will this be tweaked for tanks DLC? :)

 

 

Another thing that grinds my gears:

 

 

When I fired the shot the camera switches to unit next to the vehicle.

setAccTime set to 0.075 so at 60fps one frame will take roughly 1ms.

 

Notice how within 1ms after the projectile hit the vehicle the first unit is suddenly turned in...

The hint displays time, video has a resolution of ~1ms in game time per frame.

 

 

Now imagine you got 2 snipers coordinating their attacks to take out both targets at once, this would fail if the guy in the front will be hit first, since the guy in the rear turns in within a millisecond.

Considering bullet positions are only updated every frame and one frame at 60fps takes 16.67ms this could get really frustrating...

 

Cheers

 

[Dedmen 2700]

On 18.2.2018 at 12:18 AM, d3nn16 said:

I installed the dev build version of the Linux server using steamcmd and 'app_update 233780 -beta development validate'.

And installed dev build version of the Windows client using steamcmd and 'app_update 107410 -beta development validate.

 

When I start the dev build client on my Windows machine I see correct version 'Development Build version 1.81.144316'. But when I go to the server list I see my server with a red cross and its info card shows 'version: 1.80'. I disabled addon check in server config file (verifySignatures=0, onDifferentData = "", onUnsignedData = "", onHackedData = "", BattlEye = 0). In the server logs I see incorrect version 'Arma 3 Console version 1.80.143869 x86'.

 

Does this mean it is not always possible to test the latest dev build on a Linux server? Do I need to wait until the Linux Server Dev Build 1.81.144136 is released?

AFAIK there are no dev branch builds for linux server.

1.81 will never be released. Next release version is 1.82

[x3kj 1247]

4 hours ago, Grumpy Old Man said:

Notice how within 1ms after the projectile hit the vehicle the first unit is suddenly turned in...

The hint displays time, video has a resolution of ~1ms in game time per frame.

This immediate "pop" to inside could be due to animation states switching from turned outside to dead state animation (which may just exist for inside). So when he died, he immediately popped to another animation. Its a possibility. Reaction time is way too crazy nevertheless.

In general, the AI seems to have issues with beeing too precise and immediate in things that shouldn't be.

Example 1: Crew disembarking when vehicle is rendered immobile or in the condition to explode. It takes 1 frame for them to recognize this and jump out (and start shooting anyone who stands near with their personal weapons).

Example 2: Someone on discord this weekend showed a video where he thought AI can see through walls. It seemed to me however, like the AI only heard the sound of a weapon, but immediately and with 100% accuracy pinpointed the source of origin and then turned in precisely this direction to start searching visually - it didnt find anything of course, because the sound source was obstructed by the wall.

[.kju 3244]

Quote

In general, the AI seems to have issues with being too precise and immediate in things that shouldn't be.

 

another example is awareness of death/kill/destruction

 

unfortunately it seems BI programmers didnt consider to introduce some (random) delay into AI awareness and perception (at least in many areas)

[Strike_NOR 898]

@x3kj @.kju

 

Would make sense to make this into an AI skill slider parameter. "Reaction time". Could be the same parameter that influences aim time etc. 

Make that a base value, the eject time could be a coefficient of that base value. So for instance, reacting to enemies in CQB is generally fast, but as crew reacting to tank damage it gets an increased reaction time based on the initial level.

[dragon01 902]

Guess so. They're probably working on finishing the new tanks, and any new features that'll follow will likely be implemented on them first. I just hope it's not too far away...

[Sgt. Strongest Military Ever 19]

I think we'll be seeing something new this week: