[WIP][1.8.1, 1.9.1, 1.10.1, 1.11.0–2, 1.12.2–5] Principia—releases every new moon—n-Body and Extended Body Gravitation

[GrahamsNumber]

12 hours ago, maccollo said:

The planets simulate at much lower rate that vessels, so if you disable asteroids you can simulate 1000 years in 2 or 3 minutes with the stock system. If you add more stuff it will obviously be slower, but you shouldn't have to wait hours.

Thanks for replying, @maccollo. I tried following your advice, but I wasn't able to figure out how to disable asteroids. A google search turned up nothing useful (only how to get rid of "tracked" asteroids, but I am not tracking any). I got myself a mod that allows a 10M times speedup, so far that's the fastest I have been able to go. (If you know how I can simulate even faster, please let me know). Would you happen to know anything I can do to easily check that nothing went wrong (including that no moons were ejected from their systems)? In that case, I would be happy to let the system simulate for a while and then report back. I am sure that Outer Planets Mod people would appreciate to know whether or not they can run Principia safely.

Edit: I wrote before that the system simulates very slowly, but this seems to have been a temporary problem with my computer. The system is simulating quickly now, and I will report back with any results I can come up with.

Edited April 24, 2017 by GrahamsNumber
See post

[GrahamsNumber]

Hello again, @maccollo, @eggrobin, @nanomage

After some fiddling around, I finally managed to simulate the system for around 6000 years. Unfortunately, it's as feared: With Outer Planet Mods installed, the systems are not stable. Even with the custom Kopernicus configuration mentioned here, Bop flies out of the system eventually. I'm guessing this is because OPM moves Eeloo to an orbit around Jool. Aside from trial and error, is there any way I can find out the right Kopernicus configuration to make everything stable? Then I could post my results here for future reference.

[pleroy]

(Mod co-author here.)

17 hours ago, GrahamsNumber said:

I have read your FAQ entry for Kopernicus users. Unfortunately, that small entry does not state how to make sure that the rest of the system is stable (i.e. how to simulate it, other than just letting the game run at max speed for a couple of hours). Would you have any more guidance on that, and what I should do if I find that the system is unstable?

Unfortunately no, this is a hard problem.  Basically you need to explore the phase space of the parameters of the modified solar system until you find a stability zone.  It took @eggrobin weeks to do that with the stock KSP system.  I am not familiar with OPM, but basically the KSP system is a bad starting point because the planets are way too light (Jool's mass is about 2/3 of that of Earth and 1/500 of that of Jupiter) so they get moved around very easily.  The KSP system is just not very interesting from the perspective of n-body physics.  It would be much more interesting to simulate real-life extrasolar planetary systems (these are known to be stable otherwise we would not be observing them).

17 hours ago, GrahamsNumber said:

Second: Only the error on exiting EVA was fatal, and for that, I will try to reproduce it (luckily, I still have a save), but for the issue with the vessel sinking into the Mun, this doesn't even cause a crash. Anything else I could provide here for you?

I don't think so, we'll just need to try to reproduce the problem by landing on the Mun with enough logging enabled.  (But the fix won't be in the upcoming Cartan release.)

2 hours ago, GrahamsNumber said:

Thanks for replying, @maccollo. I tried following your advice, but I wasn't able to figure out how to disable asteroids. A google search turned up nothing useful (only how to get rid of "tracked" asteroids, but I am not tracking any).

Use the FAQs, Luke!

30 minutes ago, GrahamsNumber said:

After some fiddling around, I finally managed to simulate the system for around 6000 years. Unfortunately, it's as feared: With Outer Planet Mods installed, the systems are not stable. Even with the custom Kopernicus configuration mentioned here, Bop flies out of the system eventually. I'm guessing this is because OPM moves Eeloo to an orbit around Jool. Aside from trial and error, is there any way I can find out the right Kopernicus configuration to make everything stable?

As explained above, no.  And trying to do that in the game is going to be an awful waste of time.  For the stock KSP system @eggrobin used a mix of C++ and Mathematica to efficiently explore the phase space.

[Uace24]

Is this RSS/RO compatable? 

[pleroy]

3 hours ago, Uace24 said:

Is this RSS/RO compatable? 

Yes.

[nanomage]

12 hours ago, GrahamsNumber said:

Hello again, @maccollo, @eggrobin, @nanomage

After some fiddling around, I finally managed to simulate the system for around 6000 years. Unfortunately, it's as feared: With Outer Planet Mods installed, the systems are not stable. Even with the custom Kopernicus configuration mentioned here, Bop flies out of the system eventually. I'm guessing this is because OPM moves Eeloo to an orbit around Jool. Aside from trial and error, is there any way I can find out the right Kopernicus configuration to make everything stable? Then I could post my results here for future reference.

well there's always the option that doesn't require research: you can remove bop (or any other offending celestial) from joolean orbit and make it circumsolar instead, or make it orbit any other planet (it will probably be stable enough around dres?). Sure it damages the feel of OPM, but on the upside, most anyone can do it.

