• Our friends from AlphaCentauri2.info are in need of technical assistance. If you have experience with the LAMP stack and some hours to spare, please help them out and post here.

Single Player bugs and crashes v42 plus (SVN) - After April 2022

I got:

GFC Error: failed to initialize the primary control theme

with Caveman2Cosmos placed in Documents\... . No such problem if Caveman2Cosmos is junction in BtS\Mods pointing to Caveman2CosmosSVN in Documents\... .

I do not know why, but apparently length of file paths is not issue.
It never worked correctly in first place when in Documents.
You need to run as admin if game is in program files.


Installation guide:
1. Never overwrite old version of mod with new one - always delete old version, as files are often moved, renamed or removed.
2. Mod must be named Caveman2Cosmos and won't work if placed in C:\Users\<USERNAME>\Documents\My Games\Beyond the Sword\MODS.
If its located there - delete it from that place
So valid install paths will look like this:
For GoG this will be something like: C:\Program Files (x86)\Civilization IV Complete\Civ4\Beyond the Sword\Mods\Caveman2Cosmos
For steam it will be something like: C:\Program Files (x86)\Steam\steamapps\common\Sid Meier's Civilization IV Beyond the Sword\Beyond the Sword\Mods\Caveman2Cosmos
Assets folder would exist in Caveman2Cosmos folder - ensure it isn't nested!
3. If game is on system drive "c:", then you need to run as admin.
 
So why does C2C have issues when installed in My Documents ?
Because it just wasn't programmed with this eventuality in mind, and so it errors out when it finds nothing in the main game folder ?
GFC Error: failed to initialize the primary control theme

Perhaps. Something may be hardcoded to the MODS inside main folder.
 
At the moment it appears all devs are a touch distracted or just under intense demands in RL or on other projects. I don't think we aren't seeing these reports, just not having much capacity to do anything with them. Toffer has been able to get in on a few now and then. I'll see if I can find a little time soon. Just very hard at the moment for me. Not quite making enough to cover the budget and have to build out quite a bit of a $ making system to break free of that problem.
 
Playing on 11511 revision, game was started n 11510 revision. I've had lot of random crashes throughout but now the game is stuck - no matter what I do, it will crash during AI turn processing. Link to save:
 
Strange semi-consistent shutdown.
I first thought it's because of me building Vienna Congress, but it's, like I said, semi-inconsistent.
1. No Congress. No break.
2. Congress. Break.
3. Next turn after [1]. No Congress. No break.
4. Next turn after [1]. Congress. No break.
5. No save (too annoyed to bother). Next turn after [4]. After Congress. Break.
 

Attachments

Hmm, this line
algo::for_each(cities(), CvCity::fn::doTurn());

Loops through a NULL city (for player 13 - Greece), that couldn't happen before...
Germany captures and raze Knossos before this happens, so I guess Knossos isn't removed from the cities() array when it is deleted.
@MattCA: any thoughts?

Edit: Knossos is razed due to culture on its city plot falling down to zero, that was impossible before the culture rework. @Blazenclaw @flabbert
....On second thought, possible that it is razed because it culturally flip to German ownership and that it is autorazed due to only having 1 pop (should only happen on conquest though, so maybe the culture flip pass some wrong arguments when it does the ownership change?). I'll investigate further another day.
Figured this one out, when a city's culture expands it change the culture of plots around it including its own plot, if there's a city on any of the plots it will do a check to see if the city revolts and change ownership, the bug here was that it did the city revolt check on its own plot due to changing the culture on it when culture popped, and Knossos here was swamped with german culture so that greek culture was a minority and the random roll made it revolt. This will invalidate the city object as ownership change involve creating a new city object entirely, transferring stats from the old to the new and then deleting the old object. Problem with this was that the city object was in the middle of its doTurn() function when it deleted itself, but the function will continue running unless aborted (and this would not be a good place to abort it as the city deletion happens from one of the looped plot objects before all the plots had been looped). Logical solution here is to not allow the increase of owners culture on the plot of a city to trigger a city revolt, only foreign culture increases should be allowed to trigger this.

I.e. An object deleted itself inadvertently in situ of the stack processing a function member of said object, it then went on to reference NULL variables in self that was assumed to not ever be NULL at that point.

Fix coming to SVN soon.
 
@Toffer90
Probably related stuff.
I've been fooling around with WB city placement, namely putting a huge stack of cities next to each other, and ended up first with every city besides the capital going rebel AI civ, and then next turn or two also losing the capital... and becoming a totally random civ that wasn't the one that ate me, but used the same cities.
Was weird, lol.
Easily repeatable in under 5 minutes on a duel solo map, so no need for a save.
 
