Autonomous CAP and GCI AI fighter script

[lukrop]

Just relased RC2. Check it out here. This version requires the current development version of MIST.

[Gabriel8387]

Does this work with the nttr map?

 

EDIT:Yes it does

Edited February 9, 2016 by Gabriel8387

[Stonehouse]

It should.

[Gabriel8387]

Red side is working great. But blue side isnt working at all. Ive checked and checked and everything is done right. Anybody else experiencing this on Nevada?

 

EDIT:lol nvm I found spaces in the pilot name...derp

Edited February 9, 2016 by Gabriel8387

[Stonehouse]

:smilewink:

 

Happens to us all at one point or another lol

 

Red side is working great. But blue side isnt working at all. Ive checked and checked and everything is done right. Anybody else experiencing this on Nevada?

 

EDIT:lol nvm I found spaces in the pilot name...derp

[Quax456]

It should.

 

Lukrops script throws me exceptions when a ground unit dies :cry:

 

And the inAir CAP's spawning far far away of their CAP Zones...

Edited February 9, 2016 by Quax456

[Gabriel8387]

So im having a problem with AI spawning on the runway. Any suggestions?

 

EDIT: So I noticed these aircraft kept spawning exactly where the center of my Nellis AFB trigger was. So I enlarged the trigger zone and moved the center to the NW and now they seem to spawn in correctly...im going to play around some more and see if I can confirm this.

Edited February 10, 2016 by Gabriel8387

[Gabriel8387]

So moving the trigger did seem to fix the runway spawning, however now I get multiple aircraft spawning in the same place.

[Stonehouse]

Weird, unless something changed in concept in lukrop's translation the trigger zone over the base is really just to "mark" the airbase as a GCICAP base and that is why it must be the same name as the base because the old script basically used the name to get the airbase id and vec2 position of the base and the spawn should use the airbase id for spawning in on the ground with a waypoint type of ramp or runway rather than turning point. The old script used to use the radius of the zone to clean up wrecks at the airbase but that was about it and that was why it needed to be located over the base to define the clean up area.

 

Few more bugs to squash by the sounds of things.

[lukrop]

And the inAir CAP's spawning far far away of their CAP Zones...

That's because they spawn at their airbases. AFAIR it was not possible to port units with MIST which where dynamically added. I'll need to look at the MIST code and see if it's possible to fix that.

 

Weird, unless something changed in concept in lukrop's translation the trigger zone over the base is really just to "mark" the airbase as a GCICAP base and that is why it must be the same name as the base because the old script basically used the name to get the airbase id and vec2 position of the base and the spawn should use the airbase id for spawning in on the ground with a waypoint type of ramp or runway rather than turning point. The old script used to use the radius of the zone to clean up wrecks at the airbase but that was about it and that was why it needed to be located over the base to define the clean up area.

 

Few more bugs to squash by the sounds of things.

 

It's still exactly like you described. So in theory it should work. I'll have a look at it sometime. Currently (because of the placement of the airfields) I don't see the point of having GCICAP on NTTR. Sooner or later it will be fully supported of course.

[Gabriel8387]

Currently (because of the placement of the airfields) I don't see the point of having GCICAP on NTTR. Sooner or later it will be fully supported of course.

 

Actually I have a few missions this would work great on. How long are you thinking?

[Stonehouse]

lukrop I noticed you have updated the master branch again and that one of the issues mentioned in the code comments is that "takeoff" is not recommended at this time because of tasking timing issues.

Have you thought of making any ground launched CAP or GCI head to a random point say 10-20 kms from the airbase ie have a random first waypoint somewhere 10-20 kms from base climbing to CAP altitude and then reset their tasking and route to be what you need? That way you could test for being airborne and perhaps a certain distance from their take off base or even wait until they have reached waypoint 1 before doing the tasking etc. I would think this would solve your issue.

 

It would probably happen this way in real life to a degree anyway as unless it is an extreme emergency aircraft don't tend to just blast off through the base traffic pattern in a bee line to their mission destination. Even priority traffic might need to make a bit of a small dogleg near the base to clear traffic already in the area before heading out on their mission heading.

[lukrop]

I just released the third release candidate for 1.0.0.

Download it here

 

Release notes:

This version fixes the long standing bug #5 as well as making GCI flights only spawn if there is no free CAP closer than an airfield, as it is stated in the README.

 

If border checking is enabled the script now alerts the user if it can't find the group whose waypoints form the border and aborts.

 

lukrop I noticed you have updated the master branch again and that one of the issues mentioned in the code comments is that "takeoff" is not recommended at this time because of tasking timing issues.

 

AFAIR it was related to not beeing able to delay the tasking the way I wanted. Honestly I can't remember now what the exact reason was. Those are some interesting thoughts though and I'll have a look into it ASAP. For now I'll just create a issue on github. Thanks Stone!

 

edit: Ah! I wrote it down in the comment :D

if a flight is queued for takeoff and not already in the game world while getting tasked

DCS delays the spawn of the group on the runway if other groups/units are just/already taking off. So your suggestions should fix this.

