modding HUGE Problem with Modding RDR2 That No One is Talking About

[Sanah 17]

1 hour ago, HarrietOrtega said:

I might just delete it altogether because it's just been a trial of attrition to play more than 10 minutes without a crash.

Honestly just play unmodded, like I do. And I will keep doing that until a total solution is found, which is something very unlikely to ever happen.

Edited May 17 by Sanah

[HarrietOrtega 9]

25 minutes ago, Sanah said:

I have the impression the game's code is being held together by ducktape. Not surprised though seeing how much dynamic stuff goes on; from sounds, animals minding their own business, stagecoaches passing by, trains, peds, a random encounter every minute... if you think about it it's pretty reasonable that everything goes literally nuts when you touch even the tiniest thing lmao

Oh yeah, don't get me wrong, what they've accomplished is insane. The amount of game assets alone is terrifying and a lot of work obviously went into the game. BUT, it's still a big pile of poop.

 

I have tried playing unmodded several times but the blurry textures on my triple ultra wide...well, it just looks so rubbish that it takes me out of the game completely...then I do a cry.

[kaozzz777 18]

Hate to be that guy, but does R* officially support modding? Are there official mod tools and/or documentions? They offer a product and it works correctly as provided.
It's in the responsibility of mod authors to understand the engine they're modifying and make sure it's working. Not just the mod itself, but the game as a whole. This issue shows that there isn't a deep enough understanding how RDR2 works. And that users of mods don't know about possible downsides. But thanks to this thread at least people can be aware of what to look for.

I'd still wish there would be a proper and clinical debugging way of testing with actual benchmarks and numbers, but gotta make due with what we have.

 

Edited May 17 by kaozzz777

[HarrietOrtega 9]

This is true but developers know that their game will be messed with, I'd actually go as far as saying that all developers would support and release tools for the games they create if it weren't for the publishers and higher ups. I firmly believe that and it's always a good thing, Valve and Epic wouldn't be anywhere near where they are now if they had locked down their game engines and kept their mod tools to themselves 25+ years ago

 

They don't have to support it at all and I'm not miffed about it, infact I really like the challenge of trying to decipher the data so I can add ONE additional blade of grass to the entity batch without creating a localised black hole.

 

Sorry I'm ranting, I'll get back on topic, promise 🙂

[Sanah 17]

50 minutes ago, kaozzz777 said:

Hate to be that guy, but does R* officially support modding? Are there official mod tools and/or documentions? They offer a product and it works correctly as provided.
It's in the responsibility of mod authors to understand the engine they're modifying and make sure it's working. Not just the mod itself, but the game as a whole. This issue shows that there isn't a deep enough understanding how RDR2 works. And that users of mods don't know about possible downsides. But thanks to this thread at least people can be aware of what to look for.

I'd still wish there would be a proper and clinical debugging way of testing with actual benchmarks and numbers, but gotta make due with what we have.

 

I hate to be that guy too, but you are absolutely right. The Story Mode's quality, although slightly diminished through online content updates, is unparalleled. That's point blank period. Yes, some minor aspects such as TAA implementation could have been improved, but this is honestly a pretty solid PC port for probably the biggest and most ambitious game in existence.

R* didn't give us a modding toolset to begin with, so it's totally on mod authors to deliver in such regard. If they cannot do that, or they don't know how, then I would suggest them to just refrain from making mods altogether. This might be a hot take for many people but 99% of the mods for this game are a complete fucking mess. They hardly do what they promise to do, and when they do, it's at the cost of the main game's features. And the funniest part of all this is that nobody gives a damn apart from us who are trying to keep this thread alive.

 

Sorry if someone gets offended by this but it's just what it is. The most problematic mods -which by the way are usually the most endorsed- have restricted comment sections on Nexus. What a coincidence. And hey, comments might be restricted but donations? Nah, always open.

Edited May 17 by Sanah

[HarrietOrtega 9]

1 hour ago, Sanah said:

I hate to be that guy too, but you are absolutely right. The Story Mode's quality, although slightly diminished through online content updates, is unparalleled. That's point blank period. Yes, some minor aspects such as TAA implementation could have been improved, but this is honestly a pretty solid PC port for probably the biggest and most ambitious game in existence.

R* didn't give us a modding toolset to begin with, so it's totally on mod authors to deliver in such regard. If they cannot do that, or they don't know how, then I would suggest them to just refrain from making mods altogether. This might be a hot take for many people but 99% of the mods for this game are a complete fucking mess. They hardly do what they promise to do, and when they do, it's at the cost of the main game's features. And the funniest part of all this is that nobody gives a damn apart from us who are trying to keep this thread alive.

 

Sorry if someone gets offended by this but it's just what it is. The most problematic mods -which by the way are usually the most endorsed- have restricted comment sections on Nexus. What a coincidence. And hey, comments might be restricted but donations? Nah, always open.

Yeah, I'm not very vocal on forums but reading the comments section (if it's not locked like you said) it's so sad. I just don't get why you would make a mod and not fix it if it's broken etc...there's no pride in the work it seems or to try and understand what it is you are changing and the roll on effects it can have down the line. Makes me so mad.

 

oh and Discord...everything is friggin Discord.

[Dagger323 32]

5 hours ago, HarrietOrtega said:

Yeah, I'm not very vocal on forums but reading the comments section (if it's not locked like you said) it's so sad. I just don't get why you would make a mod and not fix it if it's broken etc...there's no pride in the work it seems or to try and understand what it is you are changing and the roll on effects it can have down the line. Makes me so mad.

 

oh and Discord...everything is friggin Discord.

It's gotten to the point where I won't even bother downloading a mod if the mod page does not allow bug reporting and have an open comment section. I refuse to support lazy mod authors who just want to put stuff out there but are unwilling to be responsible when it comes to fixing their own creations.
 

An example of this I can immediately recollect is the mod author of "Roleplayer's Dream". The mod originally had an open comment section (and bug reporting if I remember right), in which I posted a request to fix an issue present at the time where the mod would stop functioning almost entirely if the player was to respawn or reload from another save point in-game and not remain in the same continuous game session. Instead of responding the mod author immediately closed the comments down and removed bug reporting as soon as I posted the comment. Needless to say I will not be downloading any future releases from said author.

Edited May 18 by Dagger323

[Gunter Severloh 46]

I dont speak for other mod authors, but ya i agree if a author doesn't want to fix their mod and

locks or closes the comment section of the mod page then ya best move on.

   However i can say that for those honest authors that released a mod, and maybe have posted an update

there are other factors involved, their time, their interest and their know how.

    So if they dont know how to fix the issue whatever it is in their mod, then imo they should ask someone,

research or something.

 

Im fortunate enough to work with 2 other creators as we are a small team for our mod Bandit Hideouts,

crossed99 and i been working on our mod since early 2023 together, and fixing, updating, issues

among many other things.

   Personally i enjoy the interaction with the community on my mod page, it lets them ask questions,

and get to know my routine, and see that im trustworthy and am dedicated to the mod.

I love what we created was my idea to begin with, i enjoy playing with it, and enjoy helping others

solve issues, or explain to others with questions about how certain things function

or about settings in the mod.

    Ya sometimes its a headache to get a bug or an issue that you cant seem to solve, but i will be honest

with the players of the mod and tell them i dont know whats causing it, or we tried this, or are looking

into the issue, ect,. we have solved many issues since the mod has been out.

   Overall thats just me, being dedicated to your work and having reasons to work on the mod, for me

its the fans and the love of the mod and most of all the game, again i dont speak for others but

to each author its what they want and decide to put into what they create. Cheers!

[HarrietOrtega 9]

22 minutes ago, Gunter Severloh said:

I dont speak for other mod authors, but ya i agree if a author doesn't want to fix their mod and

locks or closes the comment section of the mod page then ya best move on.

   However i can say that for those honest authors that released a mod, and maybe have posted an update

there are other factors involved, their time, their interest and their know how.

