[1.0.9.9] Windows users, help please with Firaxis and crashes

[Qwan]

Windows users, please, I need your help.

I took PIR's suggestion, and contacted Firaxis directly with the crashes I've been having. They've been extremely unhelpful. The first reply was a canned response -- "I understand you have questions regarding Civ VI for Mac" -- that showed they weren't even reading the problem I gave them. Subsequently, they asked for DxDiag and MSInfo32 files, and gave me a link to an article on how to generate them on a PC, despite my having told them I'm on a Mac. Then they said they couldn't help me, because I'm on a Mac, despite my having told them that two of you have tried the saves I've uploaded on Windows, and they've crashed for you as well.

All they need to do is upload my saves, take the next turn, and it will crash for them. Just as it did for two of you on Windows, thank you PIR and bolbas for proving that. What frustrates me so, is apparently they haven't even tried my saves.

I love this game. But I can't play it anymore, as it crashes unavoidably before turn 300 -- four games in a row.

Gedemon, I have to wonder if I'm having the same issues that you have in your thread on "Late game (turn 300+) instability". While I'm only playing on a standard-size map (at standard speed), I'm playing at Deity-level, so maybe I'm running into those instabilities earlier.

So Windows users, please also download one or more of my saves, take the next turn so it crashes for you, then report it to Firaxis. Maybe if enough of you do so, they'll take this issue seriously -- enough so they try loading one of the saves that crashes for us all.

What follows is four separate posts, one for each of the games where it crashed for me the next turn. Yes, I've posted all these before, but I'm consolidating them into one thread, to focus attention. (Plus my synopsis now is more succinct.)

 

[Qwan]

Case #1: Pachacuti 180 1190 ADe.Civ6Save (Feb 14, 2021)
Crashes when you take the next turn.

Also, if one starts with the autosave from turn 176 (or later), and simply do nothing, it will still crash when trying to get turn turn 181. (To do nothing, force the next turn with shift-return.) [This proves it's nothing I did leading up to turn 180 that caused the crash.]

 

[Qwan]

Case #2: Pachacuti 212_Crash_Eleanor2.Civ6Save (Feb 6, 2021)
Crashes when you take the next turn.

Also crashes when you take the next turn from Pachacuti 196 1350 ADe.Civ6Save.

BUT, it will NOT crash from turn 196 IFF you send 3 envoys to Armagh, thus wresting away suzerainty from Eleanor.
==> I'd think this would be rather helpful to Firaxis, as they could investigate why this small change would make a difference.
And yet, this workaround is temporary: after I got to turn 212, nothing I tried worked to avoid a crash.

 

[Qwan]

Case #3: Pachacuti 272 1802 AD-B4rail.Civ6Save (Jan 11, 2021)

Crashes going to turn 274 -- if on this turn 272 you railroad the Holy Site east-adjacent the Military Engineer that has yet to move. While originally I did things on turn 273, such doesn't matter, as you can simply shift-return to skip doing anything on turn 273. (And escape thru any AI dialogue.)

This crash actually generates an autosave for the next turn (274), but I never get control on that turn, and the autosave for that turn (274) will crash when trying to load it.

If you do not railroad on turn 272, but railroad a different tile with a different Military Engineer on turn 273 (e.g. the sheep pasture twice northeast of Kashamarka), it will still crash going to turn 274.

If you do not railroad any tile on turn 272 or 273, you CAN get to turn 274. But it will crash going to turn 275, even if you do nothing. Note that on this turn 274, the World Climate screen says "Next sea level change expected in 1 turns."

I tried doing many other things, but nothing else had any affect on the above.

Again, I think what I investigated to tightly bracket this crashing ought to help considerably in figuring out the problem.

 

[Qwan]

Case #4: Pachacuti 231 crash next.Civ6 Save (Dec 30, 2020)
Crashes when you take the next turn.

Like in case #3, this crash actually generates an autosave for the next turn (232), but I never get control on turn 232 and the autosave for turn 232 will crash when trying to load it.

 

[Qwan]

And this is the note I just sent in reply to their latest message:

Posted on CivFanatics Forums under Bug Reports: <https://forums.civfanatics.com/thre...-help-please-with-firaxis-and-crashes.667643/>

Kamil,

Please pass this case, and particularly the save games I sent, to one of the debuggers for Firaxis, as you clearly are not one.

