Battle CTD Reports ONLY

[Kull]

To help the EB Team track down the elusive Battle CTD bug, we're going to need your help. If you experience a battle CTD, please provide us with the following information:

1) Factions involved.
2) All unit types (from each army).
3) The units actually engaging in combat when the CTD occurred (if you can remember)
4) Did you or the enemy have reinforcements available for the battle?
5) Do you have ANY other graphics mod installed on top of your baseline EB v.73 game? (things like Skymod, Darth formation mod, a personal unit mod, etc.) If so, please tell us which one.
6) Is the CTD repeatable? (i.e. you tried the battle again, and it did/did not CTD)

Note: If you experience a repeatable CTD, PLEASE post a link to that file (upload it to any of the many free file hosting services)

[Cheexsta]

This is the same one I mentioned in another thread.

1. Makedon (me) and Epeiros (AI).
2. My army had 2 generals, 10 units of Taxeis Hoplitai, 3 units of Akontistai and 4 units of (under-strength) Hippeis Thessalikoi. The Epeirotes had two armies: first one (the one being besieged) had a general, about 2-3 units of phalanxes (Taxeis Hoplitai/Phalangitai and Misthrophoroi Phalangitai). The second "army" consisted only of Pyrrhos himself.
3. The CTD was after viewing all the stats at the end of the battle, but before the loading bar was shown. So there were no engagements that caused it; it was straight after clicking on the tick after viewing the post-battle stats.
4. Yes; Pyrrhos was attacking my army, the besieged force was reinforcing.
5. No; clean 1.2 install with 0.72b and 0.73 installed directly over it.
6. No. I tried reloading the save, but Pyrrhos decided not to try attacking this time. Instead, Alexandros attacked and I fought the battle as normal and won. No CTD.

The only thing I did differently in the Pyrrhos battle was click "End Battle" rather than "Continue" as I had in 2 or so other battles prior to that. In the Alexandros battle, I clicked "Continue" and let it go for a few seconds before clicking "End Battle." Unfortunately, the autosave file from before both battles has been overwritten.

[Zero1]

1. This involves me playing as the Sweboz and fighting against the Eleuthori

2. I had three general units, three Fraemamen, Two Ridarhorjaz*sp?*, and one unit of archers. The enemy had two units of Fraemamen, one of those Swordsbondsmen guys, one unit of those Sweboz skirmisher units "youthlings" or something like that, one unit of archers, and one unit of Riderhorjaz.

3. My three general units were engaged with two units of Fraemamen, and my three Fraemamen were engaged with the youthlings, and the Riderharjoz

4. No, neither me nor the AI had re-enforcements

5. No, clean RTW install, 1.2, EB .72 with .73a installed over

6. Yes, I tried once with the script on and it CTDed, I tried without the script and it CTDed at roughly the same time, I tried a third time with the script off and with the blank traits file and still got a CTD and again at roughly the same time in the battle I.E. the main melee at the town center.

Edit- Now that I think of it, the CTD always happened when my men engaged the youthlings skirmisher type soldier.

[khelvan]

Zero, as you single out these "youthling" units, it would be most helpful to have their exact name, so I can test some things. Is it possible to get that? It would help a lot!

Do you mean the Swainoz?

Thanks!

[Zero1]

Yes, the Swainoz...don't know where I got "youthlings" from

[Kull]

Cheeksta & Zero1 - Thanks guys, that's EXACTLY the kind of detail we'll need in order to track these down. Much obliged, guys!

[Alleycat]

5 sweboz ctds on town battles w/and without walls against rebels

ctd happens when I direct Riderhorjaz calvery to attack. I can't reproduce specifically since battle go a bit diff. each time but....

1 they've been directed to attack different units during the crash(ei...they aren't attacking the same unit type each time as in the example above)

2 it always happens around/close to the town square

3 played sweboz till 190 on the 7.2 build and never encountered this problem

4 riderhorjaz won't attack when grouped with other riderhorjaz if target is near town center. (yes skirmish is off and ammo gone)

clean install and all that jazz. btw the 3000 boost on mines has made the sweboz a lot more enjoyable IMO.

oh...no reinforcements envolved, also it doesn't happen every time i attack w/ riderhorjaz in a town square. I'd say maybe a 1 in 5 chance so far.

[Geoffrey S]

When giving 0,73a a quick whirl I played two turns as Sweboz. Attacked the town in Denmark with base units nearby. No reinforcements, no other mods. The game crashed in-battle as soon as my Riderhorjaz charged the enemy. When I've more time I'll fill the form out properly.

