Random crashes

[Rhye]

I'm getting frequent crashes to desktop - with no error message (simply the window closes) or any hint in the logs of what could be going wrong. It didn't happen before, it's since a few recent versions.
What's worse, they don't seem to be replicable. Reload the game and everything goes fine, but it might crash again later on.
Does any of you have installed the virtual studio and is capable of debugging such occurrances?

 

[DImensiondog]

No but my game froze and then i had to shut down as it wasn't responding, my game as phoenicia was ruined.

 

[PiR]

No but my game froze and then i had to shut down as it wasn't responding, my game as phoenicia was ruined.

Ruined? I can try to load your last autosave if you share it.

 

[Rhye]

I might have found when these crashes begun. I am testing autoplay of v75 and it seems fine, while v76 crashed once.
Sadly, there are one million changes from v75 and v76 to investigate.

 

[Mezzy]

I get crashes (0.80) whenever I try to start a new game with a civ I've already started a game with before. No error message or anything, it just closes down the whole thing.

 

[Rhye]

I get crashes (0.80) whenever I try to start a new game with a civ I've already started a game with before. No error message or anything, it just closes down the whole thing.

that seems to be of a different kind.
Is your installation clean from other mods or dlc?

 

[Mezzy]

I'm mistaken, I've succesfully started with Phoenica a couple of times now. Crashes on Spain and America so far, although the very first game I started successfully was America, but those after crashes when I click start. Clean from all other mods. Reinstalled to be sure, but still crashes.

 

[Rhye]

OK I need everyone to upload here the savegames which lead to a certain crash.
I will examine them and ask for help to someone who can debug it.
So far the crashes I am getting are only in modern times, and are happening randomly, so if I reload they disappear. That's good that you can continue the game but makes the crash very difficult to catch.

 

[Mezzy]

Sorry, my bad. The crashes I'm getting are caused by my computer locally. I'll be on a more suited computer next week to resume testing.

 

[PiR]

OK I need everyone to upload here the savegames which lead to a certain crash.
I will examine them and ask for help to someone who can debug it.
So far the crashes I am getting are only in modern times, and are happening randomly, so if I reload they disappear. That's good that you can continue the game but makes the crash very difficult to catch.

What if we run autoplay and share the latest autosave? Would it help? (less time-consuming)

 

[Rhye]

if it reproduces the crash, it would be useful.

I'm still struggling as v80 hasn't crashed recently and v75 has crashed once, so I'm totally confused.

 

[hahastone]

As I know,civ5 always crash for many famous mod .I have met cases like that for so many times.I doubt that civ5 has Memory Leak

 

[Rhye]

memory leak? That could be an explanation... when too many cities and data are on the map; which gets somehow solved temporarily reloading the game

 

[Rhye]

If this is true, we need more plagues and natural disaster to sweep up the map!

 

[PiR]

"memory leak civilization v" gives quite many results, it seems a very popular theory but I couldn't find anyone proving it via log files.
Does anyone see the RAM usage increasing or the even swap on C:/ before the crash?
If there is a way to plug a C++ debugger to Civ V probably we could catch the error to prove that.

If this is true, we need more plagues and natural disaster to sweep up the map!

But maybe the "swept things" are not removed from memory, and that's the leak.

 

[hahastone]

"memory leak civilization v" gives quite many results, it seems a very popular theory but I couldn't find anyone proving it via log files.
Does anyone see the RAM usage increasing or the even swap on C:/ before the crash?
If there is a way to plug a C++ debugger to Civ V probably we could catch the error to prove that.



But maybe the "swept things" are not removed from memory, and that's the leak.

Do we have any source code of civv?
I totally agree with u.The difference between before and after we reload the save is whether the swept things were removed from memory.And a simple way for me to suspect is that when civv crashs the fan of computer always begin to be fierce.

 

[PiR]

Idk if the Civ V SDK has something that coulh help

 

[Rhye]

alright then , I will do some tests focused on the complexity of the scenarios.
If i take a savegame that often leads to crashes, delete stuff around the map, save and load, and doesn't crash anymore, that's a strong hint

 

[PiR]

Idk if there are different ways of deleting stuff on the map, and maybe some code do not really delete.