Bannerlord might be highly moddable, but its not exactly mod-friendly. Way too many CTDs

Users who are viewing this thread

The modded game seems to be overly susceptible to CTDing at the drop of a hat. Every minor error or conflict seems to result in an instant CTD and building a stable modlist is a real chore. Could the game be more robust?

I need to use a mod called OnGovernorChangedWorkaround just in order to play the game. What this mod does is it intercepts various intermittent OnTick event errors that normally cause CTDs and forces the game to ignore them. It doesn't fix any errors, mind you, it just makes the game ignore them, and the crazy thing is, it f**king works. I would get CTDs every 10 minutes without the mod but I can play for hours with it.

So that makes me wonder, if those errors can be safely ignored without seeming to cause any long-term issues, how many other game-crashing errors could also be safely ignored? Does the game really need to hard crash every time it comes across a bad texture or a faulty crafting piece or object reference not set to an instance of an object?

Is it possible that the devs coded the game engine to be overly sensitive to minor errors in order to catch more bugs during EA? And if so, is it possible to turn down the sensitivity and make the game more resilient?

Now that the game is officially released and the Workshop is up and running, there's going to be an influx of inexperienced mod-users complaining about their modded games crashing non-stop, and its no longer an acceptable excuse to say "its an unfinished game, mod at your own risk". Ultimately, this game lives or dies on the basis of its mods, so TW needs to take mod stability seriously.
 
Another way of looking at it is they should have made a feature rich game that didn't need so many mods. As of 1.8 my record was 76 active mods working to make the game palatable, right now I am down to a measly 27 but at least that is all the "must haves" in that stack.
 
Another way of looking at it is they should have made a feature rich game that didn't need so many mods. As of 1.8 my record was 76 active mods working to make the game palatable, right now I am down to a measly 27 but at least that is all the "must haves" in that stack.
That's true, but even if it was a perfect game, mods were always going to be a huge part of the experience. The only difference is that then mods would be nice extras, like spice to flavor the meal, whereas now they're essential to enjoying the game
 
Blaming the game on mod CTD's smh.
Keep shaking your head. Its very helpful.

Mods are a key element to the long term success of this game, but at the moment, modding can be very frustrating experience. Its in Taleworlds' best interests to make the modding experience as smooth as possible. The game was supposedly built from the ground up with modding in mind and yet its prone to crashing at the slightest mod-induced error. If there's some way to tune the game so that its more robust, then it would benefit everyone.
 


So here's a good example of a bullsh*t error that probably doesn't need to actually crash the game but it does. It looks like the game is checking to see if some unit somewhere should be getting xp on the tick but something's not properly defined and the game engine doesn't know what to do so it just CTDs. Is it caused by a mod? Who knows? The error report doesn't say.

All I know is that there's no way to get past this crash and so its Game Over and yet another 50-hour save goes straight to the recycle bin. Does it really need to crash on this error? Can't the game just ignore it and whatever unit should have gotten xp on that tick just doesn't get the xp?

This is a serious problem. I'm a fairly experienced mod user. I've built stable 300+ Skyrim modlists , so I know a little bit about load orders and managing mod conflicts, but this bullsh*t happens to me every time in Bannerlord. I'm ready to give up on the game. I can only imagine the difficulties a newbie mod user is having with Workshop mods.

Modding is absolutely essential to this game. If Taleworlds wants Bannerlord to have the long life that Warband enjoyed, they really need to look into how to make this engine more robust and error-resistant because if the mods don't work, this game is f**king dead.
 
All I know is that there's no way to get past this crash and so its Game Over and yet another 50-hour save goes straight to the recycle bin. Does it really need to crash on this error? Can't the game just ignore it and whatever unit should have gotten xp on that tick just doesn't get the xp?

It can, but that's generally a bad idea while developing a game. If an object reference is missing, it usually means serious and difficult to trace bugs down the line. They have 4 options:
1) Write a long error message detailing every variable at that moment that could possibly cause a crash (time consuming to code)
2) Do a memory dump (kills performance)
3) Ignore the bug entirely (this sometimes works if the object doesnt actually need to exist)
4) Just crash the engine with an exception and use the debugger.

Using the latter option is the most common during game devlopment because you generally want to remove low level bugs like this as soon as possible. Having dangling pointers (or whatever caused this issue) in a fully released game is the equivalent of building a tank, only to discover 2 years into the war that the welders used plastic instead of steel.
 