Also, all Eleuthoroi flags were white.

[struckat]

I too have started with the Sweboz. I dumped all my horses first turn to save $$.

I have the same battle crashes as the others posting here, script on, script off, no matter what I do. It always happens when the emeny general is killed.

It seems to me that somthing was changed with the rebels in this area because all of them have white banners now. Prior to the patch I only encountered this in the Numidian desert.

What did you change with the skandahoovian rebels(all 3 cities to the north of the Sweboz)?

The crash happens exactly when the enemy general dies.

Struckat

P.S
I wont be on line long, my screeching woman is near, you should all be afraid, very afraid.

[Zero1]

Seem like there are alot of problems with the Sweboz, I wonder why?

[struckat]

It may be due to an underlying anti Germanic sentiment.......

[Birka Viking]

I play as the Aideui and everytime me or the ai get reinforcment from another army the game ctd after the battle when the loading screen from battle to campaigne comes up...Sorry for my bad english..

[struckat]

So I finally managed to take Carrodunum without a battle crash. I just cant figure out what the difference was. So I fix my attention on Eburonum.

First I attack the city and fall victim to a second stack of rebels. The battle was going well though still in doubt, then it crashed. Most of the rebel troops were Aljaz Gae and Lugi that looked like fleas until you zoom to ground level.

So then I reloaded and moved to the woods south east of town. The full stack Eleutheroi followed and fell right into my ambush. What a slaughter.

There was still five or six units that had not yet routed, all fleas as described above, and then CTD.

I am too clueless about modding to help figure out what is going on. I would blame the bad sprites but the earlier crashes were against armies with normal sprites.

In the last version the only battle crashes I encounterd were fighting against rebels with white banners in Numidia.

I have also restarted the Sweboz campaign four times with similar results, battle crashes with almost all of the Eleutheroi armies.

It seems that the issues with the last version were fixed and a few new ones have popped up. I may be toostupid to help, but I must say that you guys have created the best and most interesting mod ever.

Struckat

[Cheexsta]

I ran into another battle CTD. Playing as Makedon against the Seleukids, I ordered a unit of Hypaspistai to use their spears (alt+attack). To my pleasant surprise, they actually ended up using their spears (never seems to work with anything but cavalry and missile troops) with the overarm animation. But after a second or two, the game crashed with no error.

Something wrong with the skeleton/animation? I'll try to reproduce the CTD and figure out if changing the skeleton fixes it...

Edit: just ran a few tests. With all of them, the Hypaspists broke out their spears just before engaging the enemy, before switching back to swords after hitting the enemy unit. Now, for each of them:
Test 1: Standard EB settings. Crash after selecting alt+attack.
Test 2: Replaced fs_javelinman (Hyp secondary attack skeleton) with fs_o_f_speaman (the Thureophoroi primary attack skeleton). Crash.
Test 3: Replaced fs_javelinman with fs_spearman (generic spearman skeleton). Crash.

So I'm pretty sure it's not the skeleton that's the problem. Must be something wrong with the actual secondary attack, though it looks just fine in EDU. Might test one or two more skeletons and, if that doesn't work, just disable the whole secondary attack...

Edit 2: just ran one more test, this time using fs_swordsman_barb. It worked! The Hyppies were busy whacking the enemy Clyddabre with their spears and it wasn't causing any CTD's. I'm 100% sure the other skeletons work on different units, so it must be a model problem with the Hypaspistai? Are they missing a bone or something?

[Spendios]

1) Arverni(Me) against Aedui (AI) Siege of Bibracte
2) Arverni : 1 Family Member, 3 Leuce Epos, 2 Sotaroas, 2 Mala Gaeroas, 2 Botroas
Aedui : 1 Leuce Epos (General) 2 Sotaroas, 2 Mala Gaeroas
Aedui Reinforcments : 2 Family Members, 2 Leuce Epos, 1 Ridoharjoz, 1 Botroas, 1 Sotaroas, 1 Mala Gaeroas

3) CTD occur at the end of battle when clicking to come back to campaign map

4)Yes reinfocments of the sieged city by another Aedui army

5) No

6) Only way to avoid this CTD was to lift the siege, attack the reinforcment army and the re-siege Bibracte

[Dooz]

A pre-battle CTD from another thread, but I believe it's relevant. If not, shoo me away .

https://forums.totalwar.org/vb/showp...7&postcount=51

[Teleklos Archelaou]

I ran into another battle CTD. Playing as Makedon against the Seleukids, I ordered a unit of Hypaspistai to use their spears (alt+attack). To my pleasant surprise, they actually ended up using their spears (never seems to work with anything but cavalry and missile troops) with the overarm animation. But after a second or two, the game crashed with no error.