 

 

edit2:

Actually I have a few missions this would work great on. How long are you thinking?

Could you please test with the last release candidate? I didn't have problems using it on NTTR the last time I tried. I'll see if I can quickly build a NTTR showcase mission.

edit3: Just created a showcase mission, demonstrating usage of GCICAP on NTTR. Download here.

Edited March 2, 2016 by lukrop

[wdrago]

I just released the third release candidate for 1.0.0.

Download it here

Hello Lukrop, I just tested your RC3 on both 2.0 and 1.5.3,

 

you can find my mission with RC3 inside here and there.

 

On Nevada map RC3 is smoother than RC2 (less lag issue), end CAP waypath are less look like roller coaster, didn't had runway stuck for GCI.

- Red CAP doesn't respawn if shootdown or after landing (doesn't unspawn).

- Blue CAP also doesn't respawn if shootdown, but respawn after landing (even sometime before landing).

- GCI (blue/red) spanw once only.

- CAP set start from air, sometime those stupid AI collide themself on spawn.

 

- like for RC2 CAP waypath are not allways respecting CAPzone.

 

On 1.5.3

- CAP set start from air, too much ofen those stupid AI collide themself on spawn.

- I need futher testing (uploaded version linked above CAP are set to take off from parking).

 

In old version CAP was spawning over CAPzone, now over airfield, is that a choice you made, or a need for new dev. ??

 

I really appreciated the improvement you made on RC3, that's super cool keep going Buddy :thumbup:

 

I will give you futher feedback on 1.5.3.

If you want to test the mission I made with your RC3, I attached the Lua files for both missions.

M2K Nevada Full.lua

M2K Star Script plus-c.lua

Edited March 3, 2016 by wdrago

[Pikey]

Same issue with split flights as before. If you kill 1 out of 2 and the surviving aircraft RTB's then it will stay parked, never despawn and the cap will never replenish.

 

Tested on the provided sample mission at x10 speed, took about 5 minutes until it broke.

[lukrop]

Hello Lukrop, I just tested your RC3 on both 2.0 and 1.5.3,

 

you can find my mission with RC3 inside here and there.

 

On Nevada map RC3 is smoother than RC2 (less lag issue), end CAP waypath are less look like roller coaster, didn't had runway stuck for GCI.

- Red CAP doesn't respawn if shootdown or after landing (doesn't unspawn).

- Blue CAP also doesn't respawn if shootdown, but respawn after landing (even sometime before landing).

- GCI (blue/red) spanw once only.

- CAP set start from air, sometime those stupid AI collide themself on spawn.

 

- like for RC2 CAP waypath are not allways respecting CAPzone.

 

On 1.5.3

- CAP set start from air, too much ofen those stupid AI collide themself on spawn.

- I need futher testing (uploaded version linked above CAP are set to take off from parking).

 

In old version CAP was spawning over CAPzone, now over airfield, is that a choice you made, or a need for new dev. ??

 

I really appreciated the improvement you made on RC3, that's super cool keep going Buddy :thumbup:

 

I will give you futher feedback on 1.5.3.

If you want to test the mission I made with your RC3, I attached the Lua files for both missions.

 

Same issue with split flights as before. If you kill 1 out of 2 and the surviving aircraft RTB's then it will stay parked, never despawn and the cap will never replenish.

 

Tested on the provided sample mission at x10 speed, took about 5 minutes until it broke.

 

Thanks to both of you. I'll see and try to fix the issues with split flights. Regarding the issues on NTTR I'll prioritize them lower for now. Sorry about that.

 

In old version CAP was spawning over CAPzone, now over airfield, is that a choice you made, or a need for new dev. ??

This needs a fix. Teleporting them with MIST wasn't working the last time I tried. I need to workaround that somehow.

Edited March 4, 2016 by lukrop

[Pikey]

Lukrop, it's a pig of a script to get working, sorry for the constant feedback. I hope you can burn through this, we are all waiting.

[wdrago]

sorry for the constant feedback. I hope you can burn through this, we are all waiting.

 

Not sure to correctly unsderstand what you mean.

 

But from my part, I see my feedback, as humble contribution to the project, my intention is not to bother, or being impatient. as you said this is a "pig script," so I supposed he/they doesn't get time to test everything.

 

So I hope feedback are taken for what they are ? I really hope my feedback are not received as stress or so...

I also guess that, seeing what we are doing with the script or how we mixing it with some others scripts, may help them in dev progress...

If i'm wrong tell me ,)

Edited March 4, 2016 by wdrago

[Steggles]

Pikey seems to be the resident DCS script tester. He helped a lot with finding bugs in EWRS for me too. I'm sure all of your inputs and time is appreciated. I know I appreciated pikey's time and help when I first released mine.

[Pikey]

To clarify, by a "pig of a script" its a large stubborn immovable object that takes a lot of work. GCICAP is really useful and played a part in my missions for more than a year now. The level of effort to repair it post 1.5 perhaps wasn't appreciated by Lukrop at the time, I'm sure it is now, but it doesn't make the job any easier.