    So if they dont know how to fix the issue whatever it is in their mod, then imo they should ask someone,

research or something.

 

Im fortunate enough to work with 2 other creators as we are a small team for our mod Bandit Hideouts,

crossed99 and i been working on our mod since early 2023 together, and fixing, updating, issues

among many other things.

   Personally i enjoy the interaction with the community on my mod page, it lets them ask questions,

and get to know my routine, and see that im trustworthy and am dedicated to the mod.

I love what we created was my idea to begin with, i enjoy playing with it, and enjoy helping others

solve issues, or explain to others with questions about how certain things function

or about settings in the mod.

    Ya sometimes its a headache to get a bug or an issue that you cant seem to solve, but i will be honest

with the players of the mod and tell them i dont know whats causing it, or we tried this, or are looking

into the issue, ect,. we have solved many issues since the mod has been out.

   Overall thats just me, being dedicated to your work and having reasons to work on the mod, for me

its the fans and the love of the mod and most of all the game, again i dont speak for others but

to each author its what they want and decide to put into what they create. Cheers!

I really appeciate this response, I use your mod and it's a shining example of hard work and dedication. Thanks to you both for making and maintaining one of my fav mods!

[kaozzz777 18]

My benchmark method with monitring through Rampage. There's always some variance, but you can establish a range which can be hit consistently. After a few trial runs you see the same stuff. You can verify first by checking vanilla vs SH+Rampage. See if same stuff and events spawn, then take numbers and compare what you get with mods running.
V2 scripthook is advised as ABs been struggling with pool management for a while.

clean document folder and a recent clean chapter 2 savegame from nexus. no old saves, no cloud saves. Delete cache files.
 

I check script count and ped pools, example:
1) Leaving camp on horse around 9am, sunny weather. heading to the area in screenshot right after the tracks. getting always 6 herb spawn scripts and around 26-32 scripts overall; 120-130 ped count average. Check animal species. look up a wiki that lists them for an area.
2) ride to valentine. buy a train ticket to St.Denis. If it loads without crashing and has passengers, head to the Bayou, compare scripts/ped amounts and animal species.
ESP in rampage ped manager greatly helps tracking stuff, especially small animals.

edit: 3) trains see discussions before and later

To get vanilla ped counts and events, native scripts have to run as intended as they control world events, chance encounters, wildlife, pick-ups etc.
 

 

Edited May 22 by kaozzz777
cleanup

[Sanah 17]

Interesting. Will test trains later with V2 once again, to see if I see any difference. Doubt it, because I've done multiple tests with V2 in the past.

Edited May 18 by Sanah

[Dagger323 32]

28 minutes ago, Sanah said:

Interesting. Will test trains later with V2 as you said.

I've been using V2 the entire time for the very reasons kaozzz777 stated. I'm still having problems with trains, but perhaps the issue is even worse with AB's Scripthook; having not used it for months, I couldn't say either way.

Edited May 18 by Dagger323

[Dagger323 32]

Once again, though, I must reiterate to others what I've said before about testing for TRAINS, not just peds and wildlife. It seems many newcomers to this thread are focusing too heavily on everything else and not troubleshooting the train spawning nearly enough, particularly in the manner that I clearly laid out a couple pages back in this thread. Train spawning breaks quicker than other spawning does and must be tested the most thoroughly out of everything. I encourage people to go back and read what I've said about train spawns previously before assuming their game is fixed or in an acceptable state just because they're seeing mostly normal amounts of peds and animals.

[Sanah 17]

36 minutes ago, Dagger323 said:

I've been using V2 the entire time for the very reasons kaozzz777 stated. I'm still having problems with trains, but perhaps the issue is even worse with AB's Scripthook; having not used it for months, I couldn't say either way.

Was about to say something like this. I have done many tests with V2 during these past months and didn't see differences worth noting. I did go back to AB's though just because of its wider compatibility with old mods, but like you said, issues persist no matter which one is used. I really didn't see much of a difference in terms of spawns regardless of what V2's author states on his Nexus page.