Something wrong with the skeleton/animation? I'll try to reproduce the CTD and figure out if changing the skeleton fixes it...

Edit: just ran a few tests. With all of them, the Hypaspists broke out their spears just before engaging the enemy, before switching back to swords after hitting the enemy unit. Now, for each of them:
Test 1: Standard EB settings. Crash after selecting alt+attack.
Test 2: Replaced fs_javelinman (Hyp secondary attack skeleton) with fs_o_f_speaman (the Thureophoroi primary attack skeleton). Crash.
Test 3: Replaced fs_javelinman with fs_spearman (generic spearman skeleton). Crash.

So I'm pretty sure it's not the skeleton that's the problem. Must be something wrong with the actual secondary attack, though it looks just fine in EDU. Might test one or two more skeletons and, if that doesn't work, just disable the whole secondary attack...

Edit 2: just ran one more test, this time using fs_swordsman_barb. It worked! The Hyppies were busy whacking the enemy Clyddabre with their spears and it wasn't causing any CTD's. I'm 100% sure the other skeletons work on different units, so it must be a model problem with the Hypaspistai? Are they missing a bone or something?

This is definitely a problem Cheex. I have tested it separately and gotten the same result. It's definitely the Hypaspists. We are trying to figure out whether we are going to release another hotfix with earlier animations or a corrected one (and the port problem fixed), or decide what else would be the best thing to do. (So technically I'm reporting in with a CTD in this "ONLY" thread too )

[Kull]

Edit 2: just ran one more test......

You have a nice analytical approach, and we appreciate you isolating this problem down to the Hypaspitai. Could you do me a HUGE favor and start up a Sweboz game, and see if you can likewise isolate the offending unit/skeleton in those CTDs? It would really help us if we had another "bad unit" to look at. Many Thanks!

Edit: O'ETAIPOS told me that the Sweboz general uses the same model as the Hypaspistai.....so we might be honing in on a bug that's causing both CTD's!

[paullus]

Sad, I was about to report that the hypaspists seemed to have a problem, but that's already posted. Good luck fixin it!

[Cheexsta]

TA and Kull: definately sounds like a model error then. I only use Gmax (can't afford 3DSMax...) so I might be able to stare at it blankly and look like I'm doing something intelligent, but my guess is that the animation/skeleton is referring to a bone that isn't present in the model. As soon as the game tries to move this non-existant bone, it freaks out and CTD's. I don't know for sure, since I only have a rudimentary knowledge in the area, but that's all I can think of...

Edit: just checked the model in GMax (after a slight frustration with the pak extractor). There don't seem to be any bones missing. In fact, I would swear it's exactly the same model as a Pezhetairoi...

Edit 2: ok, a couple more tests done. I replaced the Hypaspistai model in DMB with that of the Pezhetairoi and fired up the game. As I expected, the skins were all over the place, but the game ran fine. Unit charged in, I used alt+attack and BAM - CTD.

So I did the same thing, this time replacing the model with that of the Taxeis Iudaioi (since they were right underneath Hyps and they used the fs_javelinman skeleton). I KNOW that this model works fine with the skeleton, otherwise I'm sure you'd get more CTD reports with them. Exactly the same as the Pezhetairoi skin; a CTD.

I think that rules out a model problem. My next test will involve switching the primary/secondary attacks around; perhaps the game doesn't like having a spear weapon as a secondary weapon?

[Kull]

I'm getting in way past my knowledge level....but what controls the actual movement of the sword? Is there some kind of txt file which specifies max-min movement limits, and are we somehow over-riding them?

Or, since the CTDs always seem to occur when the unit switches from spaer to sword, what governs that switch....and in what way is it different from those of other units who make the switch, yet don't experience the CTD?