I spent over 30 years as a software engineer, and was an expert at debugging. I found what helped most was a reproducible error. That's exactly what I've given you. Actually, I've given you several.

All your debuggers have to do is load one of the save games, and take the next turn. It will crash for them. Just as it has for two Windows users on CivFanatics Forums. When it crashes for them, they will be able to generate all the auxiliary files you're asking from me. More importantly, in two cases, I was able to avoid the crash temporarily by doing something--or not doing something--specific. That ought to help considerably figuring out the problem(s).

Your continued claim that you "are not equipped to deal with the Mac version" is without merit. If Windows users on CivFanatics Forums can load my saved games and it crashes for them, so can you.

--steven (Qwan on CivFanatics)

 

[Qwan]

Below is the latest reply from Kamil of 2K Support Customer Service. The fact that Kamil continues to tell me to do more implies none of you Windows users have sent this problem to <support.2k.com>.

Windows users, please please take one of the cases I've posted, follow the steps I did to get a crash, and send a report to Firaxis. Also see the pinned message at the top of this forum by Gedemon for how to generate additional files Firaxis wants with the crash report. Or you can follow the instructions Kamil sent here: <http://support.2k.com/hc/en-us/articles/201335133-How-do-I-attach-a-DXDiag-and-msinfo-file-from-my-PC->

Please do not let this become yet another "Kitty Genovese" debacle, where all of you assume someone else will act. The game Civ VI which we love is broken, and the more of you who provide Firaxis with concrete information about it, the more likely it will be fixed.

The simplest case to test is the first one, as all you have to do is load the saved game, and take the next turn. Although the next two are more interesting situations, because the crash can be avoided temporarily.

(Yes I know that Firaxis should be able to do this themselves, but evidently Kamil has no experience debugging software, and doesn't even play Civ VI.)

Kamil K., Feb 20, 2021, 5:40 PST:

Hey Steven!

I am glad to see that you are so passionate about one of our games!
I want you to know that I would have moved the Earth for you, if need be.
That being said, this is 2K Support Customer Service, and not a development team. The team that is capable of aiding Mac users is located here. I can guarantee you that, were you to contact them, they would have done everything in their power to help you. The issue could have probably been solved by now, if I may say so.

Were you to have any questions regarding out 2K Titles that do not involve Mac, I am your man.

My kindest regards,

Kamil K.
2K Support

 

[HiRezAudio]

Hi Kamil.

Just because 2K keep telling you to do the same stuff does not mean that nobody else has sent in reports. Not by a long shot.
I have had exactly the same experience with PS4 2K support, which is hopeless to the point of being patronizing at best and downright rude at worst (FWIW, my PC Windows version has so far been just fine and we will return to your bug later)
The problem I have repeatedly tried to report to them is the constant late game crashing on turn end on Consoles. This may be related to your bug.
All I ever get from 2K are potted responses to clear the cache (PS4 does not even have one) & reinstall the game, and when you follow the link on how to clear the cache you get sent to a page that....tells you the PS4 does not have one.
They also repeatedly require video footage of the bug - this is impossible to do on a console unless you get someone to stand beside you on every turn end, filming, just in case the bug triggers. Yet when I try to point this out the requests for videos, screenshots etc still come. I then threatened to make their emails public, and finally got a reply asking me to simply 'answer the following questions and that is all - we will pass this to the developers' and guess what happened?
Yup - the very next email (which I was told would not occur) asked for.....screenshots and videos.
I even tried to point them at these forums, linking to threads where other console users are getting the same problems but to no avail.
They just do not bother to read the reports or read the answers, otherwise they would not keep going round and round in circles - I am just waiting for them to blame this on Covid - it will happen one day I am certain.

Anyway.
With your bug - I will try the saved games & see what happens but from reading this it will be a crash.
I have a couple for you though - do these crashes only happen with Pachacuti, and are you running any Mods?

 

[PiR]

Hi Kamil.

Kamil is the support guy from 2K I think.

btw, sorry for you @Qwan that must be very frustrating.
However looking at my own experience with 2K support I would from now on dare open only one ticket at a time https://forums.civfanatics.com/threads/what-happens-when-we-report-a-bug-here.666334/#post-16004211
Which do you prefer me to open?


See also this feedback
https://forums.civfanatics.com/thre...return-was-some-existential-questions.667316/


Maybe try complaining on Twitter... that's what works best with companies these days...

