Single Player bugs and crashes v36 download - After the 24th of October 2015

[voltar]

Will take a look when I can. Got a bit of a debug workload building up here. Is this on stock v36 or SVN? Just to be clear.

It is v36

 

[Thunderbrd]

It is v36

It'll take me a little longer to look into this one because it is stock v36 and being a mini it means I'll need to get a stock v36 fileset here. I can't promise that I can fix it since it's a mini without the full compile files but it could still illuminate a problem spot. I'll try.

Been needing a stock v36 fileset on the system here anyhow.


EDIT: yeah, so the mini does give some information but what it's trying to say the actual crash is I'm entirely uncertain. I'm sure we'll run across this again at some point though and if it is repeatable and a save can be loaded and given a deeper evaluation to the crash then it should be easier to sort out. It's unfortunate but I just am not getting enough info from the mini to truly get a grasp on what the problem is. My apologies.

 

[Dancing Hoskuld]

Added a comment in the first post on how to turn off the Lotus Temple so as to stop it causing a crash.

 

[voltar]

It'll take me a little longer to look into this one because it is stock v36 and being a mini it means I'll need to get a stock v36 fileset here. I can't promise that I can fix it since it's a mini without the full compile files but it could still illuminate a problem spot. I'll try.

Been needing a stock v36 fileset on the system here anyhow.


EDIT: yeah, so the mini does give some information but what it's trying to say the actual crash is I'm entirely uncertain. I'm sure we'll run across this again at some point though and if it is repeatable and a save can be loaded and given a deeper evaluation to the crash then it should be easier to sort out. It's unfortunate but I just am not getting enough info from the mini to truly get a grasp on what the problem is. My apologies.

Dang I foolishly deleted the saves, but yeah it was repeatable, I got to a point where the game would crash during AI turn every time, couldn't figure out a way out...

 

[Thunderbrd]

Well... if we find it again we can address it then.

 

[BigSmac]

We do really need to find such a situation and capture a save from just before placing such a city. The only thing it looked like it could have something to do with on the one example I've seen the map on was that a Volcano was adjacent to the city when placed. However, I've recently placed a city adjacent to a volcano and didn't have that issue then either.

Could it be some kind of xml flaw in a build/feature perhaps? I THINK there's some kind of bonus amount of production that comes from placing a city on forests isn't there? Could this mechanism be screwed up severely somewhere?

I have such a save game for you. It happened earlier in this game but I missed it then. The city I am supposed to create is centered right on the salt. What I did is, I built the city, then I made a que of production buildings (all i was able to built right after founding) the next turn all buildings from the que are created and I get about 37milj gold from not being able to carry over production.

I use the latest released version, not the SVN

 

[Thunderbrd]

Good job! Will be interesting to see what's taking place here.

 

[BigSmac]

I've been doing some reloadings to see if I could reproduce the situation, but it seems to be random. If I do an exact copy of things that turn, it won't happen again...

I added a save game a few turn after the event maybe you can get something from comparing those 2....?

 

[Toffer90]

I've been doing some reloadings to see if I could reproduce the situation, but it seems to be random. If I do an exact copy of things that turn, it won't happen again...

I added a save game a few turn after the event maybe you can get something from comparing those 2....?

That must mean there is a cashing bug that triggered before the first save was made. Very hard to debug.
cashing is temporary files that is used by the game (for performance reasons) and deleted when you exit the game.

Edit: I always spell caching wrong when I'm not paying attention. ^^ Annoying word.

 

[Thunderbrd]

Yeah, hadn't mentioned it yet but this test case did NOT replicate. I'm quite certain the compare isn't going to work either. This... is going to be tough.

(@Toffer - probably true. But sp is 'caching' Just teasin'.)

 

[afairris]

Hi there,
I get a repeating CTD on building a helicopter factory for the first time. I've attached my dump and a save. If you just hit enter after the save it forces you back to desktop. May I ask what I can do to fix it?

 

[Thunderbrd]

You've done your part. Now I need to do the rest. Will look into it later this evening.

 

[Thunderbrd]

Still getting this assert:

Assert Failed

File:  CvGlobals.cpp
Line:  5294
SVN-Rev:  8914
Expression:  stricmp(szType, "NONE")==0 || strcmp(szType, "")==0
Message:  info type 'ATTACHABLE_CHIMNEYSMOKE' not found, Current XML file is: xml\GameInfo\CIV4EspionageMissionInfo.xml

----------------------------------------------------------

 

[afairris]

I'm really sorry, I don't know what that means.
Do I need to get a fresh copy of the XML file?

 

[Thunderbrd]

That wasn't for you really... it was for the team. I'm looking into your save now and I think I've found the problem but it may be I've found a symptom of a larger issue... we'll see.

EDIT: I should warn you now though that you'll be needing to update to the SVN to get the fix once it's resolved. You certainly have an interesting game going there - perhaps a little bit of a runaway but you've gotten a long ways. Was that all from a Prehistoric start?

 

[Dancing Hoskuld]

Still getting this assert:

Assert Failed

File:  CvGlobals.cpp
Line:  5294
SVN-Rev:  8914
Expression:  stricmp(szType, "NONE")==0 || strcmp(szType, "")==0
Message:  info type 'ATTACHABLE_CHIMNEYSMOKE' not found, Current XML file is: xml\GameInfo\CIV4EspionageMissionInfo.xml

----------------------------------------------------------

I am not finding ATTACHABLE_CHIMNEYSMOKE in any of the XML I have searched including base Civ IV, BtS and Warlords. Or any mods I checked.

 

[Thunderbrd]

That's the problem. It's called for through the getInfoForString command. I can't find where it's called from though because it's showing up nowhere in the python or xml files in the mod nor is it coming up as a call from the codeset anywhere. Very strange.

 

[Dancing Hoskuld]

It could be inside a nif somewhere

 

[Thunderbrd]

Must be because a grepwin search of the entire Civ 4 folder structure found no reference to it.

BTW, the issue in this save may be something I will need to refer to a higher power than I. I've narrowed it down to a problem with where the unit thinks itself to be (by plot x and y) and where the unit's selection group seems to think itself to be. The unit itself appears to be a military captive trying to enact a build mission and running up against a problem in that all the checks up to a point are asking if the group's plot is valid but not concerning itself with whether the UNIT's plot is valid. I think the captive was just captured and is going about its merry way without having had a plot set for some reason. The selection group is controlling the activity and thinks its at a valid plot but the unit has a crazy negative iX and iY. At a point it crashes. It's like the unit thinks its there when its not. Perhaps it's trying to update what it should do before the unit is fully re-initialized after capture or not fully destroyed after death. Either way, it's confusing the hell out of me.

 

[Thunderbrd]

Yeah, I've officially hit a brick wall trying to solve this. Having trouble with the map data storage on promos now but I suspect the problem is much more fundamental. I've asked Alberts2 for assistance but that may take a while to get a response on.

My apologies... its not often I can't solve the problem.

However, I CAN say with confidence that the building of the helicopter factory is unrelated to the crash.