exterminate/enslave/occupy CTD - any known solution?

[florin80]

The exterminate/enslave/occupy CTD happens when the player wins the battle and is presented with the 3 choices before taking the settlement. But on clicking one(any) the game CTDs. As I`ve said the exact option doesn`t seem to matter much because I had a saved game in TE when this happened to me time ago and I kept trying all of them and the result was the same. But after a number of tries I was able to get passed it(without changing anything). So it might not be a campaign killer, but it sure is annoying if you play a big long battle and are happy with the result you`ve gotten only for the game to crash.

It happens in quite a few mods(RTRPE, TE, RG, IBFD, SPQR and EB are just some I`ve read that had this problem at one time or another) so I thought to ask if anyone knows of a solution. Or at least a theory that can be tested further. Since it happened to so many mods I`m curious if anyone seriously tried to fix it yet.

[Squid]

The reason behind the CTD is well know, the problem is since the battle isn't in the settlement (it's on the tile your army is on outside the settlement) the garrison army when defeated attempts to retreat, but the game tries to do a standard retreat and not a retreat to the settlement. This normally isn't a problem, but where you start to end up with difficulties is when the garrison can't retreat as if it was a non-garrison army you get the CTD. I'm guessing the CTD happens when the button is pressed because that's when the retreat of the armies is calculated.

Possible solutions are to make sure that all settlements have multiple methods of access, so if one way towards the settlement is blocked an army can "retreat" another if need be, so islands with only one way on will almost always cause problems. This is difficult to do, and is not always successful. The real solution would be for CA to release a patch that has a retreating army check if it is garrison and then not run the standard retreat algorithm.

[florin80]

I don`t think the retreat issue is the one/only one to explain it. In the example I gave above it happened like this: I was besieging a settlement(with dacia iirc) and the enemy(macedon iirc) attacked me with a second army from the outside during its turn. I chose to play the battle on the battle map and I completely destroyed both armies. And after that I was presented with the options to take the settlement because it was empty. There were no troops whatsoever left to retreat in any direction.

[Squid]

I don`t think the retreat issue is the one/only one to explain it. In the example I gave above it happened like this: I was besieging a settlement(with dacia iirc) and the enemy(macedon iirc) attacked me with a second army from the outside during its turn. I chose to play the battle on the battle map and I completely destroyed both armies. And after that I was presented with the options to take the settlement because it was empty. There were no troops whatsoever left to retreat in any direction.

Yes this is exactly what I was talking about, I know it doesn't seem logical, but it seems a garrison army will always attempt to retreat, even if there are no soldiers left in the army, as I said altering the map (and making sure you siege from a side that doesn't block retreat) helps to reduce the occurrences of the CTD.

I'm pretty sure it isn't trait related, because RS has/had problems with this CTD, and we aren't using NumMenInBattle, also any PostBattle conditions I believe are evaluated before you return to the campaign map (which is where you get the exterminate/slave/occupy option) because if you get a trait with a gain message it's waiting for you when the campaign map comes back up.

[florin80]

But could you please give an example of a blocked retreat with an image? I`m not sure I understand how to see which tile is blocked and which isn`t.

[MasterOfNone]

Unless this rare error can be duplicated you are going to have a hard time tracking them down. The fact that it cannot - afaik - be duplicated suggests a random element.

[CaesarVincens]

That also has problems, though I don't know why. But a solution that should work is to not destroy the garrison army and then take the city on your turn.

[MasterOfNone]

I'm not sure if killing all units means there are no survivors - can some be healed?

Anyway, I must say I have rarely come across this problem - and I do not remember seeing it in FATW(though I may be wrong). Do some mods suffer from it more than othes and, if so, why? Does it occur a lot in vanilla?

[florin80]

I'm not sure if killing all units means there are no survivors - can some be healed?

I think I did enough of a good job so that nobody got healed. Seriously, noone could have survived, because otherwise there wouldn`t have been an empty city and I saw no army remaining on the strat map.

Anyway, I must say I have rarely come across this problem - and I do not remember seeing it in FATW(though I may be wrong).

Judging from the reports it`s not happening very often or at least it`s not reported often. But it does happen and it sucks if it does. Kills alot of the mood so a sure solution would be great. So far I`ve only read playing "tricks" on how to avoid it so if it could be figured out like the revolt ctd it would be much better.