 

I appreciate the insights, and I find the RampageTrainer stuff about scripts and ped pools interesting, but this is honestly nothing particularly new. We are at the same spot.

Edited May 18 by Sanah

[Sanah 17]

7 minutes ago, Dagger323 said:

Once again, though, I must reiterate to others what I've said before about testing for TRAINS, not just peds and wildlife. It seems many newcomers to this thread are focusing too heavily on everything else and not troubleshooting the train spawning nearly enough, particularly in the manner that I clearly laid out a couple pages back in this thread. Train spawning breaks quicker than other spawning does and must be tested the most thoroughly out of everything. I encourage people to go back and read what I've said about train spawns previously before assuming their game is fixed or in an acceptable state just because they're seeing mostly normal amounts of peds and animals.

Yeah, it's very tempting to check if a couple herbs, deer or birds spawn near you, but the main issue are trains. So please, stay on topic and check trains before posting any results.

[kaozzz777 18]

I retried my benchmark again and could replicate results. With trains.
Did my script monitor check in heartlands, then did Daggers train route.
Trains at ~10AM and ~11AM around Flatneck Station ; ~1-2PM Wallace Station
Btw there will basically always be birds on bridges (just another visual indicator)
after that I run my test for flatneck described here (direct post link)
again trains passing at around 1am,12am,0am with 150ms and force update checked. 22-27script count (with horse).

 

if the engine runs all its scripts even after i teleported and change game minute timing, I don't know what else is supposed to benchmark and stress test it. And trains you hear in the game world aren't "fake" ambient sounds. Plus don't forget that trains are by default a high variance entity in the game (inconsistent schedule, mostly reactive to player actions). Relying on a train script mod to check if vanilla train scripts run correctly isn't optimal if we can't guarantee it's 100% reliable. Not saying the mod causes issues, I had good tests with it.

Again, every train runs on the script "trainrobbery_ambient". Including the ones you buy a ticket for.
It's not about herbs, it's about vanilla scripts loading! they are the one clear benchmark. The entity counts are a result of vanilla scripts executing, not the other way around. Same goes for pick-ups like herbs. Every area has a "population" script with accompanying ambient scripts as a baseline.

And again, I thought i get the right results with same modlist when using AB's SH and all being ok, but noticed inconsistencies after looking at unmodded vanilla. Script count now is higher on average. Also everything being exactly the same, buying a train ticket resulted in FFFFF with AB and no issue with V2.
Rampage showing 0 ped count and reduced script amount on latest version with AB isn't just a stat bug, it's an indicator of something going wrong. Ramapge dev/debug tools and ease of use are excellent for benchmarking and isn't listing scriptmods you using.

Edited May 18 by kaozzz777

[kaozzz777 18]

On 5/18/2024 at 9:35 AM, Sanah said:

Yeah, it's very tempting to check if a couple herbs, deer or birds spawn near you, but the main issue are trains. So please, stay on topic and check trains before posting any results.

This is a crude oversimplification of what I was about. Just like the thread started, it's also about all variations of entities spawning: NPCs, Vehicles, Objects/Pick-ups, wildlife species including subspecies.
Just in this small area here you can spot at absolut minimum 14 animal types, same for the Bayou spot
 

Edited May 22 by kaozzz777
cleanup

[Sanah 17]

3 hours ago, kaozzz777 said:

This is a crude oversimplification of what I was about. Just like the thread started, it's also about all variations of entities spawning: NPCs, Vehicles, Objects/Pick-ups, wildlife species including subspecies.
Just in this small area here you can spot at absolut minimum 14 animal types, which I do with my modlist, same for the Bayou spot
 

Err, yeah... sure, you have totally fixed it. Congrats.

Edited May 18 by Sanah

[kaozzz777 18]

Some info that was kindly shared with me on RampageDiscord about scripts and pools:

[kaozzz777 18]

The goal here imo is to make aware of the problem and how to test reliably, If you want to know which mods run with native systems working as intended.
For that you need a test suite with reproducible results and compare numbers. Monitoring scripts, pool sizes and know which entities to get spawning. At set weathers, times and locations. That can be done if you combine Daggers tests with my approach.
 

