Bug Reports and Technical Issues

[Leoreth]

Hi Leoreth,

I have a game crashing at the end of turn 339, playing the Egyptians. I replayed the last few turns different ways, it always crashes.

Possible reasons (dunno, just a guess):

-The borders of Jerusalem expand at turn 339, right before the crash, in a strange way, two plots below the destroyed city of Babilu.
-I had issues with the Chinese in previous games. I couldn't enter the world builder for about 30 turns at some point in my game playing Babs (game crashed when hitting ctrl+W)), and then it was possible again. My scout during this Babs game couldn't reveal the map in China, around the city of Luoyang. The game crashed when I tried.

Here is my file, if ever you can take a look at it. The game version is 1.16.

Does not crash for me, any special settings? With VD?

 

[Dan Quayle]

Does not crash for me, any special settings? With VD?

Nope, not that I know of. I haven't touched any of the settings in the options menu (and of course haven't touched the files themselves), except screen resolution.

 

[bluepotato]

Hi Leoreth,

Yesterday I found bug while playing on commit 4ec00cfd6f9461e63a0ce263048a894598135041, under Linux+wine-6.3.
It manifests first as a Python error and then a turn later a crash. Stack trace of the Python error:

Traceback (most recent call last):
  File "BugEventManager", line 378, in _handleDefaultEvent
  File "Events", line 12, in handler_func
  File "Stability", line 58, in checkScheduledCollapse
  File "Stability", line 615, in completeCollapse
  File "Stability", line 569, in secedeCities
  File "Stability", line 599, in secedeCity
  File "RFCUtils", line 762, in completeCityFlip
  File "RFCUtils", line 364, in createGarrisons
  File "Core", line 468, in makeUnits
RuntimeError: unidentifiable C++ exception

Based on my quick investigation, this seems to happen because completeCityFlip in RFCUtils calls flipCity (line 752 in RFCUtils.py):

flipCity(plot, False, False, iPlayer, [iOwner])

but then uses tPlot, which is the now invalidated city, if bFlipUnits is false:

if bFlipUnits: 
	flipUnitsInCityAfter(plot, iPlayer)
else:
	createGarrisons(tPlot, iPlayer, 2)

which ultimately results in makeUnits trying to spawn units at coordinates like XY:636,-2147483648.

My naive solution was to change the flipCity line to this (flipCity just so happens to return the new city):

tPlot = flipCity(plot, False, False, iPlayer, [iOwner])

which fixed the Python error and subsequent crash on the next turn for me.

(I've attached a save file in case you need it.)

 

[Leoreth]

Oh thanks. I had seen this bug a couple of times when I was working on something else but haven't had time to investigate it.

 

[Enyavar]

Hi, I just think it might be a bug: Out of nowhere I get a +20 bonus on foreign stability, when I have just -1 from unstable Tibet, in the turn before.

Some things had changed in the previous turn: My oldest friend the Tamil Empire collapsed (they were at non-fought distant war with Spain+England); I also made a trade deal with the US in turn 393. The next turn, 394, I progress to the Global Age (as the first I think, not 100% sure about German progress) and I complete Hollywood. If any of those things explain that sudden diplo bonus, I'm curious how to do that in other games since a positive relationship bonus is certainly needed late in the game.

Spoiler Screenshot :

See attached savegame (DoC v.1.16.3) if needed - it is from the turn before when foreign stability was still at -1.
Just found that sudden change very suspicious; I had a similar situation once before in the same game but thought little of it then.

 

[freethink]

The year is always 1991 or 1984, or 1991, and the great Romania has always been at war with Oceania and friends with Eurasia.
Always

 

[freethink]

However, six years later in the year of our lord 1997, this all came to an end. Yet the Great Rome has stood the test of time (or the test of 1991)

 

[Leoreth]

View attachment 589236

The year is always 1991 or 1984, or 1991, and the great Romania has always been at war with Oceania and friends with Eurasia.
Always

Should be fixed.

