Bug Reports and Technical Issues

[Shadow Merc]

At some point in 1.12, I changed the conditions for releasing civs so that they are the same as for a civ being able to respawn. This means:
- impossible ten turns after spawn or respawn
- impossible 20 turns after dying
- impossible when an incompatible other civ is alive (e.g. Rome and Italy, Greece and Byzantium)
- only possible during historically appropriate time periods

Thanks, this is helpful. I was worried that it was accidentally removed after the whole core-disappearing after collapse bug.

 

[Leoreth]

Crash to desktop at the start of the attached autosaved turn. I replayed from a few turns before with the same result - a save from the previous turn is attached. Constantinople fell to the Turks ~10 turns before, so perhaps a Byzantine collapse. Or possibly related to a Heavy Swordsman in the Middle East that has nowhere to go when the Turks advance.

SVN 810. Why oh why do I give in to the temptation to play the SVN....?!

You were close with the last guess. The crash is caused by the Moorish Great Prophet in Cordoba (Uthman ibn Affan), I still don't know how to fix it, but you can delete the unit if you still want to continue this game.

 

[Fresol]

I'm getting CTD at turn 36 in my China game, here's the save before the crash. SVN 820.

 

[The Diocletian]

Batavia (on the Coffee in Java) does not get the correct tile yield, despite being an island of the requisite size. Colombo (on the Iron in Sri Lanka) does get the correct tile yield, however.

Spoiler :

edit: Should probably attach a save.

 

[Leoreth]

Yeah, I increased the island size threshold from 3 to 5 (as mentioned), but somehow that change didn't wind up being committed yet. It'll come with the next revision.

 

[Jarlaxe Baenre]

I am playing on a mac through wine, though I'm pretty sure that's not why this is happening.
In 1.12, the third Colombian goal (gold through selling resources by 1950) fails when a dead civilization respawns. I was able to fix it by commenting out two lines to do with the mayan astronomy goal. (else: fail the goal). The goal doesn't fail when a civ respawns. However, now a bunch of python exceptions appear when the civ is reborn.

Traceback (most recent call last):
  File "BugEventManager", line 400, in _handleDefaultEvent
  File "CvRFCEventHandler", line 743, in onTechAcquired
  File "Stability", line 94, in onTechAcquired
  File "Stability", line 339, in checkStability
  File "Stability", line 366, in triggerCrisis
  File "Stability", line 376, in territorialCrisis
  File "Stability", line 700, in secedeSingleCity
  File "Stability", line 602, in secedeCities
  File "Stability", line 2007, in resurrectionFromCollapse
  File "Stability", line 2062, in doResurrection
RuntimeError: unidentifiable C++ exception

They don't seem to have caused any noticeable problems, or even be related except for the tech thing (I had not just acquired a tech, so I think it's the new civ). The goal failing seems to be because the goal is achieved if the player is both mayan and not reborn, and is failed in any other cases — if they're mayan and reborn, for instance.

 

[Dan Quayle]

Playing the US I can't access the WB anymore, immediately after moving to turn [July 1883]...so far I could enter it anytime, but as soon as I enter this turn it crashes to desktop when I ctrl+W, before doing anything.

I wonder if it has to do with the great writer ranking civs popup at the beginning of turn [July 1883], which I've looked to disable but can't find out how.

Also, to try and isolate the problem, (obviously something happening during the AI turns) I erased all eurasian cities on the previous turn [jan 1883] and it worked...but obviously this is not a solution...

Here's the Jan 1883 save, if you try it out, you'll see you can enter WB, but not once you get to the next turn. WOuld there be any way to check out which AI unit/building, etc. does this ?

 

[Leoreth]

I am playing on a mac through wine, though I'm pretty sure that's not why this is happening.
In 1.12, the third Colombian goal (gold through selling resources by 1950) fails when a dead civilization respawns. I was able to fix it by commenting out two lines to do with the mayan astronomy goal. (else: fail the goal). The goal doesn't fail when a civ respawns.

Can you tell me where you changed those lines? I thought I had fixed everything related to the Mayan goal.

Playing the US I can't access the WB anymore, immediately after moving to turn [July 1883]...so far I could enter it anytime, but as soon as I enter this turn it crashes to desktop when I ctrl+W, before doing anything.

I wonder if it has to do with the great writer ranking civs popup at the beginning of turn [July 1883], which I've looked to disable but can't find out how.

Also, to try and isolate the problem, (obviously something happening during the AI turns) I erased all eurasian cities on the previous turn [jan 1883] and it worked...but obviously this is not a solution...

Here's the Jan 1883 save, if you try it out, you'll see you can enter WB, but not once you get to the next turn. WOuld there be any way to check out which AI unit/building, etc. does this ?

I'll look into it.

 

[Jarlaxe Baenre]

2947 and 2948, but the problem is 2945.

 

[Dan Quayle]

Concerning my ctd problem when entering WB, I retried the erasing of all cities in Africa, Europe, Asia and Australia in Jan 1883, and I confirm it solves the problem: after moving to the fatal turn July 1883 I can successfully enter the WB mode.

However, if I try to zero closer to the culprit, by erasing cities/destroying civs, civ by civ, on those continents, I'll end up with the same darn ctd...

by the way, I haven't found anywhere a way to disable "the world's greatest civs" ranking pop up. Is it possible to do so, and would that be responsible for the crash when entering WB ? seems unlikely...

may I also suggest, and this is valid only if it eventually saves time and energy for Leoreth and other modders who look into these problems, to have a manual to spot out these problems ourselves ? I can imagine though no bug is recurrent enough to justify this, but this is of course just a suggestion !