[GrahamsNumber]

19 hours ago, nanomage said:

well there's always the option that doesn't require research: you can remove bop (or any other offending celestial) from joolean orbit and make it circumsolar instead, or make it orbit any other planet (it will probably be stable enough around dres?). Sure it damages the feel of OPM, but on the upside, most anyone can do it.

I guess that is possible (I mean, who cares about which planet it's orbiting?). But I doubt that this would solve the problem. Sarnus (equivalent of Saturn) has a lot of moons, and these are not necessarily stable either. I'm guessing there would be a way to find a region of stability, but as pleroy mentioned above, it's not easy and would take a lot of work. If we want a stable, bigger solar system, I guess that means waiting until RO and FAR are out of beta. (For those interested: Yes, if you are determined, you can get Real Solar System, Realism Overhaul, FAR, and a whole bunch of other mods working in 1.2.2. But, fair warning: it is a pain in the neck, since you can't use CKAN, sometimes you need to build the mod yourself from the source code, run Perl scripts and in general have to live with some bugs until these are ironed out). Personally, I would love to help in the development, but I think my programming skills are pretty amateurish

[eggrobin]

For the new moon (lunation number 214), the new release, Cartan, is out, bringing with it a utility for rendezvous: the target local vertical local horizontal reference frame. There is a guide to low orbit rendezvous using the new reference frames (screenshots to be added in the near future).

See the change log for details.

Please read the concepts page carefully, it has been expanded since Cardano. Remember to also have a look at the FAQ if you have a question.

Thanks again to @Iskierka for testing the Linux build.

[Clubbers]

So has Vall always been unstable on Cardano save and when updating it recognizes that?

Edited April 25, 2017 by Clubbers

[pleroy]

20 hours ago, Clubbers said:

So has Vall always been unstable on Cardano save and when updating it recognizes that?

The Jool system was unstable.  We don't patch existing saves, so if you keep using a Cardano save, the unstability will still be there.  You'll need to create a new save with Cartan to avoid running into the unstability.  We apologise for the inconvenience.

Edited April 26, 2017 by pleroy

[Gordon Dry]

Hmm, I checked the log for something else and stumbled upon this:

[LOG 13:03:41.794] AssemblyLoader: Loading assembly at E:\Spiele\Kerbal Space Program\GameData\Principia\x64\principia.dll
[ERR 13:03:41.800] Failed to load assembly E:\Spiele\Kerbal Space Program\GameData\Principia\x64\principia.dll:
System.BadImageFormatException: Format of the executable (.exe) or library (.dll) is invalid.
  at Mono.Cecil.PE.ImageReader.ReadOptionalHeaders (System.UInt16& subsystem, System.UInt16& dll_characteristics) [0x00000] in <filename unknown>:0 
  at Mono.Cecil.PE.ImageReader.ReadImage () [0x00000] in <filename unknown>:0 
  at Mono.Cecil.PE.ImageReader.ReadImageFrom (System.IO.Stream stream) [0x00000] in <filename unknown>:0 
  at Mono.Cecil.ModuleDefinition.ReadModule (System.IO.Stream stream, Mono.Cecil.ReaderParameters parameters) [0x00000] in <filename unknown>:0 
  at Mono.Cecil.ModuleDefinition.ReadModule (System.String fileName, Mono.Cecil.ReaderParameters parameters) [0x00000] in <filename unknown>:0 
  at Mono.Cecil.ModuleDefinition.ReadModule (System.String fileName) [0x00000] in <filename unknown>:0 
  at AssemblyLoader.ScanForBadTypeRefs (System.String file) [0x00000] in <filename unknown>:0 
  at AssemblyLoader.LoadExternalAssembly (System.String file) [0x00000] in <filename unknown>:0 

Full log:
https://www.dropbox.com/s/0e6xron5avp01zt/2017-05-06-1 KSP.log.7z?dl=1

[eggrobin]

6 minutes ago, Gordon Dry said:

Hmm, I checked the log for something else and stumbled upon this:

[LOG 13:03:41.794] AssemblyLoader: Loading assembly at E:\Spiele\Kerbal Space Program\GameData\Principia\x64\principia.dll
[ERR 13:03:41.800] Failed to load assembly E:\Spiele\Kerbal Space Program\GameData\Principia\x64\principia.dll:
System.BadImageFormatException: Format of the executable (.exe) or library (.dll) is invalid.
  at Mono.Cecil.PE.ImageReader.ReadOptionalHeaders (System.UInt16& subsystem, System.UInt16& dll_characteristics) [0x00000] in <filename unknown>:0 
  at Mono.Cecil.PE.ImageReader.ReadImage () [0x00000] in <filename unknown>:0 
  at Mono.Cecil.PE.ImageReader.ReadImageFrom (System.IO.Stream stream) [0x00000] in <filename unknown>:0 
  at Mono.Cecil.ModuleDefinition.ReadModule (System.IO.Stream stream, Mono.Cecil.ReaderParameters parameters) [0x00000] in <filename unknown>:0 
  at Mono.Cecil.ModuleDefinition.ReadModule (System.String fileName, Mono.Cecil.ReaderParameters parameters) [0x00000] in <filename unknown>:0 
  at Mono.Cecil.ModuleDefinition.ReadModule (System.String fileName) [0x00000] in <filename unknown>:0 
  at AssemblyLoader.ScanForBadTypeRefs (System.String file) [0x00000] in <filename unknown>:0 
  at AssemblyLoader.LoadExternalAssembly (System.String file) [0x00000] in <filename unknown>:0 

Full log:
https://www.dropbox.com/s/0e6xron5avp01zt/2017-05-06-1 KSP.log.7z?dl=1

This is normal, that's just KSP trying to load a native DLL as a managed one. Nothing to see here. When there's something to see in the logs it's typically in our logs, not in KSP.log; we don't log there.

[bobkabob]

I'm noticing an instant ~200m/s acceleration on every vessel when transitioning from space to atmosphere (70km) on Kerbin. Making reentry impossible for my early career craft. It's a mod heavy install, but a quick removal of Principia makes the problem go away and a reinstall makes it reappear. Later I'll try it on a fresh install of KSP and report back.

 

EDIT:

Fresh install fixed the problem, now to reinstall mods one by one to find the culprit.

 

EDIT2:

It appears to happen randomly and with different mods installed (tested all night!). Running Tasker and Resource Monitor on my second monitor shows no abnormal CPU loads or memory issues. Heck, on my last attempt fully modded it didn't even do the acceleration. I'll try to learn how to pull logs and see if those shed any light on whats going on. If anyone else is experiencing this, please chime in! 

Edited May 9, 2017 by bobkabob
Update

[lawndart]

Hi Bobkabob,

Yes, I'm getting this too. My deorbit burn should have dropped me to 23Km at periapsis, but I reached periapsis at 67Km - my velocity had kicked up by 200M/s as I hit the atmosphere and I was never able to deorbit.

I also have quite a few mods installed, including Kopernicus and the Galileo Planet Pack.

I'll also do some testing. Hopefully it's caused by mods/parts that are actually installed on the spacecraft, e.g. Kerbal Engineer Redux, or some flight characteristic modifiers, such as Persistent Rotation.

[GrahamsNumber]

9 minutes ago, lawndart said:

Hi Bobkabob,

Yes, I'm getting this too. My deorbit burn should have dropped me to 23Km at periapsis, but I reached periapsis at 67Km - my velocity had kicked up by 200M/s as I hit the atmosphere and I was never able to deorbit.

I also have quite a few mods installed, including Kopernicus and the Galileo Planet Pack.

I'll also do some testing. Hopefully it's caused by mods/parts that are actually installed on the spacecraft, e.g. Kerbal Engineer Redux, or some flight characteristic modifiers, such as Persistent Rotation.

That reminds me that I had the same problem. Also mod heavy install, but the predictions that Principia would give me were constantly off. I.e., I would be in orbit, burn retrograde, and be told that I would reach 20k m at periapsis. Except I would end up somewhere near 60k. Never noticed it before, but now that you mention it, the predictions always were heavily off when it came to deorbiting. I ascribed this simply to the integrator that Principia uses being too inaccurate, but now that I think about it more, the difference of a couple ten thousand meters is pretty relevant after all.

Not sure I can make a useful bug report for now though, can't play KSP right now (Graphics card troubles)

[lawndart]

OK, think I have found something regarding the 200m/s addition.

I made a copy of my current Principia Cartan install which is modded with such things as Kopernicus/Galileo, as mentioned previously. 

I rolled back to Cardano and there were no obvious problems in a new sandbox game. I then deleted Cardano and installed Cartan and started a new sandbox game with the 200m/s problem reoccurring. I built an entirely stock ship for testing.

I then noticed that suborbital flights on Gael (which has the same planetary values as Kerbin) were not triggering the 200m/s addition. Putting a ship in orbit then reducing the periapsis so it fell below the atmosphere entry point at 70km did trigger the addition as the ship crossed the boundary. So I launched a ship into a suborbital trajectory, saved, then experimented with different periapsis heights to see if the problem occurred at a specific point. And it does.