 

[Qwan]

With your bug - I will try the saved games & see what happens but from reading this it will be a crash.
I have a couple for you though - do these crashes only happen with Pachacuti, and are you running any Mods?

Thank you for looking into this!

I have only been playing Pachacuti since December. (I had wanted to win with Pachacuti before going on to another leader.) However, I found an even older save which crashed on the next turn, with Frederick. I thought I had posted it to the Forums, but can't find it. Details in my next post.

The last time I finished game was a Diplomatic Victory as Alexander (Deity-level, standard map-size) on turn 272 November 7th, 2020, i.e. after the Frederick crash.

While I play with UI-only mods normally, I tested all of these loading the saves without any "community created content" mods.

The UI-only mods I normally use are: Better Civilization Icons, Better Civilopedia, Better Deal Window, Better Report Screen, Better Tech Tree, Better Trade Screen, Extended Diplomacy Ribbon, Happiness and Growth Indicators, Leugi's Unique District Icons, Radial Measuring Tool, Real Housing from Improvements, Sukritact's Simple UI Adjustments, and Z Unit Marks. Oh, and normally I have the Scout Cat turned off; I prefer the look of the Dog.

 

[Qwan]

Case #5: Frederick 134 CrashNextUnless.Civ6 Save (Oct 13, 2020)
Crashes when you take the next turn.

Will NOT crash if you sell India one (1) iron.

Unfortunately, I didn't finish this game, so I don't know if avoiding the crash was temporary. While I have correspondence that says it crashes from autosaves 131 onwards (using shift-return to skip any actions), I don't have any other saves from that game anymore.

 

[Qwan]

More info on Case #1

As several of these cases crash on an AI's turn, I thought to try one in multiplayer, and look at that AI's turn.

Loaded Pachacuti 180 1190 ADe.Civ6Save in Multiplayer: Hot Seat (after moving the file appropriately).
Changed Nubia to Human (player #2), then continued, taking my (player #1) Next Turn, to get to player #2.

If I immediately re-save (as Amanitore 180 1190 AD s1)
then re-load it in HotSeat, changing Nubia back to AI
(thereby letting the AI do everything for Nubia)
it will still crash. Unsurprisingly.

BUT if I first set production of Nuri to anything, then re-save
then re-load it in HotSeat, changing Nubia back to AI
(thereby letting the AI do everything else for Nubia)
it DOES NOT crash.
I can get to turn 181 of player #1.

Oddly, when I first get control of Nubia player #2 as above,
there's a prompt to choose production for Qasr Ibrim,
but after doing so, there's no prompt to choose production for Nuri
despite "Nothing currently being produced" from Nuri.
And despite that, there's 26 production towards a Cuirassier.

Again, if I do not set production for Nuri, game will crash; if I set it to anything, it will not crash.

(I do not have to set production for Qasr Ibrim to avoid a crash.)

I went back to prior autosaves, and used Hot Seat to see what Nuri was building:
on turn 177: RenWalls 64/300
on turn 178: RenWalls 164/300
on turn 179: RenWalls 266/300
on turn 180: nothing, but had 26 towards Cuirasser
Nubia is employing Limes (+100% production towards walls).

On turn 177, production was 25.
On turn 180, production was 26.
In-between, it fluctuated a little, because the pop went up, causing -1 amenities temporarily.

From my limited testing, when I play Deity, the AI gets +100% production.
Limes is +100% production. 25 x 4 = 100, which accounts for the increment in the AI building walls.

Note that Nubia does NOT get the Deity bonus on the turn I switch it to Human.
So you have to add that in to see what would have happened if Nubia had stayed AI that turn.
Adding 50-52 on turn 179 to 266/300 means the RenWalls would have finished that turn.
The overflow of 16-18 is less than the Limes bonus, so my understanding is that it's discarded.
Thus if I switch Nubia to Human for turn 180, I should see 26 production (no Diety bonus, no carry-over) in whatever the AI decided to start on turn 179. That's there's 26 production towards a Cuirassier implies that's what the AI decided to start.

It's like we have a Schrödinger's box: the AI has decided to produce a Cuirassier and NOT produce a Cuirassier.

========
Meanwhile, I'm reluctant to continue forwards with the turn 181 I can obtain this way, because Nubia is deprived for a turn of Diety-bonus production, which doesn't seem fair.

