It may be time to give Fallout 4 the 2023 AE treatment
I love my janky, crazy-quilt Fallout 4 built that I made for Charley's story. It contains tons of custom assets, lots of playthrough-specific details and tweaks, and it was a solid learning experience that directly informed the development of my highly stable 2023 AE build for Skyrim. And, no small detail, it gave me the platform for a 94-chapter fanfic that dwarfs any of the other stories I've written for this blog.
Unfortunately, it's also become super unstable. I booted it back up this weekend to try and write some more for Charley's story, and I was shocked by how badly it was performing. Not only were the load times prohibitive and the actual game performance kind of weak, but I was encountering serious problems like infinite loading screen from the main menu or signs that the game was getting close to overflowing its memory constraints almost as soon as I loaded in (even though I was nowhere near my full memory capacity). Of course, FO4 has always been a little janky when it comes to memory handling.
With a real sense of regret and disappointment, I think it may be time to pull the plug on Charley's story and build a new world to play with. Of course, this won't be an overnight process. I have no intention of deleting all the custom assets and other work I did for Charley's world, so I'll need to take some time to carefully go through all of that material and sort it out, and ideally peel apart my big patch plugin at the same time so I won't have to reinvent the wheel in terms of the database.
And I can already tell some changes I'll want to make in the new world. First and foremost, we won't be starting up the game with a million settlement add-ons active. I learned very pointedly in Charley's adventure that not only was there no reason to start the game with those loaded, but that at times it was actively counterproductive to do so. Case in point, Salem and its whole loyalty cascade problem with the gun turrets. There's also a lot of bloaty mods that, while they were superficially good ideas, I've learned my lesson about and won't be including again. One example is Boston Natural Surroundings, which adds tons of trees that cause all kinds of geometry conflicts and also add to the draw call load. Another example is High Poly Faces (and indeed, 8K skin textures in general), which are surely "part of the problem" for memory-related breakdowns in the facegen system.
Of course we'll want to get Sim Settlements 3 installed and running, and we'll want to lean on it more for our settlement building, with far fewer manual-placement type mods installed. There'll be a few certainly for the NSFW stuff or for narrative purposes, but settlement add-ons are definitely "part of the problem" when it comes to bloat in this current build, something that shows with occasional resource related script lag weirdness. Most that add random new weapons will also be on the chopping block; I keep trying these out and they never live up to expectations. On the other hand, clothing mods like VTAW's and TheKite's stuff will definitely be staying, since making fun custom outfits was a very satisfying part of Charley's playthrough.
The process of building the next world will probably take a while, in no small part because many mods have been updated or developed alternatives, and everything is going to need to be scrutinized carefully for what it adds. Still, that should be a fun process in its own right, and I'll look forward to sharing my progress with you all as that starts to take shape.
Finally, in wrapping up Charley's story, I'll share with you my notes for what would have happened next. I had kind of roughly outlined where the story would go through Book 8, and while the technical limits of the engine mean that story won't be written in full, I'd like to share at least a glimpse into what I was planning on writing. So stay tuned, Charley's story isn't quite done yet.
2 Comments
Recommended Comments