[Cheexsta]

Again, I only have a basic knowledge of models and animations, but this is what I've come to understand:

A model consists of a mesh. This mesh is divided into "bones," such as bone_pelvis, bone_RThigh (right thigh) and so on. Animations are just a set of "commands" (for a lack of better term) that dictate what bone does what. What skeletons do is they link a model to a certain set of animations. For example, fs_javelinman refers to a set of animations that makes the model raise its right arm and move in a way that makes it look like it's throwing the javelin. This allows the game designers (and modders) to be able to change the animations for models quite easily, rather than ingraining an animation within the model file (which makes it difficult to change).

As for primary/secondary weapon switches, the game by default uses the primary weapon (which is one of the weapons labelled as such in the model; for example, with Hypaspistai this is the sword) but as soon as the player executes the secondary attack (alt+attack) the unit changes to their secondary weapon and uses that attack stat (as defined in export_descr_unit) and skeleton (units with secondary attacks that are different have two skeletons assigned to them) instead.

Hope that makes sense...

[Kull]

Thanks - that does help. One other test you might run. Try testing the animations/models included with the v.72 download and see if they cause a CTD too (drop them into your v.73a game). Might help to compare the two EDU files, also.

[Cheexsta]

Ok, I'll give that a go. Just quickly, I tried another test to see if my "spear" hunch was right. I reset everything to "standard" EB 0.73 settings (might also be noteworthy that I'm not using the hotfix...) and changed the skeletons so that the fs_javelinmen was the primary skeleton and swapped the primary and secondary attack stats around. The Hypaspists changed to their secondary weapon just before the charge (like in all previous tests) and then swapped back once they were engaged. As soon as they swapped back to their primary weapon, the game CTDed. I'm once again convinced that it's a model/anim problem ~:D

So yes, I'll give the 0.72 anims a go. It might take a while since I'll have to reinstall EB over another copy of Rome (or even to a temporary folder, that might work).

I'm so confused as to what the problem is. It just doesn't make sense: it could be the animations, but then why would said animations work perfectly on all other units? And then why did only the fs_swordsman_barb skeleton work with the spears? It's definately not a EDU problem, otherwise the latter would not have worked, either.

You might have to get one of your modelers to look into this. I think it's over my head...

Edit: ok, there's a few too many new skeletons in the patch to try to use the 0.72 anims, it would require editing quite a bit of DMB. I might give it a go later on when I have the patience to edit all of that

[QwertyMIDX]

Many other units have the same spear/sword combination, a combination which has always made RTW a little unhappy. Try testing the Arverni Arjos, the Soldorus, the Gastiz, etc.

[Kull]

OK guys, this bug is solved....no further testing is necessary (at least on this one).

It's a problem with the new sword/spear animation. Evidence:

Cheexsta tested the Hypaspistai, a unit using that animation. Result = CTD.
QM tested Arjos, a unit using that animation. Result = CTD.
Numerous individuals haave reported problems with the "Sweboz General". It turns out that his bodyguard uses that animation. Result = CTD.

I'll keep you posted.

[Dooz]

*sniff sniff*

I smell hotfix. Mmmmmm.

[Kull]

*sniff sniff*

I smell hotfix. Mmmmmm.

Shhhh. Don't tell anyone, but I think I just found the pre-battle CTD too.....

[Teleklos Archelaou]

-At least some pre-battle sprite problems fixed that caused CTDs
-One animation or skeleton that was causing a CTD in some battles
-Placement of a port building by scripts that wasn't showing up on maps and later caused CTDs
-A trait problem that may have been causing CTDs

All of the first three could be fixed with a hotfix and still be save-game compatible. I don't know if the last one would be. But a hotfix with these issues would remedy a lot of the problems and probably be save game compatible too.

[khelvan]

The only thing that won't be saved-game compatible is the ports issue; it turns out the script placement doesn't actually solve the problem, which is why we had it in 0.7.2 as well (hence all the frustration with later-game CTDs with that version). This is unfortunate, because every single player WILL run into this problem at some point, unless they immediately take Kypros.

The 0.7.3x versions have been very buggy and we realize this. We introduced some new bugs that weren't caught in initial testing, but with your help we have found all of them, we think, and while I've posted some temporary solutions, a new patch version should be out in a few days.

0.7.4 will be the stable version we all thought it would be. Assuming the 0.7.2 nasty CTD bug was the port problem, and I firmly believe it was, all of the 0.7.2 known CTDs will have been fixed in 0.7.4.

I can't promise there will be none at all, but it will be what 0.7.3 was supposed to be, and more.

Also, see my latest edit here for how to fix that port problem without losing your current campaigns. The power is yours, if you are a good player.