Although even if somehow Firaxis got this information despite 2K "Support", I doubt they'd have it fixed by the imminent February update. Hmm.

 

[PiR]

Case #3: Pachacuti 272 1802 AD-B4rail.Civ6Save (Jan 11, 2021)

Crashes going to turn 274 -- if on this turn 272 you railroad the Holy Site east-adjacent the Military Engineer that has yet to move. While originally I did things on turn 273, such doesn't matter, as you can simply shift-return to skip doing anything on turn 273. (And escape thru any AI dialogue.)

This crash actually generates an autosave for the next turn (274), but I never get control on that turn, and the autosave for that turn (274) will crash when trying to load it.

If you do not railroad on turn 272, but railroad a different tile with a different Military Engineer on turn 273 (e.g. the sheep pasture twice northeast of Kashamarka), it will still crash going to turn 274.

If you do not railroad any tile on turn 272 or 273, you CAN get to turn 274. But it will crash going to turn 275, even if you do nothing. Note that on this turn 274, the World Climate screen says "Next sea level change expected in 1 turns."

I tried doing many other things, but nothing else had any affect on the above.

Again, I think what I investigated to tightly bracket this crashing ought to help considerably in figuring out the problem.

Reproduced & submit to 2K: http://support.2k.com/hc/requests/8441723


FYI, I have already seen (and reproduced) other bug reports in this forum about a crash at a change of climate level, but never created the conditions for this myself.

 

[jwolf713]

Short question because I think I got lost. Did you find a solution for any of the game files listed here?

 

[jwolf713]

If you don't work help tips like reinstalling a game or checking file consistency, up windows update, or whatever else, there's one more thing you should try. It's a rule of law: say a war, make peace, or break an alliance.
This rule: 'rule of law: say a war, make peace, or break an alliance.' may seem ridiculous at first glance, but it works if the cause of the crash is not a damaged mod.
I've used it successfully in over 10 games when other advice like updating everything on pc, reinstalling the game, manually increasing the pagefile size or choosing DX11 instead of DX12 etc... didn't help.
A simple denunciation of war or peace worked for me in 4 cases out of 7.

Sometimes there are more complicated situations. Namely, the declaration of war can result in the fact that within a few to a dozen turns from the declaration of war 'game will crash'. But there's a way too. Unfortunately, you have to go back to the earlier point of saving the game and instead of declaring war to wait in a peaceful relationship until you jump to the point where the 'game crash' occurred

I had this in one of the games. I declared wars of the USA, i will take several of their cities and then always in the same turn the game turned off. I loaded an earlier save, maintained a peaceful relationship with the US at all costs, and declared war on other players, and the US later dealt with

Another difficult case is when 'game crushing' is caused by an 'active alliance with one of the players'. I know it sounds a little strange, but in one of the game I confirmed it experimentally. For nearly 9 hours I played different scenarios of war spoken to other players or being a victim of war, in others I sought to maintain peace. All for nothing. Having already been a little nervous about playing groundhog day, I decided to conduct an experiment and combine with an active alliance with the Zulus. Due to the fact that there is no option to break the alliance, even at a large cost, I was left to load the earlier 'save -36turn'. This time I said no and it's strange when I came to turn 560/561 where previously there was a 'game crash', I went to turn 561, then 562 and so on.


Of course, errors / lack of updating mods relative to the game can also cause various problems, as well as a conflict between mods.

 

[PiR]

In the case I reported, this is clearly due to sea rising / climate change, so I don't think these would work, but it's interesting to see so many factors can induce a crash.

 

[jwolf713]

My experience shows that changing production in a city, e.g. instead of tanks, ships or buildings does not really change anything. Even transferring governors, buying units with gold or faith points, making trade deals. It's all for nothing, sometimes you can move the game crush point a turn or a few turns, but it will happen anyway.

 

[Qwan]

Short question because I think I got lost. Did you find a solution for any of the game files listed here?

For the game that crashed due to railroading a tile, or not doing so but waiting a few turns when the pollution of others was sufficient for the next turn to have a sea level change, no.

For the most recent crash I reported, which happened on Nubia's turn, I did find a work-around (see my post above). I copied that save into HotSeat, and checked to see what Nubia was doing. I discovered that their city Nuri had just finished producing something, and that the game wouldn't prompt me-as-Nubia to produce something there. If I set Nuri's production, then returned control to the AI, I could continue.