 

There won't be a magical solution. Play vanilla, assured everything works like it should. Mod only after you done everyting and go wild. Or go the dark souls route.

 

Have fun

[Dagger323 32]

16 hours ago, kaozzz777 said:

I retried my benchmark again and could replicate results. With trains.
Did my script monitor check in heartlands, then did Daggers train route.
Trains at ~10AM and ~11AM around Flatneck Station ; ~1-2PM Wallace Station
Btw there will basically always be birds on bridges (just another visual indicator)
after that I run my test for flatneck described here (direct post link)
again trains passing at around 1am,12am,0am with 150ms and force update checked. 22-27script count (with horse).

 

if the engine runs all its scripts even after i teleported and change game minute timing, I don't know what else is supposed to benchmark and stress test it. And trains you hear in the game world aren't "fake" ambient sounds. Plus don't forget that trains are by default a high variance entity in the game (inconsistent schedule, mostly reactive to player actions). Relying on a train script mod to check if vanilla train scripts run correctly isn't optimal if we can't guarantee it's 100% reliable. Not saying the mod causes issues, I had good tests with it.

Again, every train runs on the script "trainrobbery_ambient". Including the ones you buy a ticket for.
It's not about herbs, it's about vanilla scripts loading! they are the one clear benchmark. The entity counts are a result of vanilla scripts executing, not the other way around. Same goes for pick-ups like herbs. Every area has a "population" script with accompanying ambient scripts as a baseline.

And again, I thought i get the right results with same modlist when using AB's SH and all being ok, but noticed inconsistencies after looking at unmodded vanilla. Script count now is higher on average. Also everything being exactly the same, buying a train ticket resulted in FFFFF with AB and no issue with V2.
Rampage showing 0 ped count and reduced script amount on latest version with AB isn't just a stat bug, it's an indicator of something going wrong. Ramapge dev/debug tools and ease of use are excellent for benchmarking and isn't listing scriptmods you using.

I appreciate the information you've provided. But have you tested for train spawns in towns? Rhodes, Valentine, Saint Denis, etc? Because they are the biggest determiner of train spawns working correctly. The route past Flatneck Station that I mentioned before is good for testing wilderness train spawns, but that's only half the equation. Using the Railroad Engineer mod, and waiting in a populated town, should nearly always result in a train spawning and coming through the town once an icon shows on the map that a train is approaching and you wait for it to come past you. In vanilla, that's what happens; however, if mods are breaking the spawning, then the icon will remain grayed out and no train will appear no matter how long you wait. The tests must be conducted at every town and settlement. Rhodes, Saint Denis and Annesburg (not counting the two trains that are always sitting idle in the Annesburg depo) almost never show trains when mods are breaking spawns. Valentine will, but usually only around 9–10 P.M. at night. To conduct these tests, you must be IN TOWN and able to witness the train pull into or past the station, not be out in the wilderness and merely follow the train into a town or settlement, and not ride into town on your horse and witness the train already there waiting at the station. You must ALREADY be in a town or settlement and see trains regularly pulling through. If you can confirm that the trains are spawning normally in settlements with the current mods you have installed, then I'd say your game is functioning acceptably. If not, then your spawns are still not working correctly.

Edited May 19 by Dagger323

[Sanah 17]

7 hours ago, kaozzz777 said:

The goal here imo is to make aware of the problem and how to test reliably, If you want to know which mods run with native systems working as intended.
For that you need a test suite with reproducible results and compare numbers. Monitoring scripts, pool sizes and know which entities to get spawning. At set weathers, times and locations. That can be done if you combine Daggers tests with my approach.
 

 

There won't be a magical solution. Play vanilla, assured everything works like it should. Mod only after you done everyting and go wild. Or go the dark souls route.