Do some mods suffer from it more than othes and, if so, why? Does it occur a lot in vanilla?

At some point I seemed to know it happened in vanilla too, but so much time has passed that I can`t say I remember it actually happening there. But when I made the post I didn`t want to name any mod wrongly so I did a search in each of their subforums and all named above had at least one report of the problem.

[Aradan]

Does this CTD occur even with non-generals (with regular captains I mean)?

[florin80]

You`re thinking at a trait issue?
In my example there was a family member leading the army. I don`t know for others. It`s not something that gets asked I think. But I`ll try to ask someone who reported this bug for RG relatively recent.

[Aradan]

Yeah, I'm thinking traits. They are known to cause CTDs post-battle, when the engine gets confused (example: the NumEnemiesInBattle), so it might be that.

But I`ll try to ask someone who reported this bug for RG relatively recent.

[florin80]

I got the reply from Octavius Vatco. Turns out his army was lead by a family member(the faction leader I think). Which pretty much leaves the question open instead of eliminating the trait suspicion. But in case it is a trait, any idea which could do this? According to MC the traits in vanilla RTRPE were fine and this bug was reported a number of times for it. Traits are not my speciality...

[Aradan]

Dunno, needs to be researched I guess. Anything with "PostBattle" is a suspect, especially traits including number of enemies etc. Also traits that fire upon settlement occupied/sacked/exterminated I guess.

[HouseOfHam]

Possible causes:
- A bug in the game where it's trying to give a trait to the losing army's general who survived the battle but died as a result of the settlement being captured (some kind of internal null pointer exception).
- A variation of the revolt CTD
- No retreat route CTD
- Trait level going below 0. Can't be the reason in Vanilla, since there aren't any triggers with negative Affects values.

Questions:
- Is the enemy army being lead by a family member?
- Is the player army being lead by a family member?
- Does it happen when AI sallies out?
- Does it happen when AI sends a reinforcement army to relieve a besieged settlement?
- Does it ever happen if both armies are led by captains?
- Does it happen after an auto-resolved battle?
- Are there ANY survivors of the sallying army?

If the cause of the CTD is a problem with traits, it would seem logical that it wouldn't happen every time, since most triggers use Chance < 100.

Possible ways to test:
- Change all trigger Chance values to 100 and see if you can reproduce the CTD more consistently
- Remove all triggers and see if you can ever get that CTD

[florin80]

Questions:
- Is the enemy army being lead by a family member?

I don`t remember in my case. I`ll try to ask Octavius again about his experience.

- Is the player army being lead by a family member?

Yes.

- Does it happen when AI sallies out?

With no AI reinforcement army attacking you from the outside?

- Does it happen when AI sends a reinforcement army to relieve a besieged settlement?

Yes. I think this is the most often case.

- Does it ever happen if both armies are led by captains?

Don`t know.

- Does it happen after an auto-resolved battle?

In my case it did. But only after I had it previously in a played battle. Meaning after the crash I loaded the saved game from before the battle and kep auto resolving it and the game kept crashing until one time it didn`t. I didn`t made any changes though.

- Are there ANY survivors of the sallying army?

Not in my case. I`ll ask Octavious for more info. He had the crash with RG while capturing Damascus and choosing enslave.

[Squid]

I, have no personally seen this CTD, but if you search for reinforcement CTD over at the org you'll find a thread started by mak on the subject and it mentions Masillia in vanilla having the issue, so you may be able to create it there.

[MasterOfNone]

Here is that thread: http://forums.totalwar.org/vb/showthread.php?t=75911

[Lusted]

One thing I have found to greatly reduce this bug(or remove it entirely as I haven't seen it reported for TE Gold) is to use the corrected_regions.tga to replace the existing descr_regions.tga and rebuilding map.rmw.

[florin80]

One thing I have found to greatly reduce this bug(or remove it entirely as I haven't seen it reported for TE Gold) is to use the corrected_regions.tga to replace the existing descr_regions.tga and rebuilding map.rmw.

Yes, I remember you proposing this solution, but has it been extensively tested? I wasn`t sure if it got included in Gold and weather or not the bug was reported again for it.

Here is that thread: http://forums.totalwar.org/vb/showthread.php?t=75911

Thank you for the link! There`s no one clear answer resulting from there, but it has useful examples of the crash as well as some experiments.