View attachment 589237 However, six years later in the year of our lord 1997, this all came to an end. Yet the Great Rome has stood the test of time (or the test of 1991)

This doesn't help me. Please post the full stacktrace from PythonErr.log or a save where I can trigger this error.

 

[Dominus the Mentat]

After the aztecs fall, most of their cities flip to Spain without never been colonized by them. Is that intended?
P.S.: I was playing with the aztecs before chose to play with Colombia. I have left the aztecs at war with Spain when I tried to capture slaves from spanish military units, which I have get none. Save attached. Aztecs fall in the next turn.

Another report is those python exceptions that have show up after Colombia spawn. Images attached.

 

[ariel552]

Hi Leoreth,
I hope you are doing well. I'm always very grateful about the game you did, it's the best in all history, really! However I found a bug that crashes the game the next turn, attachment below, happens with all civs in the XX century with the fix aswell. Thank you!!!

 

[ariel552]

 

[Fresol]

Playing as China, when I conquer the city of Kyouto it says "You have failed goal 1 of your historical victory!"

Well, the goal is cathedral building and has nothing to do with Japan losing a city.

Game version is 1.16.

 

[merijn_v1]

Playing as China, when I conquer the city of Kyouto it says "You have failed goal 1 of your historical victory!"

Well, the goal is cathedral building and has nothing to do with Japan losing a city.

Game version is 1.16.

This is caused by a bug in the Japanese UHV. The code accidentally makes the attacker lose the UHV instead of Japan. It is already fixed in the development version.

You can reset the victory to its old status in the WB to continue your game.

 

[HighFunctioningAlcoholic]

Super unimportant and probably not worth your time, but I should be able to steal techs from Russia according to the espionage screen, but it's not an option. It's Polynesia, and I did extend the end date through the World Builder (if that might factor), but I can still steal techs from other civs.

(Most recent update, minus whatever you did with Japan.)

 

[HighFunctioningAlcoholic]

Super unimportant and probably not worth your time, but I should be able to steal techs from Russia according to the espionage screen, but it's not an option. It's Polynesia, and I did extend the end date through the World Builder (if that might factor), but I can still steal techs from other civs.

(Most recent update, minus whatever you did with Japan.)

As a follow-up, the game is starting to bug-out a bit. I'm able to draft units despite not using the "Nationhood" civic, and if I conquer a city and immediately liberate it to another civ, it still asks me what I want to build in that city.
All these errors are post-changing the end date, so, I say knowing nothing about the code, it might be related. I wouldn't waste your time - who plays Polynesia into 2027?

 

[1SDANi]

Settling a city as India doesn't found Hinduism. Is this intentional?

EDIT: I can't reproduce it anymore. I guess it was specific to the game I rolled?

 

[Enyavar]

Huh. Japan on the home stretch in 1965, still in the same game as the error I reported on Monday.

But that war that I tried to ignore here, blows up my game. To replicate in 1.16.3, please just click the turn button, then it throws errors. It has something to do with the secession of some civ, and another turn later again, the game simply crashes.
(Couple of Python Errors)
I can provide older savegames if so needed.

(I will engineer my way around the whole war I got currently in: I'm going to reload a save from 1957, station a bunch more troops in my allies' cities, and just finish the UHV without war against the Iberians).

Edit: Thanks for looking into it Leo! In the meantime, I did just like I wrote and finished the Japanese UHV in 1970. Whew.

 

[Leoreth]

I clicked end turn on develop and got no errors, pretty sure I have fixed this already.

 

[Fresol]

In latest development version AI is willing to give away surplus strategic resource(iron, horse etc.) for free, when you already have a copy of it. There could be exploits around it.

 

[Enyavar]

* Byzantium
* Spawn: 330 AD at Constantinople
* Conditional: requires Greece to be dead and Rome to be alive and unstable

Pre-Bugreport question: Is that textfile up to date, or should I report it as a bug when I am playing a solid Rome and Byzantium spawns to take away half my empire?