 

I constantly test new releases looking for fixes so I can use this in live public servers, but its not ready yet. I hate saying it as much as everyone else but i'm not giving up on this project, it is to large squadron PvE what CTLD is to helicopters. Simply putting down cap zones as respawnable aircraft isn't enough, we tasted blood with the previous 1.2 versions when they worked.

 

Our dedicated server sits mostly locked to the public since 1.5 first came out because of the remaining issues. I don't want a PvP TvT server, I want an environment that just one player can climb into or 20 players, GCICAP gives us that, the use of AWACS and does it all so well.

 

So that's why i'm trying every release asap and have for months now.

 

Not sure to correctly unsderstand what you mean.

 

But from my part, I see my feedback, as humble contribution to the project, my intention is not to bother, or being impatient. as you said this is a "pig script," so I supposed he/they doesn't get time to test everything.

 

So I hope feedback are taken for what they are ? I really hope my feedback are not received as stress or so...

I also guess that, seeing what we are doing with the script or how we mixing it with some others scripts, may help them in dev progress...

If i'm wrong tell me ,)

[lukrop]

I'm not taking any offense or feel pressured by your feedback guys. I appreciate it very much. For example I didn't even notice/forgot/missed the bug Pikey just mentioned.

 

Actually I'm just play-testing a fix of this very bug. So far it's looking good and I hope this resolves some other issues too. First half-destroyed group just landed and despawned.

[Quax456]

Dear LUKROP,

 

unfortunatly I have found some new BUGS! :(

 

 

1. Your scripts handles TriggerZones over FARPS like AIRBASES. This causes some funny behavior of spawning GCI groups :D
   
2. Using '__CAP__blue1' as unit name within a template is ok but not with a single group having the same Name as its unit! Then you get LUA execeptions I have never seen :cry:
   
3. Your newest version spams GCI flights, but doesnt respawn CAP. I mean as far as I have seen it in a short period of time.
   
4. CAP flights do not RTB if they are wounded or 1 of 2 got shot down..
   

[Pikey]

Good news then! :)

I was thinking of trying an airfield tidy up and seeing if the 'static aircraft' that is created when they land and shutdown as a singleton can be removed to get the flight respawning again. I'm sure this problem came with this new feature in 1.5. where parked aircraft turn into statics.

I'm not taking any offense or feel pressured by your feedback guys. I appreciate it very much. For example I didn't even notice/forgot/missed the bug Pikey just mentioned.

 

Actually I'm just play-testing a fix of this very bug. So far it's looking good and I hope this resolves some other issues too. First half-destroyed group just landed and despawned.

[lukrop]

1. Your scripts handles TriggerZones over FARPS like AIRBASES. This causes some funny behavior of spawning GCI groups :D
   
2. Using '__CAP__blue1' as unit name within a template is ok but not with a single group having the same Name as its unit! Then you get LUA execeptions I have never seen :cry:
   
3. Your newest version spams GCI flights, but doesnt respawn CAP. I mean as far as I have seen it in a short period of time.
   
4. CAP flights do not RTB if they are wounded or 1 of 2 got shot down..
   

 

Are you using RC3 or commit 7cc6d5d? This commit fixed the issues with CAPs not despawning if the group lost units.

 

I'm testing with Marioshata's Battle for Nalchick mission. He named all the template groups exactly the same as the pilot name e.g. __CAP__blue1 for both, pilot and group name. No problems here. Also this shouldn't matter since GCICAP only looks at the pilot names and if it doesn't find the ones it needs it already tells you which one is missing.

 

CAP is respawning in the mission too, as well as GCI not spamming. Actually I didn't see any GCI spawn/take off in the last hours this mission ran since the free CAP flights in the area got vectored to the intruders.

 

Since half damaged groups now despawn the mission ran over night and in the morning I still saw both sides patrolling.

 

Edit: Oh and what did you name the trigger zone over the FARP? Need to test that.

Edited March 6, 2016 by lukrop

[Quax456]

Are you using RC3 or commit 7cc6d5d? This commit fixed the issues with CAPs not despawning if the group lost units.

 

I'm testing with Marioshata's Battle for Nalchick mission. He named all the template groups exactly the same as the pilot name e.g. __CAP__blue1 for both, pilot and group name. No problems here. Also this shouldn't matter since GCICAP only looks at the pilot names and if it doesn't find the ones it needs it already tells you which one it's missing.

 

CAP is respawning in the mission too, as well as GCI not spaming. Actually I didn't see any GCI spawn/take off in the last hours this mission ran since the free CAP flights in the area got vectored there.

 

Since half damaged groups now despawn the mission ran over night and in the morning I still saw both sides patrolling.

 

Edit: Oh and what did you name the trigger zone over the FARP? Need to test that.

 

 

1. I used commit commit 7cc6d5d
   
2. The Zone is named like the Farpsname e.g. "blue farp 1"