I've only been working on it for a couple days but here's some improvements so far:

• No more unlit shaders, now there is actual lighting

• The low res filter is gone

• Physics have been improved (more weight, no more vertical climbing, overall feels better)

• Actual 3d models inside the orb

• Animated backgrounds

• Replay system

• Less reliant on UI

• Much improved code base (hopefully it stays that way)

• Freely rotatable camera

I've been thinking about an Acceleroll 3.0 for a while, but let this be a lesson in code quality. Due to the game jam scope, pretty much after 2.0 the code base was entirely unmanageable. This is part of why things like target needed a full rewrite. Also adding new features was cumbersome. And I was not using Unity's systems properly, so any update to a shared object in levels, every single level needed to be manually updated any time a change was made. And honestly Acceleroll was always a bit crappy, but I liked the idea behind it.

An example of the code base being terrible is Unity's Input class was referenced basically everywhere instead of one universal class in my project. So when I made the Apple Silicon port, controller support broke. And fixing it would mean manually locating all references to the Input class, and reworking it to use Unity's new input system (which is terrible, but it's sort of built for things like that). This was one of many examples, and for a hobby project it was a tall order. So it's time to start fresh.

There were also some design things I did not like in Acceleroll. I felt like a lot of the levels were boring/too easy, and the hard levels were more fighting against the physics than an actual challenge (good example of that was the tightrope level (Hard mode W1-2). And difficulty was super weird and strange, it wasn't a curve, more like a spaghetti noodle. A hope with Acceleroll II is to bring back some challenge.