It can, but that's generally a bad idea while developing a game. If an object reference is missing, it usually means serious and difficult to trace bugs down the line. They have 4 options:
1) Write a long error message detailing every variable at that moment that could possibly cause a crash (time consuming to code)
2) Do a memory dump (kills performance)
3) Ignore the bug entirely (this sometimes works if the object doesnt actually need to exist)
4) Just crash the engine with an exception and use the debugger.


Using the latter option is the most common during game devlopment because you generally want to remove low level bugs like this as soon as possible. Having dangling pointers (or whatever caused this issue) in a fully released game is the equivalent of building a tank, only to discover 2 years into the war that the welders used plastic instead of steel.


I get why they'd want to tune the game to be extra crashy in EA to generate more crash reports and catch more bugs, but now that the game's an officially launched product with an official workshop, I would hope that they start to prioritize stability and resilience. Maybe they need to do another pass through the code and harden up those areas that are prone to crashing. And they shouldnt just ignore crash reports from modded games. Of course they can't support mod problems, but if they can see common errors cropping up with lots of mods then maybe there's something they can fix on their end.

I get that many errors are in fact fatal and need to crash before they screw up the whole game, but I'm sure a lot of them only lead to minor problems if any. A common source of crashes is a bad asset or crafting component. I use this one mod that adds parties to the map, but certain troops use a particular bugged horse so the game crashes if you try to fight them. Couldn't the game just spawn those troops without their problematic horses? I'd rather have that than a hard crash. I'd rather be able to play the game with some weird glitches than not be able to play it at all.
 
Lately it's the Modders who are causing trouble to each other and for the users, what happened with MCM 5 was a misery, mainly on Steam, it was catastrophic for the confidence we could have in the
workshop.
 
Lately it's the Modders who are causing trouble to each other and for the users, what happened with MCM 5 was a misery, mainly on Steam, it was catastrophic for the confidence we could have in the
workshop.
For sure. I got burned by that update and learned my lesson about using the workshop for anything but the most trivial of mods.

Surely there's got to be a better way to manage non-backward-compatible updates of the major prerequisite mods than just dropping them with no warning. Was it critical that the mod release that day? Could the developer have held it back a week and issued a heads up so that dependent modders could have a chance to update? But that's something modders will need to work out for themselves.
 
Surely there's got to be a better way to manage non-backward-compatible updates of the major prerequisite mods than just dropping them with no warning. Was it critical that the mod release that day? Could the developer have held it back a week and issued a heads up so that dependent modders could have a chance to update? But that's something modders will need to work out for themselves.
AFAIK any mods uploaded to steam workshops auto-update subscribers during their next session. The modders have no control over this.
 
AFAIK any mods uploaded to steam workshops auto-update subscribers during their next session. The modders have no control over this.
True, but that's why he could've posted an alert before actually publishing the update and maybe with a link to a preview version for modders to update their mods ahead of time. There really needs to be some kind of agreed upon best practices for posting game-breaking updates or it'll really hurt the overall mod scene. Casual users are only going to put up with their savegames getting bricked so many times before they give up for good.
 
True, but that's why he could've posted an alert before actually publishing the update and maybe with a link to a preview version for modders to update their mods ahead of time. There really needs to be some kind of agreed upon best practices for posting game-breaking updates or it'll really hurt the overall mod scene. Casual users are only going to put up with their savegames getting bricked so many times before they give up for good.
MCM is a utility used by other mods. Getting all mods that use it to upgrade simultaneously seems unlikely. Better for modders to publish mods without dependancies (other than Native etc) or to avoid steam workshop.
 
Is MCM now only on the steam workshop? Why not just have different/older versions on nexus? IDK what other benefits steam has aside from more straightforward installation.
 
Of course they wouldn't, but at least he would have done his due diligence and any issues would be with the modders who didn't update. Like it or not, prerequisite mods seem to be here to stay so these issues will keep happening unless there are some common processes that people follow. Modders are a smaller group than mod users so it's easier to influence their behavior. And users seem to just want to use the workshop
 
Mods with dependencies are a problem with or without the steam workshop, its generally just bad practice. Steam's auto-update system is a pain in the ass, and i don't know if there's a setting to disable it at least for mods, but any time mods rely on other mods you're going to have problems like this. Back in the SimCity 4 days there were massive dependecy chains of hundreds of different prop libraries you had to install. Even with physical games and manual mod installation, this quickly became unmanageable to update anything.

A lot of this is the fault of the developers, you often get mods that make the game more moddable or fix API issues. But these mods are usually tiny, so having them as a separate dependency for every mod rather than just bundling it is an unnecessary risk.

Then again most of the internet is dependency chains, and bugs in some ancient github can cause half the internet to shut down and cost billions in repairs. True story
 
Back
Top Bottom