v20 bugs and crashes

@Dancing Hoskuld
then you looking through logs this "modder message" as obstacle not giving any good so if it can be silenced logs will be more friendly?

i cant start any of map scripts....it is only me?
 
AIAndy's new music script is not working for me, now just playing the OLD menu music every time.

EDIT: Odd, now (rev. 1532) it works again...
 
OK in the NEW dll, i get this minidump and here is the Max Comp, i didnt even get 1 turn in after doing so??

The minidump shows it crashing in a call from the outcome code, but I can't see anythign obviously wrong. The turn plays fine for me, but I have not yet updated my assets to v20, so I don't have any outcome XML, so likely am not seeing the same circumstances as you are.
 
What directory do minidumps get stored in when there is a crash? sorry for the noobish question, but the game doesn't crash that much for me.

EDIT: Here is the save anyway, it experiences a repeatable CTD on End Turn.

View attachment 312034
The minidumps are saved in the directory with the BtS exe.

@Koshling: I can't see anything wrong either but I can reproduce the crash. Investigating with debug DLL now.

EDIT: Using a debug DLL the crash does not happen. The code also looks correct and it is a standard city loop.
 
RESMGR: Texture Eurasian_Jav_Thrower_GLOSS.dds failed to load
RESMGR: Texture Eurasian_Jav_Thrower_128.dds failed to load
RESMGR: Texture plantation2a.dds failed to load
RESMGR: Texture Wheat2.dds failed to load

ERR: FAudioManager::GetAudioTagIndex(): Could not find string AS3D_UN_FOOT_UNIT_LOW2 of TagId 4
ERR: FAudioManager::Do3DSound(): Illegal scriptId -1 attempted./n

finally started game with C2C_Global_Highlands

in logs i see WARN : EventSigns.addSign() gSavedSigns is not initialized!
but without CTD...will try playing

one more - in Checksum.log
many lines like this
[33590.992] (null) : 1672616655
it is ok?
 
The minidumps are saved in the directory with the BtS exe.

@Koshling: I can't see anything wrong either but I can reproduce the crash. Investigating with debug DLL now.

EDIT: Using a debug DLL the crash does not happen. The code also looks correct and it is a standard city loop.

If it's reliable for you non-debug try running that under the debugger. If you have the PDB and are prepared to switch to assembly-with-source view it's a pain but doable.
 
OK in the NEW dll, i get this minidump and here is the Max Comp, i didnt even get 1 turn in after doing so??

@Koshling: I can't see anything wrong either but I can reproduce the crash. Investigating with debug DLL now.

EDIT: Using a debug DLL the crash does not happen. The code also looks correct and it is a standard city loop.

If it's reliable for you non-debug try running that under the debugger. If you have the PDB and are prepared to switch to assembly-with-source view it's a pain but doable.

OK here's what i did for "my" CTD, i used the OLDer dll provided and it "proceeded" withOUT any problems what so ever, to the next turn, so i put back the current SVN dll and poof crash before end of turn:crazyeye:


EDIT: OK attached is the NEXT turn withOUT any problems if you can compare, dont know for sure, sorry. (All logs and Max Autosave included)

OK i am adding in the NEW minidump, here is what i did, i used the OLD dll to get to next turn (no problem), then i changed back to the NEW dll and poof crash again (attached second minidump) so it has to the the NEW dll that is doing it.

EDIT EDIT: OK proceeded now over 25 turns withOUT any problems using OLDer dll.
 
@SO: I believe you and I are experiencing the same crash, as when I took a look at the python debug log the last entry referred to an AI subduing a kangaroo (and therefore calling the Outcome code). Maybe one of Koshling's DLL changes today screwed up the Outcome system, but I'm no expert.
 
@SO: I believe you and I are experiencing the same crash, as when I took a look at the python debug log the last entry referred to an AI subduing a kangaroo (and therefore calling the Outcome code). Maybe one of Koshling's DLL changes today screwed up the Outcome system, but I'm no expert.

I haven't touched the outcome code at all, so I don't see how. However, once I get done with the current memory analysis I'll update to latest assets (so I'm using the outcome system myself) and try to reproduce these crashes.

@AIAndy/DH - is all the subduing now done by the outcome system and the old Python removed (or at least not called) in the latest SVN?
 
I haven't touched the outcome code at all, so I don't see how. However, once I get done with the current memory analysis I'll update to latest assets (so I'm using the outcome system myself) and try to reproduce these crashes.

@AIAndy/DH - is all the subduing now done by the outcome system and the old Python removed (or at least not called) in the latest SVN?
Yes, all the subdueing is now done by the outcome system. The Subdue Python code is not called.
 
These were the instances I was talking about earlier, thought it was initially a one time thing with Beijing, but I also noticed it in 2 other cities.

I don't know what the cause of it is though, and that is what bothers me.
 

Attachments

  • Civ4ScreenShot0005.JPG
    Civ4ScreenShot0005.JPG
    458.3 KB · Views: 59
  • Civ4ScreenShot0006.JPG
    Civ4ScreenShot0006.JPG
    466.5 KB · Views: 59
These were the instances I was talking about earlier, thought it was initially a one time thing with Beijing, but I also noticed it in 2 other cities.

I don't know what the cause of it is though, and that is what bothers me.

Do you have saves/autosaves from a bit before it happened? If we can get a save game in which it is about to happen (and does on a few turns autoplay) it should be easy to pin down.
 
@SO: I believe you and I are experiencing the same crash, as when I took a look at the python debug log the last entry referred to an AI subduing a kangaroo (and therefore calling the Outcome code). Maybe one of Koshling's DLL changes today screwed up the Outcome system, but I'm no expert.

Having updated to latest assets I can now reproduce SO's crash, so I'll see if I can figure it out when I get to C2C time later today (i.e. - after I'm done working)
 
Do you have saves/autosaves from a bit before it happened? If we can get a save game in which it is about to happen (and does on a few turns autoplay) it should be easy to pin down.

Unfortunately I don't, my autosaves are regular, however I tend to check every 100 turns, and when I notice this, I go through my autosaves but I can't find the before or after (since it autosaves every 4 turns or so.)

However, when it occurs it leaves the city in permanent no development, and things do not appear to get accomplished. :/

I'll be keeping a better eye on it though in case it appears again.
 
some more "log scavenging"

RESMGR: Texture minke.dds failed to load
RESMGR: Texture tiger_blink.dds failed to load
RESMGR: Texture AnimalTeeth.dds failed to load


Warning: building 0 is not associated with a CvCityLSystem node; it will not be visible!
building:Art/empty.nif does not contain a node called SHADOW. Its shadow, if present, will be corrupt!
Warning: Node Leaf_2x1 has no generic art matching constraint 180.000000
this warnings is normal ? i have many of them...

i have CTD,,,with some strange log message

Player 50 Unit 16409 (TXT_KEY_LEADER_BARBARIAN's Skunk) moving from 72:48 to -2147483647:-2147483647
to where he moved?

created new game - same situation...

Player 1 Unit 8193 (TXT_KEY_LEADER_CHIEF_SLEEPY_EYE's Stone Thrower) moving from 73:49 to 72:48
i changed to english only civ version for clean testing purpose.. but why this "text place holder" here?

will attach save if needed

messages like this :
info type not found
info type NONE not found
in xml.log is ok?

"Band of Homo sapiens sapiens" it is typo? or made for purpose?
 
Top Bottom