Again crash without any error.
yeah been getting alot of these in the Ancient era also, never used to but lately ALOT of them . //
Strange semi-consistent shutdown.
I first thought it's because of me building Vienna Congress, but it's, like I said, semi-inconsistent.
1. No Congress. No break.
2. Congress. Break.
3. Next turn after [1]. No Congress. No break.
4. Next turn after [1]. Congress. No break.
5. No save (too annoyed to bother). Next turn after [4]. After Congress. Break.
Playing on 11511 revision, game was started n 11510 revision. I've had lot of random crashes throughout but now the game is stuck - no matter what I do, it will crash during AI turn processing. Link to save:
New SVN is out, this should fix your crashes
 
Figured this one out, when a city's culture expands it change the culture of plots around it including its own plot, if there's a city on any of the plots it will do a check to see if the city revolts and change ownership, the bug here was that it did the city revolt check on its own plot due to changing the culture on it when culture popped, and Knossos here was swamped with german culture so that greek culture was a minority and the random roll made it revolt. This will invalidate the city object as ownership change involve creating a new city object entirely, transferring stats from the old to the new and then deleting the old object. Problem with this was that the city object was in the middle of its doTurn() function when it deleted itself, but the function will continue running unless aborted (and this would not be a good place to abort it as the city deletion happens from one of the looped plot objects before all the plots had been looped). Logical solution here is to not allow the increase of owners culture on the plot of a city to trigger a city revolt, only foreign culture increases should be allowed to trigger this.

I.e. An object deleted itself inadvertently in situ of the stack processing a function member of said object, it then went on to reference NULL variables in self that was assumed to not ever be NULL at that point.

Fix coming to SVN soon.
I'm also concerned - why would it raze itself during a culture swap? Shouldn't culture swaps not ever lose a population from the peaceful change of ownership? I thought that was one of the whole points of the culture takeover strategy - that it did not do any damage to the buildings or population when it happens. Is that not how we have things here?
 
I'm also concerned - why would it raze itself during a culture swap? Shouldn't culture swaps not ever lose a population from the peaceful change of ownership? I thought that was one of the whole points of the culture takeover strategy - that it did not do any damage to the buildings or population when it happens. Is that not how we have things here?
Probably why it was a BUG (not to be confused with BUG, lol) to begin with, no?
 
I'm also concerned - why would it raze itself during a culture swap? Shouldn't culture swaps not ever lose a population from the peaceful change of ownership? I thought that was one of the whole points of the culture takeover strategy - that it did not do any damage to the buildings or population when it happens. Is that not how we have things here?
It being razed was just a minor if else conditional bug and didn't cause the CTD, the CTD happened if city A called its doTurn() and while applying culture to plots for this turn triggered itself to revolt and change ownership (owner change always kill the original object instance and makes a new one), which was not a good thing to do in that spot of the code deep inside city A's doTurn() function. I changed it so that culture flip can only happen if the culture that increase is from a foreign nation, that way no city should be able to trigger culture flip in cities of their own nation while increasing the native culture of their own civ on these city plots.

Population loss didn't happen on non conquest city ownership change, but the condition for autoraze to trigger was a bit wonky so that autoraze would always occur for cities with only 1 pop even if it was not conquest.
 
the condition for autoraze to trigger was a bit wonky so that autoraze would always occur for cities with only 1 pop even if it was not conquest.
That was the afterthought additional concern that struck me. I realized the rest of the issue too... just wanted to ask about this because that didn't seem right even if it was given the proper timing. Thanks for communicating the situation as plainly and effectively as you do. Really very helpful!
 
Hello!

I've just done a fresh install of Civ4 and C2C on my new computer, Windows 10 Pro 64 bit.

Everything is loading...except when I start a new C2C game, I'm not getting any of the graphical overlays.

No unit actions, no clock/turn counter, no map in the corner, nothing.

Not having this issue on regular BTS. I have the mod installed in the correct place, in the file path with the exe.

Am I doing something wrong, please?

Thanks!

Chris
 
Hello!

I've just done a fresh install of Civ4 and C2C on my new computer, Windows 10 Pro 64 bit.

Everything is loading...except when I start a new C2C game, I'm not getting any of the graphical overlays.

No unit actions, no clock/turn counter, no map in the corner, nothing.

Not having this issue on regular BTS. I have the mod installed in the correct place, in the file path with the exe.

Am I doing something wrong, please?

Thanks!

Chris
Run game as admin
 
Back
Top Bottom