GunnerFlyboy

Did ED ever get this into one of the builds?

I can say using Claude to do troubleshooting, tuning hardware, and DCS has been quite a ride for me as well. I've not had a chance to use the AI for code development. But he did point me here and thank you for doing this.

I've spent a lot of time in the past combing through bug reports to get to a place where I could enjoy DCS. I'm retired USAF aircrew and after that career I spent another 20+ years doing software development. I even developed software solutions for FSX and P3D a long time ago. It occured to me to use an LLM AI to comb though the data not only from these forums, but to also use it analyzing simulation flight test data derived from monitoring the CPU, RAM and GPU sensors and correlating these to the DCS.log. I spent about a month in my spare time training an AI to become a simulation test data analyst. I was pleasantly surprised by how much faster problems could be tracked down in the hardware. Also it helped to tune or retune DCS (like after an update). The end result is I can determine issues almost as fast or faster than spending time waiting and wading through the forums for clues as to what is wrong. In fact, the AI has developed a "mental" picture of how DCS operates. I'm not sure its correct more test data needs to be collected to give credence to the conclusions. I'm now spending a lot more time flying with instrumentation on the hardware running DCS. So if something goes wrong a solution is no longer many hours and and in some cases - days to get the sim and hardware back in a stable state so I can spend more time flying and less time troubleshooting. Is there anyone else here that has done anything similar? And if that is true, would they be interesting in sharing their experience? Maybe the discussion would help others who are struggling to get the sim working faster?

I've been using Claude 4 Sonnet with extended thinking. Many of the errors are in modules that aren't mods. I had a box that needed tuning and OC and undervolting to be able to run most scenarios. It actually goes and reading the forums and other sites that have data related troubleshooting data. It took test flights in three different scenario sizes large (~2000 objects), medium (~1000), and small (~500). It was able to determine not only what errors mattered, it could give guidance on OC and undervolting the CPU and GPU. I even asked it the build components for a new machine. Now I use it for all flights where there is any issue.

Sweet!

I had the same problem, yesterday the sim worked fine, then today I couldn't get it to launch with the new launcher. I did all the troubleshooting and found a workaround that gets you back into the sim without a repair or more invasive troubleshooting. The issue is resolved (until ED can figure out the cause) by bypassing launcher via a modification of the options.lua file. Open the file up in notepad or notepad++ (Make a backup!). You can find it in this location: C:\Users\[Username]\Saved Games\DCS\Config\options.lua Once you have options.lua open, look around line 140, search for this: ["launcher"] Now that you have found this line of code: ["launcher"] = true change it to: ["launcher"] = false Save the file and then you should be able to start DCS like the old way. We are just using old way which didn't depend on the launcher. DCS log was truncated so there was no information that was usable, I looked also at the event viewer - nothing there out of the ordinary. My system specs: HP Omen 30L GT13-0090 (Modified for greater thermal cooling and RAM) Intel Core i9-10850K (Stock: 3.6GHz, OC: 4.7GHz with undervolting) NVIDIA RTX 3090 OC and undervolted. 64GB RAM @ 3200 MT/s Windows 11 Home Build: 24H2 DCS Version 2.9.17.12034 Doc

How did you do that? And did it work on conjunction with the other modification to the B-52H.lua file above?

I'm not sure if this has been requested before. I searched and this aircraft didn't come up. A full fidelity version would be nice, but at least an AI version of the OA-1K Skyraider II. https://en.wikipedia.org/wiki/L3Harris_OA-1K_Skyraider_II

So the laser guided ordnance doesn't need a TGP to do terminal guidance? I know there is a pylon on the right wing but no litening pod. Or is it because its AI it doesn't need a TGP?

Just to be clear, what do you mean by "game reset." I only know of repair at this point.

+1

I'm mapping the controls and the china hat seems the logical choice, has anyone else done it this way on their warthog throttle quadrant?

Strange I tried it after deleting and creating a blank autoupdate_log.txt and it updated the updater then updated the sim.

I've never had issues, rebooting the router doesn't work. And I've done more than one update here. I don't think its on my end.

I've tried several times to update DCS to 2.9.4.53990 and I keep getting these indications: With this autoupdate log entry: autoupdate_log.txt Here's the tracert: C:\>tracert cdn.digitalcombatsimulator.com Tracing route to 1231349722.rsc.cdn77.org [89.187.164.37] over a maximum of 30 hops: 1 1 ms <1 ms <1 ms 192.168.1.1 2 8 ms 12 ms 9 ms syn-142-254-131-189.inf.spectrum.com [142.254.131.189] 3 11 ms 10 ms 18 ms lag-61.hcr01mckntxwf.netops.charter.com [24.28.88.145] 4 24 ms 13 ms 14 ms lag-28.plantxmp01r.netops.charter.com [24.27.13.8] 5 10 ms 9 ms 9 ms lag-27.mcr11crtntxjt.netops.charter.com [24.175.36.177] 6 12 ms 11 ms 11 ms lag-21.rcr01dllatx37.netops.charter.com [24.175.49.0] 7 12 ms * * lag-414.dllstx976iw-bcr00.netops.charter.com [66.109.6.52] 8 12 ms 11 ms 11 ms lag-0.pr3.dfw10.netops.charter.com [66.109.5.121] 9 14 ms 11 ms 27 ms charter-dal.cdn77.com [84.17.32.144] 10 11 ms 11 ms 12 ms vl205.dal-clg3-dist-1.cdn77.com [185.156.45.47] 11 10 ms 11 ms 9 ms 746217339.dal.cdn77.com [89.187.164.37] Trace complete. C:\> Any help would be greatly appreciated.

B-52D/G/H Crewdog here. Would like to see all three models I flew developed. I'm a also a retired software developer. If someone needs a SME and developer in one, I would be glad to do it to honor those of us who flew her in combat.

Here's the latest test track at NTTR: B-1B NTTR LL Bombing Flight Test Non VR - MT.trk

Former crewmember here, would like to see at least the loadout expanded on the AI model.

I've tested the Bone in NTTR scenario and it crashes every time I test it. Nose pitches down way too much for a decent to low altitude flight and impacts the ground still in a nose down attiude. Also in the summary post mission it reports "pilot dead" before the crash.

Sorry you would think an old B-52D/G/H guy would get that right, thanks for the correction.

I'm not sure if this was an oversight, bug or what, but I see the mount for a targeting pod on the bone, but no pod. And and there should be a pylon on the right wing (facing the nose) between the engine nacelles with a targeting pod. On the left wing of the BUFF we carried a pylon and TACTS pod when we were on the range in Nevada. Also the munitions loads available for both seem very limited as they both can carry many other conventional weapons. I've not tested the bone, but the buff loses its weapon load if you try to modify it. Update: In the 2023 and Beyond the BUFF has a targeting pod (see attached image from video). null

I would add "Doom" and "Hulk" to the B-52H part of the list.

Thanks, that did it.

How to jettison fuel tanks only... Tried combat jettison but the pylon's when with the wing tanks.

Strange, I renamed the directory for the 2nd time after a repair and uninstalling and reinstalling the aircraft and I can now map the controls. Weird. Thank for the help!