However, it crashed in the same way four more times, after that turn 180 and up to turn 222. Setting Nuri's production always avoided the crash. Sometimes I needed to take a 2nd turn as Nubia, since it was losing the Diety bonus production for that turn, and so sometimes Nuri didn't finish production when I took control. (I'd switch Nubia back to AI for a turn, then back to Human, so that Nubia only lost Diety bonus production, but still got Diety combat bonus, and still made all other decisions.) [I don't feel too guilty about this, as Nubia isn't one of the top other Civs, nor are they one of my neighbors.]

In the original turn 180 instance, there was overflow production placed into something. In the second instance, there was not. What Nubia was finishing building varied: for turn 180 it was Renaissance Walls, another instance was a Military Academy, another instance was the Huey Teocalli wonder. Another instance was a Cuirassier. The last instance might have been a Musketman corps.

It hasn't happened since; I'm now at turn 250, but I may have encountered the pollution crash. That does not happen during the AI's turn, that happens at the beginning of your turn, before you get control, and an autosave is generated for that turn (which will crash if you load it). I haven't loaded the February update, because I want to finish this game first, so I don't know if they fixed the pollution-crash bug. If I can't work around it, I may try the update.

Again, none of these crashes are due to a community mod, as they will happen if you don't use any. That is, while I only use UI mods playing these games, I've loaded each without any mods, and it will still crash the next turn.

Yes, I've checked declaring war and declaring an alliance. I checked everything I could think of, for the cases #1 - #3, until I could isolate a single action (or inaction) that would result in the game crashing the next turn. As I posted.

 

[jwolf713]

You have an alliance or friendly relationship with some players, which makes it impossible to check the option of declaring war. You would need to share more files from earlier moments in the game. Or bad relations making it impossible to check the alliance short option.

Case 1,2,3 are three different games, but only files from the turn after which the game crashes. Basing only on the file from the turn in which the game crashes quite significantly limits the possibility of circumventing the problem of self-closing of the game. Unfortunately, the previous states of the game / saved game points are useful. Sometimes you have to go back even in some cases a dozen or so turns or more. It depends on the relationship with which civilization is the problem, no matter good or bad.

There are a lot of locks in the files you share, in the form of alliances or friendly relations, so that you can check various variations of war / peace, etc. In My advice, look for files from previous turns, where you do not have, for example, an alliance or friendly relations that block the possibility of declaring war.


'game crushing' maybe caused by an 'active alliance with one of the players'. I know it sounds a little strange, but I confirmed it experimentally./2 games in total

To start a new game safely, you can always copy files from the current one as a so-called backup. I don't know if and how often you save the game manually, but the files from the 'auto' folder can also be copied to a folder, and then copied back if necessary. And in the meantime, play a different / new game.

 

[Qwan]

I did declare an alliance and a surprise war on the same turn. With different Civs, of course. The crash still happened.

I don't know what you mean by "self-closing".

I have set saving all autosaves. So I can go back to any that you want. For cases #2-#4, I renamed the Single file so it wouldn't overwrite these saves when I started a new game. I can share any autosave you want -- just name which.

In the more recent cases, I did go back several autosaves, one at a time, then continued forwards doing nothing, i.e. force each next turn with shift-return. It would still crash. I don't go back further than five turns before giving up on that.

I'd be happy to have someone else working on these crashes, even if you're not Firaxis. Have you tried loading any of the cases in my initial post and seeing if it crashes for you?

You have an alliance or friendly relationship with some players, which makes it impossible to check the option of declaring war. You would need to share more files from earlier moments in the game. Or bad relations making it impossible to check the alliance short option.

Basing only on the file from the turn in which the game crashes quite significantly limits the possibility of circumventing the problem of self-closing of the game.
Unfortunately, the previous states of the game / saved game points are useful. Sometimes you have to go back even in some cases a dozen or so turns or more. It depends on the relationship with which civilization is the problem, no matter good or bad.

'game crushing' maybe caused by an 'active alliance with one of the players'. I know it sounds a little strange, but I confirmed it experimentally./2 games in total

To start a new game safely, you can always copy files from the current one as a so-called backup. I don't know if and how often you save the game manually, but the files from the 'auto' folder can also be copied to a folder, and then copied back if necessary. And in the meantime, play a different / new game.