Factorio Forums

It's happening again.

Here's the scenario folder, saves folder, factorio-current.log, and temp/currently-playing folder.

https://www.dropbox.com/s/9bde0ekmm67a3 ... g.zip?dl=0

When I started the server after updating to 0.16.16, it loaded the 0.16.15 save file (factoriorpg56-railworld.bak) and wasn't working, so I migrated the save file on my own computer to 0.16.16 and saved and uploaded it to the server (factoriorpg56-railworld.zip). Still isn't running any of the scenario scripts.

Well it looks like that isn't enough to reproduce it either. But, I've noticed something odd in the log file:
This is odd because there isn't supposed to be any file called “2018-01-04_Factorio_0.16.13.lua” in the base game. We have 2017-12-03 and then 2018-01-09. But it appears that even if I create that file with just `game.reload_script()` in it, it still doesn't reproduce the problem. Still – could you post your data/base/migrations/2018-01-04_Factorio_0.16.13.lua? Also, any idea where you got that file?

I found the bug. Psihius's server management tool was overwriting new files onto the folder and every time he'd update it would grab the all of those old files, including the migration script.

Hm, we have a web management tool done by 3RA, they have a version manager - they put each version of installed Factorio into /usr/share/factorio/x.xx.xx/... folder. But what's weird, the RPG server is on our multienv tool and that binary lives at /opt/factorio/factorio - completely separate from the multienv tool.

psihius wrote:

Code: Select all

root at factoriommo in /opt/factorio
$ cat /usr/share/factorio/0.16.15/data/base/migrations/2018-01-04_Factorio_0.16.13.lua
game.reload_script()

root at factoriommo in /opt/factorio

Hm, we have a web management tool done by 3RA, they have a version manager - they put each version of installed Factorio into /usr/share/factorio/x.xx.xx/... folder. But what's weird, the RPG server is on our multienv tool and that binary lives at /opt/factorio/factorio - completely separate from the multienv tool.

Nevertheless, it looks like a server setup issue. It is still a bit weird because, as I said, even if I create the suspicious migration file, it doesn't reproduce the problem for me, so I guess some other files must be replaced from somewhere.

I'll call this not a bug. If you can find a way to reproduce this by taking the vanilla tarball and perhaps replacing some files in it, and if you think it is reasonable to expect the game to work with those files replaced, I'll take another look.