Common Errors and/or Crashes Section

[AIAndy]

Is this a known issue? Been reported before?

I updated all my graphics drivers at the introduction of V18. It occurs as you transition from city view to planet view.

Of course I was pretty sure the savegame wouldn't be looked at cause it's not SVN.
(why would the debug team keep a vanilla version around any way. "The" cool thing to do is SVN all the way baby! )

JosEPh :/

Actually I just tried it on the newest SVN version with debug DLL. Whatever the problem is, it happens both in the vanilla and the SVN version and it does not seem to be graphics driver dependent.
Might be a bug in the Civ4 graphics engine or some problem with a graphics model displayed on the screen at that time.

 

[Koshling]

Thanks a lot, Koshling, for your insightful description of the crash. I definitely get the impression that you and all the C2C team understand much more about this game than the Civ4 developers ever did!

Now I do have some serious issues with the attached save game:

1. Research is negative although it is at 80%. I did press ctrl-shift-r but I am not sure whether this issue already was present before. Some strange things go on here: See the attached screen shot of Cusco´s city view…
2. Many but not all of my Android workers never finish their jobs. Their turn count indicated by the mouse hover increases every turn instead of decreasing. You can observe this with the Androids trying to terraform near Huamanga. I have observed this before the current SVN but then only with many workers on one tile (30 or so), now it seems to happen already with a few workers.

This is one crazy game state - just shows that Galatic era needs some balancing work! Anyway:

Regarding the negative research:

1. The dominant contribution is from the broken production modifer. Ctrl-shift-R/T rebuild already fixed that.
2. Negative contribution from specialists turned out to be an AND bug, which I have now fixed in the latest SVN

Regarding the android workers:

• Each plot stores the progress towards each possible build type that is/has been worked on at that plot
• Each turn the total 'workage' towards a build is added to that build's progress in the plot
• To save space the progress was being stored in an array of shorts (16 bits), which were also signed (max value 32767)
• Total build costs scale with game speed. With your settings some of the terraforms were costly enough to push progress negative at which point all the arithmetic broke down
• I also discovered that the save formats for the plots were not remapping build ids, so these arrays could also get corrupted (and even be the wrong size, which would likely result in a crash later) across asset changes that added or removed builds

I have changed the progress to an int (won't be terribly expensive really) and fixed the save format, with dynamic conversion for old saves.

 

[Sleeeper]

xml.log spams with

[*] info type ATTACHABLE_CHIMNEYSMOKE not found, Current XML file is: xml\GameInfo/CIV4EspionageMissionInfo.xml

 

[JosEPh_II]

Actually I just tried it on the newest SVN version with debug DLL. Whatever the problem is, it happens both in the vanilla and the SVN version and it does not seem to be graphics driver dependent.
Might be a bug in the Civ4 graphics engine or some problem with a graphics model displayed on the screen at that time.

Thank you Andy for looking at it. I'm under the weather today and at home so I got a bit sarcastic in my reply, no excuse really so I apologize.

JosEPh (Old and Slow.....and cranky most of the time....so Watch Out!)

 

[LeAto]

I'm having a CTD issue with release 18. I just downloaded and installed 18 for the first time. Previous version was 15. I removed 15 off of my system and copied over the folder for 18. When ever I attempt to start a new game with any combination of settings it seems I CTD with a runtime error pointing to the the base bts civ4 exe. There is no other information given other than its a runtime error directed at the base exe. I'll provide any further information I can just ask it of me. Many thanks, and thanks for the effort this has been a solid mod and can not wait to try 18.

~ LeAto

 

[Dancing Hoskuld]

I'm having a CTD issue with release 18. I just downloaded and installed 18 for the first time. Previous version was 15. I removed 15 off of my system and copied over the folder for 18. When ever I attempt to start a new game with any combination of settings it seems I CTD with a runtime error pointing to the the base bts civ4 exe. There is no other information given other than its a runtime error directed at the base exe. I'll provide any further information I can just ask it of me. Many thanks, and thanks for the effort this has been a solid mod and can not wait to try 18.

~ LeAto

Try clearing the cache, by holding down the shift key while loading the game. It may be that your cache has some bits from v15 in it still.

 

[LeAto]

That worked. Many thanks!

 

[Snofru1]

This is one crazy game state - just shows that Galatic era needs some balancing work! Anyway:

Regarding the negative research:

1. The dominant contribution is from the broken production modifer. Ctrl-shift-R/T rebuild already fixed that.
2. Negative contribution from specialists turned out to be an AND bug, which I have now fixed in the latest SVN

Regarding the android workers:

• Each plot stores the progress towards each possible build type that is/has been worked on at that plot
• Each turn the total 'workage' towards a build is added to that build's progress in the plot
• To save space the progress was being stored in an array of shorts (16 bits), which were also signed (max value 32767)
• Total build costs scale with game speed. With your settings some of the terraforms were costly enough to push progress negative at which point all the arithmetic broke down
• I also discovered that the save formats for the plots were not remapping build ids, so these arrays could also get corrupted (and even be the wrong size, which would likely result in a crash later) across asset changes that added or removed builds

I have changed the progress to an int (won't be terribly expensive really) and fixed the save format, with dynamic conversion for old saves.

