Krazy1

Ah. I wasn't precise. It wasn't actually running, icon was just out of sync initially. I'll try to test but holidays may intervene. Thanks

Thanks guys. It's WAY faster now. My 150 part ship is running about 30 FPS... at least double. So much more playable. The difference between landed and flight is only 5-10 FPS now. One problem though... it's making my GPU fan louder. Can you fix that?

My kerbals have staring contests... everyone wins.

Seems reasonable. But I can't reconcile that with your post above Then why does turning it off (landed) make the frame rate triple? So does stock KSP, right? I'm deducing that the cost function is some product of N bodies and P parts. So it's doing N-body gravity on every part individually? If so... yeah that's a lot to calculate. Couldn't the gravity calculation be done just once using the ship CoM? Then assume the same gravity vector applies to each part. The parts are very close together relative to solar system scale. So computation time would be F(N CBs) + G(Parts), not F(N*Parts). I guess it would compromise rotation from differential gravity and tidal forces...? Seems worth it.

Very nice @SpaceFrog Welcome to the forum. Any chance there's a video coming?

Thanks for the notice. I've been a good boy Santa... please bring me lots of garbage collection optimizations. One thing that's still baffling is the dramatic change in frame rate between landed and in flight. It's literally 10 FPS in flight and the instant it touches down it's 30 FPS . Vacuum or atmosphere, doesn't seem to matter. Landed actually has additional interactions with the CB surface versus a free body in flight. So I'm really curious how the update will affect this behavior. Well, it's not "impossible". I just use 4x physics warp, set KAC to pause near the end of the burn and do something else. 1 hour burn takes 15 minutes, when you don't have to watch it isn't too bad IMO.

Completed a contract today to make 4 tourists pass out. The real challenge was that it had to be "sub-orbital" so airplanes wouldn't work. They got their money's worth... passed out twice! Earned Gwenfry a 14G ribbon. Jeb is jealous. @Hotel26 Bop is retrograde for stability with Principia so the situation is similar to yours when going to/from Bop from another moon. I did a Jool 5 and starting from an inner moon, I pushed AP (JP?) way way way out, then flipped inclination about 180 deg. at AP, then aligned PE (JE?) with Bop. That transfer took a whole Kerbin year but saved dV.

Yes. Both ends actually. I removed the strut, removed boosters, set them to the side, then reassembled. Looks like Nov 12 crash also had a part removed and reattached. I know it's not much to go on so far. I hope others report if they have similar issues.

Several missions over the last week or so... Flew a sightseeing trip to N76A to see more arouras... did not disappoint. Album: https://imgur.com/a/gmL3bYS Probably my last flight for awhile... had to dump Kerbal Konstructs. Filled up my save file with junk. Finished the mission to recover ore from Minmus... with parachutes this time. Found a terrain bug that destroyed my ship on loading... had to land from orbit again . Reentry got the 10m shield to 3388K, 96.8%. Anybody destroy one of these before? Album: https://imgur.com/a/qJN4nqC Returned the tourists with LKO Bus MK2. Unstable, almost burned up. Got rave reviews. Album: https://imgur.com/a/XJQlr5B Tested a Boar engine in Minmus orbit then crashed it for fun. Peak acceleration 230 m/s^2 Album: https://imgur.com/a/SBsbrJY

I had to remove KK from my JNSQ install. My .SFS had 500k lines of bloat in it and was getting very slow saving/ loading. I also had Omega's Stockalike Structures and KSRGAP. Also installed JNSQ_KK_Fixes.zip. More info here: https://forum.kerbalspaceprogram.com/topic/220799-500000-lines-of-garbage-in-save-file/#comment-4341412 Also it kept forgetting when I opened facilities... I had to reopen many of them again when I came back to them. But it wasn't consistent/ predictable. I really liked flying around to all the different sites and being able to refuel.

Another crash to desktop from VAB. I picked up a free strut floating beside the ship this time. I did some undo/redo before that. Similar error message as Nov 12. Infinity or NaN floating point numbers appear when calculating the transform matrix for a Collider. Scene hierarchy path "strutConnector/model/ReStock/Assets/CompoundParts/restock-strut-1(Clone)/strut_main_anchor/strut_main_SPHERECOLLIDER" Player log KSP log error log crash dmp I would just try removing Restock but I'm using Restock+ parts on ships in my career.

Worth a shot although I'll miss having them. Removed KK, Omega's stockalike structures (also needed for the KK buildings in JNSQ), JNSQ-KSRGAP and Pilot Assistant. Deleted all the hex-code objects for destructables. SFS is now 30 mb, was 38. "only" 900k lines now. It does seem a little faster going to KSC, and it's smoother without drawing each object after it shows the scene. The in-flight save stutter is better, still stutters a couple times but not as long.

