I am playing multiplayer via internet with a friend. Vanilla, first time on 1.1.13. Game was started in 1.0, switched to experimental in time for 1.0.9. I set up the game back then, and we are taking turns hosting because the non-host experience is not as good as the host's experience. This time, he was the host.
We started around 8 p.m. CET yesterday, at which time I saved my LazyBastard singleplayer game (as you can see in the logs). We decided to call it a day at around 3 a.m. CET. He saved and quit. I clicked "Save" in the popup, chose a file name and saved. At more than half-way through, the progress bar stopped. When I alt-tabbed out, I found that behind the Factorio window, a windows window had popped up with the message that the game had crashed.
Find attached the incompletely written savegame and the log. If you give me a hint where I can find the crash dump mentioned in the log, I can provide it to you as well.
[1.1.13] Crash On SaveAfterServerLeft (!hasReachedZeroConnection was not true)
[1.1.13] Crash On SaveAfterServerLeft (!hasReachedZeroConnection was not true)
- Attachments
-
- factorio-1.1.13-crash-on-save.zip
- (96.81 MiB) Downloaded 120 times
-
- factorio-1.1.13-crash-on-save.log
- (113.02 KiB) Downloaded 98 times
Re: [1.1.13] Crash On SaveAfterServerLeft (!hasReachedZeroConnection was not true)
Well, this is a consistency check related to electric pole not having proper state. The idea is, that better to crash when saving, than saving corrupted save.
I extended the crash info to say what entity/position is it related to, but other than that I would need to figure out how it can happen.
If you have any idea how you could reproduce it, I would welcome it, but until than, I need to put it into pending.
I extended the crash info to say what entity/position is it related to, but other than that I would need to figure out how it can happen.
If you have any idea how you could reproduce it, I would welcome it, but until than, I need to put it into pending.
Re: [1.1.13] Crash On SaveAfterServerLeft (!hasReachedZeroConnection was not true)
Also, from you description it seems that your (hosting) friend's save attempt was successful? If yes, could you also include that save file?
Also, as kovarex said, the crash is caused by attempting to save an inconsistent state, so if your friend's save was successful, the inconsistency must have occurred between those save attempts - if you remember what was happening, it would be of great help.
Also, as kovarex said, the crash is caused by attempting to save an inconsistent state, so if your friend's save was successful, the inconsistency must have occurred between those save attempts - if you remember what was happening, it would be of great help.
Re: [1.1.13] Crash On SaveAfterServerLeft (!hasReachedZeroConnection was not true)
Thank you for looking into it, will attach the save as soon as my friend comes back to me.
Remember what was happening is easy: I didn't touch any controls after saving had started. I probably was browsing reddit on my second screen. Next time I looked at my factory, I was greeted by the "Server has quit" screen. Last thing we did before both saves, was waiting out the attack waves on an artillery outpost. The waves did not damage any power poles, as far as I can remember. We did not have pending builds (ghosts).
I remember one power pole oddity however, that may be related:
When we started that save, we had one 500spm factory laid out with blueprints that was never completed (>1k ghosts, no resource inputs) ever since 1.0 (named "MegaFactory 3" in the save). I had since put down the very same blueprint in 1.1.x in newly liberated area ("MegaFactory 4") and had the robots build it and the poles had auto-connected.
During this run, I went back and had my personal robots and spidertron robots complete it. Just that about half a dozen of the power poles didn't auto-connect to the others, and I had to manually connect them for the factory to spring into action.
Maybe I can try whether that reproduces the issue, but probably only next weekend. The new log message is already part of 1.1.18 I assume?
Remember what was happening is easy: I didn't touch any controls after saving had started. I probably was browsing reddit on my second screen. Next time I looked at my factory, I was greeted by the "Server has quit" screen. Last thing we did before both saves, was waiting out the attack waves on an artillery outpost. The waves did not damage any power poles, as far as I can remember. We did not have pending builds (ghosts).
I remember one power pole oddity however, that may be related:
When we started that save, we had one 500spm factory laid out with blueprints that was never completed (>1k ghosts, no resource inputs) ever since 1.0 (named "MegaFactory 3" in the save). I had since put down the very same blueprint in 1.1.x in newly liberated area ("MegaFactory 4") and had the robots build it and the poles had auto-connected.
During this run, I went back and had my personal robots and spidertron robots complete it. Just that about half a dozen of the power poles didn't auto-connect to the others, and I had to manually connect them for the factory to spring into action.
Maybe I can try whether that reproduces the issue, but probably only next weekend. The new log message is already part of 1.1.18 I assume?
Re: [1.1.13] Crash On SaveAfterServerLeft (!hasReachedZeroConnection was not true)
I actually fixed a bug that is most probably related for 1.1.19, so I suggest using that (another stable candidate), my guess is, that your bug was related to the thing I fixed.
So I'm moving it to resolved bug until proven otherwise.
So I'm moving it to resolved bug until proven otherwise.