Crash to unity shortly after disconnected stage exploded in atmosphere (null vessel name) #4026
Comments
|
it isn't just caused by parallax and I have seen it crash in FAR dragcube functions, i have no idea where the issue lies - your KSP.log has some from scatterer and kerbalAlarmClock functions |
|
i've seen me and one other person get a similar unity crash error, my KSP.log shows that it seems to be some sort of null reference, i have had the same exact crash from a disconnected stage desintegrating but only after loading a save |
|
Thanks for reporting. This looks similar to #3872, which we could not investigate because we didn't have a crash dump. Here we have a crash dump, so there is hope. |
|
Decoded stack trace: Principia/ksp_plugin/interface.cpp Lines 886 to 890 in 9fb3103 Opening the dump under the debugger shows the following crashing stack: with:
Since we haven't seen this before, and this code has not changed in a long time, and KSP has not changed either (duh!) I am going to go out on a limb and theorize that there is a mod somewhere that sets the |
|
(Not similar to #3872 after all.) |
|
Hm, I can't think of any mod that I have that would rename a vessel. Just in case, I'll drop my mod list here. (not using RP-1 express)
|
|
Not really fixed, let's see if this reoccurs and what information we get. At least it won't be a Unity access violation. |
|
I'm using RP-1 express, except for Tundra launchpads which were causing a CKAN error.
|
|
Here's what is in common:
|
|
To be clear, the νεῶν κατάλογος of all your mods is not useful because I am not going to debug someone else's mod, not to mention 100 mods. After #4027, which will go in the July release, we'll know the GUID of the culprit vessel. That alone won't tell us much, but what I am hoping is that we'll find the same GUID somewhere in the Principia logs and we might be able to attribute it to a mod. Or not, it's entirely possible that we never get to the bottom of this. It would not be the first time that a mod would violate an invariant of KSP that we depend upon, see e.g., #2419. In some cases we were able to work with the mod author to get the problem fixed, but not always. |
|
Interesting, I'll be sure to report if I get another crash after the july release. |
Note that if this theory is correct, there is no telling in what state the vessel is left, and there is no guarantee that papering over the missing name would result in a vessel that conforms to the invariants expected by Principia. |
pleroy
changed the title
|
Haven't heard anything since Kleene, so closing. |
The folder below has all of the relevant crash logs. No fatal log was present. Frame set to ECEF.
https://drive.google.com/drive/folders/1cs6egUKGd7qV-89hH3CCd2EapnSSYrZe?usp=sharing
Error.log says that "principia.DLL caused an Access Violation", so I think its principia that caused the crash.
Running the mission again gave no issues.
The text was updated successfully, but these errors were encountered: