Now that Godot Dash has had its first releases, I'm reconsidering my choice of pushing through the growing discomfort that is using Godot for me. I want to do what makes me happy, and working with Godot feels painful (mostly because of GDScript, but also partly because the UI framework is annoying to work with).
I feel split, because abandoning Godot Dash a second time[1] feels like betraying myself, and now that I have contributors, it feels like betraying them, like their efforts were for nothing.
I think the project needs a new name that isn't tied to the engine it's written in, that way I could maybe rewrite with peace of mind.
I want to try Bevy Engine so, so bad. I haven't gotten the occasion of getting into it for the sake of finishing the game, but this thing has been tempting me for more than a year and I am tired.
It's not just about porting code from an engine to the other. Bevy operates on a different paradigm that feels coherent with Rust. Some things in the current codebase are convoluted and I'm sure we could simplify the code of the Player class more (editor code is also a mess too).
The other problem is that the lack of an editor for Bevy[2] makes me worry about how we're going to make the editor UI if we can't tweak it live. The official editor hasn't landed yet, and other solutions, like Jackdaw, egui-based plugins or Bevy Skein, but they seem oriented towards making 3D assets and not 2D assets nor UI.
-
The first time being with Raylib Dash. ↩
-
The Bevy Editor is being worked on, the
bsnformat was a prerequisite for it to happen, so it's moving along nicely! Thebsn!macro released with Bevy 0.19, and it overhauls the way scenes are built, but the associated asset loader (forbsnfiles) doesn't exist yet. ↩