Reinforcement CTD Fix - The Real Deal!

[-Praetor-]

But if you replace the NumEnemiesInBattle with the NumFriendsInBattle conditional, the engine must also count the number of soldiers in the battle, so it will also be confused and cause a CTD...

An the problem would not only be present when the player has reinforcements in the battle, but it would also arise when the enemy brings in reinforcements, because the trait that involves that conditional (NumFriendsInBattle) would also operate if one of the enemy`s armies has a general in it... since the traits operate also for the AI controlled FMs as well as for the player controlled FMs..

Perhaps a more safe solution would be replacing any conditional that involves counting soldiers in the field of battle altogether, with another like total kills, etc...

PS: Just a thought....

[BozosLiveHere]

If NumFriendsInBattle turns out to be as buggy as NumEnemiesInBattle I'll try something else. So far it has worked. I don't really have good substitutes for them (NumFriendsInBattle wasn't a perfect substitute already).

[CountArach]

w00t!!! Go EB!

Okay I just played a reinforcement battle as Koinon Hellenon, just to see if it worked. Here are the results:

Me = 1671 Deployed, 2512 Kills, 1481 Remaining
They had 3 Captain lead armies:
1 = 1221 Deployed, 88 Kills, 192 Remaining
2 = 341 Deployed, 61 Kills, 116 Remaining
3 = 1521 Deployed, 70 Kills, 263 Remaing

Overal Result = Heroic Victory
The second army was the garrison of Korinthos. There was no CTD! It came up a historic battle on the Strat map.

Great job guys! Especially the scripters!

[Zalmoxis]

Doesn't work for me.. but in my game it's saved at battlescreen. Does that make a difference?

[CountArach]

Well I started a new game, but yes, good question.

Save game compatible?

[Kull]

Save game compatible?

Yes. My game was from last March and it worked fine.l

[Kull]

Doesn't work for me.. but in my game it's saved at battlescreen. Does that make a difference?

Well you're at the battlemap here, so god only knows what has been loaded into memory. And given that you still got the CTD, the odds are the old file is already in there. I wouldn't consider this a valid proof for or against the theory.

[Kull]

For those who really want to help, here's a link to a directory containing a bunch of reinforcement CTD save games submitted by other EB players. A few files are "non-CTD" and a couple others have nothing to do with the bug, but the titles will tell which ones should be tested.

Here's the plan:

1) Select and download one of the files.
2) Post the file name you are testing in this thread (so others don't try the same one)
3) After your test, tell us what happened. Be sure to indicate the result of the battle if it wasn't a CTD. Average victories and losses often don't produce CTDs, while Heroic victories are almost a guarantee.

[Oleo]

For those who really want to help, here's a link to a directory containing a bunch of reinforcement CTD save games submitted by other EB players. A few files are "non-CTD" and a couple others have nothing to do with the bug, but the titles will tell which ones should be tested.

Cool, I lost all my saves cause I started playing some other games and I uninstalled. Now I can download and test a couple of my saves again :)

EDIT: Ok, so I tested both Belgae_2stack_ctd.sav and 1sweboz vs 2romans_ctd.sav

In both cases I get a ctd with the old file and no ctd with the new file. Works like a charm!! Nice work. Hopefully it works the same for others.

[Slartibardfast]

Damn! I've been made redundant after five minutes!

Downloading and checking tonight.

[BozosLiveHere]

Uh, Kull, that directory has a lot of savegames from the more recent builds that won't work with .74.

[Kull]

Uh, Kull, that directory has a lot of savegames from the more recent builds that won't work with .74.

True, there are some of each. But it's no harm in trying to load up a v.79 game...if it doesn't work, they can post that and move on to the next one.

[-Praetor-]

Doesn't work for me.. but in my game it's saved at battlescreen. Does that make a difference?

Since when can you save your game from the battle screen?

