It seems that iceberg group 43_01 through 43_10 is the main culprit, at least with stock (unmodded) SH3, NYGM, and WAC 5.0. And yes, it applies to the unmodded game, too.

I'm beginning to suspect that a lot of the CTD problems people have been encountering when loading a previously-saved, modded game (and not just with my depth charge noise mod) is due to the prevalent practice of testing mods using blank campaign files (due to the necessity of substantially decreasing loading times). Since the icebergs have been present from the beginning, even in the stock (unmodded) version, whatever is causing the CTD problem with them has most likely been carried over throughout the years with other peoples' mods.

I have been combing through the iceberg info in the _RND file, and I simply cannot see anything that should be causing a CTD during reload. I'd much rather fix the problem that exists by modifying the appropriate file info, instead of just brute-force deleting, but until/unless I can figure out how to fix the iceberg info I'll just delete it instead.

Have icebergs to be within rendering distance at the moment of saving for the saved game to become corrupt?

Are you 100% sure that the problem isn't caused by one iceberg unit in particular? :hmm2:

I'm no longer sure of anything now, except that I need a healthy swill of grog.

I'm using stock SH3 for the tests, since it's the fastest load time and (hopefully) the most stable. There are 6 years worth of icebergs: 1939, '40, '41, '42, '43, and '44. Each year has 10 locations. Those 10 locations overlap; for example, Iceberg 39_01 through Iceberg 44_01 start at the same location, and use the exact same waypoints.

Leaving the _LND and _SCR files intact, I modified the _RND file as follows. The results are, well . . . unexpected.

All icebergs removed: no CTD

Iceberg 39 only: no CTD
Iceberg 40 only: no CTD
Iceberg 41 only: no CTD
Iceberg 42 only: no CTD
Iceberg 43 only: CTD
Iceberg 44 only: no CTD

Icebergs 39 and 40 only: no CTD
Icebergs 41 and 42 only: no CTD
Icebergs 43 and 44 only: no CTD
Icebergs 40 and 41 only: no CTD
Icebergs 42 and 43 only: no CTD
Icebergs 39 and 44 only: no CTD

Icebergs 39, 40, and 41: CTD
Icebergs 42, 43, and 44: no CTD
Icebergs 39, 40, 41, and 42: CTD
Icebergs 41, 42, 43, and 44: no CTD

:/\\!!:timeout:

You are right, they don't make any sense, unless we can prove that the problem is not the icebergs themselves, but some limit that gets exceeded with the icebergs.

When you can, do one last test using a _RND with nothing but all the icebergs. I am pretty confident you wont get any crash this time.

Just icebergs = CTD

:hmmm:

schlechter pfennig, what did you use to renumber the random units in the campaign file once you deleted the Icebergs?

I am trying to do one for CCom12.

Honestly? I hadn't renumbered the units, since everything seemed to work without a hitch with my having just deleted the applicable icebergs. I'd released the campaign file fix more as an interim measure while I more deeply investigated what's going on, in the hope of fixing the error instead of just deleting the units as it currently stands.

So it still worked without renumbering them? That's good then. I want to see if it works for CCoM.

Actually I am surprised the file will work because if you open up the random cfg with the mission editor after deleting the Icebergs. Nothing shows up in the editor. It's like its blank or something. Check it out and see for yourself.

I'm not using youre Mod, so i cannot tell anything related about.
But let me write some notes about the SH3-nature in general.
SH3 that you use, may not be the same some other people use. There seems to be different versions out there, even they all called v1.4. Shortly i have learned, that the Steam version is different from what i have to use, (Sf-version).

When looking into all Icebergs in unmodified SH3 RND-campaigne i also cannot find any reason why there may be a problem that results to a crash when reloading a SaveGame. Especially, i used a long time an unmodified game. Saved and Reload in all possible situations, even in between a Convoy-battle without problem. My SaveGame-Crashes begun just about after heavily modifying my game-installation. Or the Number of the Saves reaches a point of no return, at the same time, i dont't know for shure.

So, is there any chance that you can test youre Test-Series again with A.G. Filemanager-ini-Comment-Crash-Mod?
http://www.subsim.com/radioroom/showthread.php?t=222565
This was made for the Steam-version. I dont't know if there is an effect on other versions also.

This may or may not relevant. Icebergs which does not need any Equipment such as Radar have EQP-files containing only a Comment. So maybe this can help. Also in the last post #6 of this Thread A.G.'s words about reproduce i can not write better.

Greetings.

I'm afraid I've given icebergs a bad name. They just happened to be in the wrong place (literally) at the wrong time.

Since I knew something was buggered up between the Depth Charge Noise mod and the Campaign_RND file, I started examining what unit(s) in there were the cause, and I began by "process-of-elimination'. Unfortunately, the icebergs are at the top of the list, and removing those (and subsequently mis-renumbering the remainder) gave the impression that they were the root cause.

They are not.

Several hundred test mission later, and I have no real answers. I've edited the existing Campaign_RND; I've created my own. The results are inconclusive, although give some small clues.

• Anything with more than 40 RndGroup in the _RND file never reload without CTD
• RndGroup size between 20-40 randomly successfully reloads or CTDs, with it usually CTDing, with the CTD frequency increasing as the RndGroup size increases
• RndGroup between 10-20 randomly successfully reloads or CTDs, with it usually successfully loading at the low end, with the CTD frequency increasing as the RndGroup size increases
• RndGroup between 5-10 mostly successfully reloads, with random, occassional CTDs
• RndGroup between 1-5 almost always successfully reloads, with very rare CTDs

Even within subgroups there is no consistency. I've run a mission with a 40 RndGroup, had a successful save-and-load, only to run the same mission immediately a second time and have it CTD upon reload, yet if I reload the first mission save it still successfully reloads.

It would appear that the game is either corrupting, or incompletely saving, the "saved game" information. That's the only conclusion I can reach, as once a mission has been successfully saved, it can be reloaded time and time again.

What I don't understand is the correlation between having Bolds appear outside of being ejected out of a Uboat, the Campaign_RND file, and the corruption of saving a game.

At the moment, understand that using the Depth Charge Noise mod causes CTD upon loading a saved game, if you've been depth charged already.

Maybe its a hardcode thing. Something in the hard code isn't jiving with the fact that bolds appear without actually being launched from a U-boat.

It's something, yes. Odd thing is, I can have dropped bombs convert into torpedoes, and said torpedoes actually hit, explode, and sink, ships, without causing a save-and-reload CTD. Those torpedoes aren't being launched from a U-boat. Granted, they are stuck going due north, because no info is being transferred to them as it would if launched from a U-boat.

It also makes no sense that the error comes from saving a game's campaign_RND info, just from calling up a Bold's .sim activity.

Yes that is odd that its all supposedly unrelated as far as files go. That's to much for my brain to wrap around. That's why I don't really ever mod. Jus simple stuff for me.