Content pipeline first

I know, I should be doing game slicing. Like, implementing basic gameplay and testing ideas that way.
But for some reason I want to start with whatever I'm currently doing.

The thing is: I know, it's not easy to optimize a game so that it looks great and runs great.
I already made a game a couple of years ago called Speed for Needs (take a guess what that's about kekw) and it ran at 30fps on my 1080ti back then. It was an Initial D- or Need for Speed Carbon-inspired downhill 1v1 race at night.
It was fun, also has a trailer that I am truly proud to have made to this day, but it ran really bad. i don't want to make the same mistakes again.

This is most likely why I'm trying to figure this stuff out before going in head-first, just so i can have the peace of mind that, whenever the game is fun, I also know how to put these assets together so it'll look great and still run just fine.

LOD stuff in Godot

In Godot there seem to be two ways of doing it.

First: Auto Mesh LOD, which you can select when importing meshes from Blender for example. However I can't tell you anything about it since it just doesn't work for me for whatever reason.
Maybe Auto Mesh LOD just didn't feel like it for the type of mesh I was throwing at it. Maybe I'll find out later.

Second: Just shipping 3 meshes in one .glb file and setting them up via the visibility range.

My lod1 is way too aggressive, dropping 13k tris in less than 8m distance, which looks horrendous XD. but anyways: it works, and i kinda know how now.
Moving further away, you should not be able to make out the lesser detail (although you can, since this is literally the first model I am testing this with :D)

Visual style tangent (optimizing my way out of thinking about gameplay)

So after I was testing LODs, i've started messing with shaders a bit. i did that before, right when i started with Godot. this time, there's a valid reason though!
i wanted to get vertex snapping. But it was so hard to do so?!
I don't know why, but the visual shader graph Unity provided just plain overwhelmed me. I do appreciate this for smaller things, but it took ages to get the look I wanted, the material shader I wanted.
With Godot it's like 10 lines. gg.

And while i was in there, i also found a way to get nice black and white color banding, which makes this somewhat jpg-artifact / comic styled look:

Finally thinking about gameplay

It's just not easy thinking deeply about yourself and projecting that onto a canvas for others to play and preferably even enjoy.
I tried to do it anyway, and I might have found a loop that I can aim for. It probably won't be the final version, but at least it's something.

I've started "development" of this whole project around june 2025. It felt easy, or even natural to create environments and think about how to flesh them out.
But when it came to the actual gameplay-side of things, I just wasn't able to think about anything fun tbh.
I had a core mechanic in my mind the whole time, but just didn't know what to do with it.

here is what i came up with so far for now (text shown on hover)
You wake up in an apartment. You walk through hallways getting towards the cellar (optionally a lift can be taken). In the cellar, there is a machine. You cannot quite make out the details, but it's obvious a lot of work has gone into it.
You're actively managing voltage, synchronization, temperature, energy flow, and can "craft" new parts.
All this is surrounded by a hyperfocus-system which can be beneficial or make things worse. Each work session ends abruptly. Next day things may have changed.
i won't go much more into detail here right now, but I think this is at least a start. Things evolve over time.

Docker image idea

I thought about moving from a bare configuration to a Docker image.

There are multiple reasons why.

First of all, it isn't that great to develop on the open heart.
In my case it's mostly fine, but I've already encountered issues where I was developing some feature that was still very broken.
However i couldn't leave it like that since the whole page would be broken. so: stash changes, revert branch… unnecessary hassle.

I've also checked out Tracy and came to the conclusion that i should've used it earlier, since it would've made debugging a lot easier :D
(It also makes Server errors look much better than the standard Apache ones in prod)
This is also a big reason as to develop locally. Tracy is awesome but it will show you EVERYTHING down to the DB credentials (which is not openly exposed to the internet but anyways)
You don't want to have this just ... sitting on your website ... in plain sight.

However (and i most likely misconfigured stuff or just don't know better), after some days I'm just having even more problems than before.
Every time I pull, my config gets reset. But when I don't bake the config inside the image, I can't make runtime changes to it?
And then weird permission stuff happened where the one thing I wanted from Docker - have one image here where it works, deploy it somewhere else and it will also work - doesn't work.
In my local install, I was able to upload files. In the version on the server, I couldn't. It threw an error every time.

i really don't know. For now, I'll stay natively deployed or find another way.
I did however just think about something in the middle. Using docker locally to develop and committing and pushing my modules (as I do already). And just pull them manually whenever I feel like they are ready.
Maybe I have to just give it a try.

Side note: writing every day

Just now, writing a little every day, I notice how fucking much I'm doing. That's not meant as a brag. I mean it like: how am i supposed to keep all this information in my head? It's just a lot.
Waking up, working full-time, having issues there of course too. Coming home and just continuing to work. I'm slowly getting worried since I do notice a constant feeling of exhaustion...