 

[SeekTruthFromFacts]

You were close with the last guess. The crash is caused by the Moorish Great Prophet in Cordoba (Uthman ibn Affan), I still don't know how to fix it, but you can delete the unit if you still want to continue this game.

Thank you for another speedy and piece of bugfinding.

Swat!

 

[Leoreth]

Unfortunately I don't understand how the bug actually came to be so I cannot prevent it from occuring in the future. I hope it's a rare technical problem because otherwise there's trouble.

 

[Leoreth]

I'm getting CTD at turn 36 in my China game, here's the save before the crash. SVN 820.

Do you have a save from one turn before? That save crashes before I can open it.

Playing the US I can't access the WB anymore, immediately after moving to turn [July 1883]...so far I could enter it anytime, but as soon as I enter this turn it crashes to desktop when I ctrl+W, before doing anything.

I wonder if it has to do with the great writer ranking civs popup at the beginning of turn [July 1883], which I've looked to disable but can't find out how.

Also, to try and isolate the problem, (obviously something happening during the AI turns) I erased all eurasian cities on the previous turn [jan 1883] and it worked...but obviously this is not a solution...

Here's the Jan 1883 save, if you try it out, you'll see you can enter WB, but not once you get to the next turn. WOuld there be any way to check out which AI unit/building, etc. does this ?

Can you tell me which revision this is on? I can't open the save on 820.

I am playing on a mac through wine, though I'm pretty sure that's not why this is happening.
In 1.12, the third Colombian goal (gold through selling resources by 1950) fails when a dead civilization respawns. I was able to fix it by commenting out two lines to do with the mayan astronomy goal. (else: fail the goal). The goal doesn't fail when a civ respawns. However, now a bunch of python exceptions appear when the civ is reborn.

Traceback (most recent call last):
  File "BugEventManager", line 400, in _handleDefaultEvent
  File "CvRFCEventHandler", line 743, in onTechAcquired
  File "Stability", line 94, in onTechAcquired
  File "Stability", line 339, in checkStability
  File "Stability", line 366, in triggerCrisis
  File "Stability", line 376, in territorialCrisis
  File "Stability", line 700, in secedeSingleCity
  File "Stability", line 602, in secedeCities
  File "Stability", line 2007, in resurrectionFromCollapse
  File "Stability", line 2062, in doResurrection
RuntimeError: unidentifiable C++ exception

They don't seem to have caused any noticeable problems, or even be related except for the tech thing (I had not just acquired a tech, so I think it's the new civ). The goal failing seems to be because the goal is achieved if the player is both mayan and not reborn, and is failed in any other cases — if they're mayan and reborn, for instance.

Looks like you commented out the tech assignment on respawn, that is bound to cause trouble further down the road. The UHV problem will be fixed with the next commit.

 

[Dan Quayle]

I can't find where to check the revision number, I've only found the txt file 1.13, and updated with svn on October 2nd, but I'll give you the number of course when I know where to find it...

 

[Leoreth]

October 2nd should be enough to figure it out, but you can always right-click the mod folder and use Tortoise SVN -> Show Log, the current revision is the one in bold.

 

[Dan Quayle]

right, I read in bold in the log box you've indicated :

"Revision: 794
Author: leoreth
Date: jeudi 2 octobre 2014 01:47:31
Message:
- increased the number of immigration events
- York is now properly renamed to Toronto
----
Modified : /Assets/Python/CityNameManager.py
Modified : /Assets/Python/StoredData.py
Modified : /Assets/Python/UniquePowers.py
Modified : /Assets/XML/Text/CIV4GameText_RFCHints.xml"

thanks by the way to look at this, I've enabled logging and disabled "hide exceptions" in the civilizationIV.ini file as indicated in your initial post on this thread, but there is no in game message spotting a problem from turn Jan to July 1883...

 

[Leoreth]

Alright, thanks. If I have time I'll look at it tomorrow.

 

[Dan Quayle]

If this can help, I may point out that my civ is at war with three rivals, on turn Jan 1883. Whether I make peace or not with all three (on turn Jan or July 83), it won't change anything, still can't access WB on turn July 83.

However, if I make peace with all three on July 82, WB is not accessible as soon as Jan 83 !

This very unusual ctd when accessing WB seems more like a general breakdown that's bound to happen sooner or later...I've had CTDs from time to time, but occuring during the actual game, and so far I managed to isolate and fix these problems "ingame" (precisely by entering WB .

Here, the closest "solution" I found was to erase a certain number of cities following a curve from Mecca to Sapporo, on Jan83, which allowed me to enter WB on July83. A game conflict related to interwining borders in those areas maybe ?

Also, none of my units are under attack from jan to july83 and I forced universal peace with the diplomacy tool in WB on turn jan83 on all civs, dead or alive, including indies, natives, etc, and haven't spotted barb units which may provoke this, to no avail. So my guess is that its not a battle or any war event between AIs that cause this either...

btw, I've recently changed graphic cards and got meself a Asus Geforce GT 640, 2Gb memory. I guess that would more be the problem if you eventually find out that you can open WB on turn July83...

 

[Fresol]

Do you have a save from one turn before? That save crashes before I can open it.

Here's another save.

 

[Leoreth]

Thanks.