If your ship's periapsis is below -70km the 200m/s does not get added. Above -70km (e.g. -69km) the 200m/s is added to the ship's velocity.

Please could someone confirm this?

[lawndart]

More testing...

I made a completely new install of KSP and added Principia Cartan as the only mod. A simple ship of Mk1 capsule, small nose cone, T800 fuel tank, T400 fuel tank, 4 winglets and a T1 Aerospike engine. Saved just before apoapsis.

Without Kerbal Engineer Redux installed I can't see the numbers for periapsis until it is raised above 0km. I'm also getting mixed results. The bad news is the 200m/s addition is still occurring.

Suborbital flights do not trigger the 200m/s addition where the periapsis is below Kerbin ground level. Once above ground things get complicated. Suborbital flights with a periapsis as high as 68km were not triggering the addition. Others did trigger the addition, but only where the apoapsis was higher than ~84km. 

So the issue may be apoapsis dependent. 

 

 

[pleroy]

1 hour ago, lawndart said:

I made a completely new install of KSP and added Principia Cartan as the only mod. A simple ship of Mk1 capsule, small nose cone, T800 fuel tank, T400 fuel tank, 4 winglets and a T1 Aerospike engine. Saved just before apoapsis.

Without Kerbal Engineer Redux installed I can't see the numbers for periapsis until it is raised above 0km. I'm also getting mixed results. The bad news is the 200m/s addition is still occurring.

Suborbital flights do not trigger the 200m/s addition where the periapsis is below Kerbin ground level. Once above ground things get complicated. Suborbital flights with a periapsis as high as 68km were not triggering the addition. Others did trigger the addition, but only where the apoapsis was higher than ~84km. 

Great, it seems that we are getting close to something that we should be able to reproduce.  Could you give us your save (or saves) as close as possible to the point where the problem shows up, as well as your INFO log?  The best way to go is to create a github issue with detailed explanations of what you are seeing and links to one or several gist with the save/INFO files.  Thanks for your perseverance in trying to nail down this bug.  With some luck we might be able to fix it in the upcoming Catalan release, planned for May 25th.

[lawndart]

Hi Pleroy,

I have been trying to save just before the issue manifests itself, but it appears the act of pausing the game to save the file prevents the issue from occurring, or at least it has in the dozen or so attempts I have tried so far.

I have also noticed the on-screen altimeter occasionally jitters for about half a second as the spacecraft transitions into the atmosphere but I have not had time to check whether this happens only when the 200m/s is added.

I have also sent a spacecraft to Eve, but haven't been able to replicate the velocity change on hitting the atmosphere we get at Kerbin as yet. I've only made half a dozen re-entries at Eve so far, so it may manifest itself at some point.

I'll try to get more testing in when I can, but I am supposed to be doing some decorating

[pleroy]

Closing the discussion on the 200 m/s acceleration: thanks to a save and detailed instructions provided by @lawndart we were able to reproduce, debug and fix the problem (see this issue for details).  The fix will be in the Catalan release, planned for May 25th. 

[msnbcorp]

Hi everyone, just reporting the same problem that the guy one page before. I was debugging an other mods when i run into these 2 lines :

[EXC 19:48:21.824] NullReferenceException: Object reference not set to an instance of an object
    principia.ksp_plugin_adapter.PrincipiaPluginAdapter.OnGUI ()
[EXC 19:48:21.826] NullReferenceException: Object reference not set to an instance of an object
    principia.ksp_plugin_adapter.PrincipiaPluginAdapter.OnGUI ()

 

Don't know why exactly but it seems that in some case during loading principia launch nullreference exception

Just to let you know

[Abstract_Kerman]

Can I smash gilly into Kerbin by attaching rocket engines to it, since it's not on rails.

Edited May 15, 2017 by Abstract_Kerman

[Planetace]

I heard this mod allows planets to have an axial tilt... Has anyone of you got evidence of this...

[pleroy]

23 hours ago, msnbcorp said:

I was debugging an other mods when i run into these 2 lines :

[EXC 19:48:21.824] NullReferenceException: Object reference not set to an instance of an object
    principia.ksp_plugin_adapter.PrincipiaPluginAdapter.OnGUI ()
[EXC 19:48:21.826] NullReferenceException: Object reference not set to an instance of an object
    principia.ksp_plugin_adapter.PrincipiaPluginAdapter.OnGUI ()

I guess we should make an effort to fix these exceptions.  To the best of our understanding they don't cause problems, but it would be nicer to avoid spamming the logs and the UI.  I created an issue to track this.

[msnbcorp]

6 minutes ago, pleroy said:

I guess we should make an effort to fix these exceptions.  To the best of our understanding they don't cause problems, but it would be nicer to avoid spamming the logs and the UI.  I created an issue to track this.

Glad to be helpful ^^