Thanks. I've been using that. It didn't seem to help much with the save-stuttering issue or scene change load time but I haven't done a real direct comparison. It did help with the KSP load times and terrain-morph stuttering, like when a mountain changes shape as you get closer.

I usually set damping = 2, spring= 1.5. That will help keep wings level when turning. Brakes stronger at the rear... like 80% rear, 60% front. And avoid the large or extra large gear. The multi-axle (bogie) setup is not stable... it vibrates on the ground and slides with no external force. Tweakscale the medium gear for large planes.

I had a crash to desktop while building in the VAB. I put a Place Anywhere 7 RCS on the side of a docking port then later moved the docking port and it crashed. Was not able to reproduce. KSP log Player log

Every time it autosaves, it stutters badly 2 or 3 times. I've seen this before with large save files... apparently KSP puts file saving directly in the simulation loop without caching - very bad. It also reloads the whole file at every scene change... like KSC to VAB, VAB to launchpad... each change takes 20+ seconds. So slow. My save file is 38 MB partly due to 500,000 lines of this junk: I'm guessing it's Kerbal Constructs? I'm playing JNSQ and I've flown to every airport several times. It frequently does not remember when I open a resource at a KK facility. An early save only has maybe a hundred of these entries.

Thanks for fixing the +1 nodal revolution button. Very useful, like when going from eq. LKO directly to Minmus without a plane change. Also for this one "...we were missing the application of the angularDrag computed by KSP." I believe I was seeing this effect recently during launches, when rockets were spinning without stopping passively with fins. One problem I'm still having is very low framerate, only 10 FPS when flying a 124 part airplane. It's 30 FPS on the ground (taxiing) which is tolerable, but it drops dramatically at liftoff. It's almost the same in in orbit- 150 part ship getting 10-15 FPS. A 30 part ship is 40 FPS. KSC overview is 80+ FPS. Tracking station is 40. It's very sensitive to part count. It's roughly 2x faster with stock physics for the same part count. It's definitely CPU limited, not graphics. Running JNSQ with stock aero (not FAR). Ryzen 5 3600.

That's the one. Sorry... I assumed you knew every mod for KSP

Launch attempts, mostly failures... Made it this time:

Request for Better SRB support. It basically works if you do not change the default grain geometry/ slots (thrust profile). Only default 8-slot works or stock size; can't change both. Changing slots results in engine reverting to thrust for the original engine size after reloading the ship. It still looks like the tweaked size and has the tweaked fuel quantity and mass but has stock thrust. Burn time seems to scale proportionally for the thrust and fuel. For example a Thoroughbred changed to 6-slot and 3.75m (in that order) initially shows about 7000 kN and 72s burn. That's incorrect, showing values for 8-slot. After removing and reattaching it shows 6-slot values 4879kN and 90s. After reloading the ship it's about 1600kN (same as 2.5m size) and 5 minutes burn time. Launch test confirms reduced thrust (and it certainly won't make orbit with 1/3 thrust)

After the Mun ore return contract debacle in the last update, where parachutes where forgotten, I went through all the stages of grieving... Then proceeded to aerobrake to LKO. Then launched a supply probe with 30 parachutes. It was night and hard to control so I only got this pic on my first launch attempt with wrong SRB decoupler staging. Parachutes have arrived.

Another minor issue... it always defaults to turn on when a ship is loaded. Just wastes CPU time if you're not using it.

Made another video of the auroras at S72A, time-lapse of the whole night this time. And did stuff in space for a change... Launched an lander that will return 2350 ore from Mun. Big rocket with wobbly 2.5m center section. Took a few tries to get to orbit. Saw this rare Mun, Minmus, Jool conjunction by chance. To the Mun... And on the way back from Mun I realized I have.... no parachutes! Arggg. It weighs 57 tons. The inflatable heat shield is a good start but it needs parachutes. A lot of them. The End!

Just a little realism improvement... I've flown around Kerbin a lot with KSRGAP (lots of pics on What did you do in KSP1 today thread). The auroras look cool but you shouldn't be able to fly above them. The default is 15 km but they should be higher in the atmosphere. So I changed the EVE config to raise them close to the edge of space: Kerbin 70 km, Jool 470 km, Lindor 400 km. Looks much better to me (Haven't been to Jool or Lindor yet) Some pics: And a VIDEO

Last week-ish I flew yet another JNSQ circumnavigation 7-leg trip for 8 KSRGAP contracts. 130 PAX on Pilum 200. Trip pics and bloopers: Also played with EVE auroras. More coming on JNSQ thread.