If the issue was simply counting, then you are right, it wouldn't make a difference. But the fact is that we DO see dramatically different results, so it must be something else. Bovi has theorized that the problem is a bug in the CA NumEnemiesInBattle code. And if so, then we should absolutely see no more CTDs. That's why we want to see a lot more testing. I'm not kidding - you guys can REALLY help us out here by running as many of your old CTD games as possible so we can say with 100% certainty that yes, the bug is dead.

You are absolutly right. Hope that it`s just the NumEnemiesInBattle conditional!

Cheers for the trait team and to Bovi!!!

[CountArach]

Since when can you save your game from the battle screen?

I think he means where you chose whether to fight/autocalc/run. You know the screen BEFORE the battle map. From here you cannot activate the script, so it can be a problem.

[Tellos Athenaios]

Since when can you save your game from the battle screen?

Since RTW came out, you can save your RTW game from the battle screen.
Simply a matter of pressing Esc when the screen has appeared on your monitor... the next time you'll load the game, it will continue at the battlescreen.

Remeber that by then the engine already calculated the number of enemies, friends, and it predicted the odds as well. This data is simply saved together with the rest of your campaign game. That may prove to be the explanation why Zalmoxis continued to encounter the CTD: the 'likely' faulty code already had done it's buggy work.

[Slartibardfast]

At first the patch made all save games crash so I uninstalled EB and reinstalled a clean 0.74 and it seems to be working OK now. I've only tested it with the CTDn1n.zip which ended in a clear victory with out CTDing. Will try some of the others and my most recent CTD's on file latter tonight and tommorow.

Well done Bovi "the Dragonslayer".

[Kull]

This is very encouraging. Thanks to those who are running the tests for us! Please continue to do so, we do appreciate your efforts!

For the record, this fix does not limit or eliminate the award of post-battle traits to winning (and losing) generals. It is a true honest-to-god fix. While giving out credit to Bovi (who discovered the faulty trigger) let's not forget to to give a round of applause to BozosLiveHere, who figured out how to use a different trigger that would produce the desired effects without invoking the CTD.

Tremendous work, gentlemen!

[Teleklos Archelaou]

The CTD had still been present in our 1.5 build and we've made no secret about it. Simply ecstatic to see the posts here about this solving people's old problems now and very happy that now I can stop reading on other mod forums about how EB has all these problems .

[Zalmoxis]

Since RTW came out, you can save your RTW game from the battle screen.
Simply a matter of pressing Esc when the screen has appeared on your monitor... the next time you'll load the game, it will continue at the battlescreen.

Remeber that by then the engine already calculated the number of enemies, friends, and it predicted the odds as well. This data is simply saved together with the rest of your campaign game. That may prove to be the explanation why Zalmoxis continued to encounter the CTD: the 'likely' faulty code already had done it's buggy work.

Turns out I used the wrong word, it was the battle deployment. Would the code still be active?

[Bonny]

not for this particular battle, but for the rest of the game, yes

[Kull]

But if you replace the NumEnemiesInBattle with the NumFriendsInBattle conditional, the engine must also count the number of soldiers in the battle, so it will also be confused and cause a CTD...

An the problem would not only be present when the player has reinforcements in the battle, but it would also arise when the enemy brings in reinforcements, because the trait that involves that conditional (NumFriendsInBattle) would also operate if one of the enemy`s armies has a general in it... since the traits operate also for the AI controlled FMs as well as for the player controlled FMs..

Perhaps a more safe solution would be replacing any conditional that involves counting soldiers in the field of battle altogether, with another like total kills, etc...

PS: Just a thought....

If the issue was simply counting, then you are right, it wouldn't make a difference. But the fact is that we DO see dramatically different results, so it must be something else. Bovi has theorized that the problem is a bug in the CA NumEnemiesInBattle code. And if so, then we should absolutely see no more CTDs. That's why we want to see a lot more testing. I'm not kidding - you guys can REALLY help us out here by running as many of your old CTD games as possible so we can say with 100% certainty that yes, the bug is dead.

[Zalmoxis]

It didn't work for me, so I abandoned the game and I'm going to start from scratch. I'll give you more info on how it's working when I have a chance.