My first 3 weeks of game development

After consuming a huge amount of game devlogs during this situation, I decided to learn a bit and support one of the content creators that I was watching, buying this course. It was straight to the point and I really liked it. I chose Godot simply because is FOSS, by the way.

Game concept

Once the course was finished, I started this project with a clear concept about the mechanics, the scope and the quest design (more on the latter on future updates). The concept is overly simple: a First Person Hand, where the player can pull and push objects. The idea is to focus all the complexity of the game in the different objects and the interactions with them, keeping the actions of the player very minimal: move, push or pull.

Learning process and software

In this three last weeks, I have been learning more or less what it means to carry a solo indie game. Besides coding in Godot, I had to fight with Blender for 3D modeling (I basically did everything this guy says) and play with Audacity for recording and sound edition (the pig effects are not mine, those were free-licensed sounds). This brief experience has provided me with three BIG takeaways about workflow and deployment:

1. The Blender-to-Godot workflow is hard. Exporting the Rigify rig is hard. Blender 2.7 was fine with Better Collada but, after upgrading to 2.8 (two days after the first install), nothing was working, not even the official godot blender export plugin. What worked for me was to export to .glb, with apply modifiers and all the animations pushed down, whatever that means.
2. What works in debug in Godot may not work in release. Godot is an awesome tool that will just get better and more performant, but I was not expecting that. Therefore, to my future self, please use is_instance_valid() instead of != null. Also something about exporting *.lan for the SMRT-Godot dialog plugin to work. And those were 3 bugfixes.
3. The Blender-to-Godot workflow is still hard. Colors are hard. The colors in Godot change completely after importing from Blender, which I have partially solved by playing with the lights and the color spaces in the materials and world environments. It is still wrong.

Stuff I have to learn soon

I am currently using the godot-heightmap plugin for the terrain and the grass and some shaders here and there to make it look a bit cooler. The nav meshes are still a working mystery for me, and I believe it is something that is important. About level design, I am considering Qodot, but I have to do a bit of research to decide if I should go with it.

Next in the horizon

Having finished this prototype, my idea is to continue and upload a demo with the level design and quest concepts that I am looking for (hence, the Qodot part). Uploading one video devlog or two about these last 4 weeks would also be nice, but we will have to wait until I understand how to record and edit a video.