Excellent solution, works perfectly now!

Concerning balancing the Galactic Era: I could have finished that game in Renaissance Era after being able to cross the oceans. I just keep it going to play through until the end of the technology tree and get the full feeling of this amazing mod. And maybe find one bug or the other. So probably Galactic doesn´t need immediate balancing, whatever problems the AI maybe had happened long before. But of course there are many empty slots in Galactic still.

 

[Hydromancerx]

I found a graphics error, the Blackwood Archer (Tupi UU) has a big red dot for its model.

 

[strategyonly]

I found a graphics error, the Blackwood Archer (Tupi UU) has a big red dot for its model.

Thx, fixed.

Also Koshling: Big normal autosaved error (at least for me that is)

http://forums.civfanatics.com/showpost.php?p=11033479&postcount=4454

 

[LeAto]

That worked. Many thanks!

Well it worked for one run of the game. I tried clearing the cache again and now I'm getting a CTD on startup of the game. I removed all the files again, a fresh reinstall of civ4 and I get 1 run before I get a CTD before the main start selection screen. Again its a runtime error with no details.

 

[Koshling]

Thx, fixed.

Also Koshling: Big normal autosaved error (at least for me that is)

http://forums.civfanatics.com/showpost.php?p=11033479&postcount=4454

Fix pushed to SVN. It won't rescue non-max-compat saves from yesterday, but regular format will work again in this build (and max compat saves from whenever should still be fine)

 

[Dancing Hoskuld]

@Koshling, I am getting another error in the autologEventManager.py to do with attitude, but in a different place. I will look at what you did and see if I can fix it.

Traceback (most recent call last):
  File "BugEventManager", line 363, in _handleDefaultEvent
  File "autologEventManager", line 352, in onEndGameTurn
  File "autologEventManager", line 1513, in checkStuff
RuntimeError: unidentifiable C++ exception

 

[strategyonly]

@AIAndy/DH:

When a "barb civ" in Rev, comes into play, it "automatically" takes over that city withOUT a fight, now thats just not right, and sends the "old" occupied player(s) out of the city by 1 tile??

Sending an example of such, ONE turn before the "Occupation" of the revolting civ.

btw Barb Civ and Barb World are off.

EDIT: Now this is just weird, i restarted from a normal autosave and now the city doesNOT revolt? (StrategyOnly)

EDIT EDIT: OK its get even weirder now, a few more turn like (4) it revolts again but this time it say it has 2 turns till revolt but revolts right away and takes over the city??

Again i went back the the normal savedgame, so i cant say for sure whats going to happen this time???? (Prehistoric)

 

[Koshling]

@Koshling/AIAndy:

When a "barb civ" in Rev, comes into play, it "automatically" takes over that city withOUT a fight, now thats just not right, and sends the "old" occupied player(s) out of the city by 1 tile??

Sending an example of such, ONE turn before the "Occupation" of the revolting civ.

btw Barb Civ and Barb World are off.

EDIT: Now this is just weird, i restarted from a normal autosave and now the city doesNOT revolt? (StrategyOnly)

EDIT EDIT: OK its get even weirder now, a few more turn like (4) it revolts again but this time it say it has 2 turns till revolt but revolts right away and takes over the city??

Again i went back the the normal savedgame, so i cant say for sure whats going to happen this time???? (Prehistoric)

I think that's just one of te intended possibilities or REV isn't it? Anyway, it's all in the REV Python which I don't really know much about. It's not a DLL issue

 

[strategyonly]

I think that's just one of te intended possibilities or REV isn't it? Anyway, it's all in the REV Python which I don't really know much about. It's not a DLL issue

Then we will leave it to AIAndy or DH, maybe they can figure out whats going on

Anyways here is another one same city, but i am not taking any chances this time, i an posting a Max Save, exactly when the revolt has started?

 

[Stormwind]

Is there any way to remove "Trade blocked by enemy ship" squares in worldbuilder ? An enemy ship was blockading but in the next round I captured their last city, the ship vanished but the plots remain blocked.

 

[Dancing Hoskuld]

Is there any way to remove "Trade blocked by enemy ship" squares in worldbuilder ? An enemy ship was blockading but in the next round I captured their last city, the ship vanished but the plots remain blocked.

No unfortunately. I thought that the problem had been fixed quite awhile back in AND. If you post a max compat save I am sure Koshling or AIAndy will be able to look at it.

 

[ls612]

I've just recieved this 'rare' PW3 bug four times in a row.

I think there is a problem with the perfectworld3 mapscript.

 

[AIAndy]

I've just recieved this 'rare' PW3 bug four times in a row.

I think there is a problem with the perfectworld3 mapscript.

I just pushed a fix to at least one problem in that code and also removed the try/except that displayed that generic popup. So if there is another problem in there, it should display a real error message now.