 

Have fun

No one here is expecting a "magical solution". I don't even know what your point is. It appears you have not tested city train spawns, since you gave no input on that. You just jumped to the conclusion that the problem is AB's ScriptHook, although the issues are present in every possible version of ScriptHook. Besides, I have never had error FFFFF when buying train tickets anywhere in the map when using AB's.

Anyway, I really don't know what exactly do you think you have accomplished. Part of the goal here is to spread the word, right, but you are missing the other part. Not making fake claims and not doing half-arsed tests is equally as important.

Edited May 19 by Sanah

[kaozzz777 18]

10 hours ago, Dagger323 said:

I appreciate the information you've provided. But have you tested for train spawns in towns? Rhodes, Valentine, Saint Denis, etc? Because they are the biggest determiner of train spawns working correctly. The route past Flatneck Station that I mentioned before is good for testing wilderness train spawns, but that's only half the equation. Using the Railroad Engineer mod, and waiting in a populated town, should nearly always result in a train spawning and coming through the town once an icon shows on the map that a train is approaching and you wait for it to come past you. In vanilla, that's what happens; however, if mods are breaking the spawning, then the icon will remain grayed out and no train will appear no matter how long you wait. The tests must be conducted at every town and settlement. Rhodes, Saint Denis and Annesburg (not counting the two trains that are always sitting idle in the Annesburg depo) almost never show trains when mods are breaking spawns. Valentine will, but usually only around 9–10 P.M. at night. To conduct these tests, you must be IN TOWN and able to witness the train pull into or past the station, not be out in the wilderness and merely follow the train into a town or settlement, and not ride into town on your horse and witness the train already there waiting at the station. You must ALREADY be in a town or settlement and see trains regularly pulling through. If you can confirm that the trains are spawning normally in settlements with the current mods you have installed, then I'd say your game is functioning acceptably. If not, then your spawns are still not working correctly.

OK, to fully understand train spawns more precise data is needed.
Do you need to have reached a certain story point for that to happen? For example, if you use a clean chapter 2 start save without progess, you haven't done the train mission yet. The game has flags called world states.
Where are you waiting, in the train station or simply near it is enough? Do you have a range for regularly? Is 2 enough, or 3 or 4?


If you test a modded version with RaildroadEngineer, you would need to know what RE is doing/causing. To many variables. The blips alone might or might not be correct. I still think it's not suited to validate results.

[kaozzz777 18]

4 hours ago, Sanah said:

No one here is expecting a "magical solution". I don't even know what your point is. It appears you have not tested city train spawns, since you gave no input on that. You just jumped to the conclusion that the problem is AB's ScriptHook, although the issues are present in every possible version of ScriptHook. Besides, I have never had error FFFFF when buying train tickets anywhere in the map when using AB's.

Anyway, I really don't know what exactly do you think you have accomplished. Part of the goal here is to spread the word, right, but you are missing the other part. Not making fake claims and not doing half-arsed tests is equally as important.

Half-arsed is bitching and moaning while adding nothing. If you don't understand importance of monitoring scripts and setting up a test envoirnment then don't comment on it. I didn't claim anything, I said I think I got a modlist working and that V2 should be used, because more native scripts were running in my benchmark. If i forgot to include something, shit happens. What bugs you had or not doesn't matter for other peoples bugs. Enough time wasted on you.
 

[Sanah 17]

3 hours ago, kaozzz777 said:

The blips alone might or might not be correct

And, by your logic, RampageTrainer somehow gets everything right. That makes lots of sense.

My bugs don't matter, but yours do?

By your same logic, no one should give a flying fuck if you got "a modlist working" or not.

You still have not done proper train spawn tests yet, but somehow you have gotten "a modlist working". Yeah, you found a sweetspot, congrats, like most people have. You kept modding the game until it totally broke. You don't need RampageTrainer for that.

 

Your tests are half-arsed and that's a fact. Even you have admitted to refuse doing proper tests because "too many variables". If you are just going to keep spreading misinformation I would suggest you to get off here and keep playing with your "working modlist" while the rest keep investiganting the issue.