1] Use a SCM tool (git, plastic, etc...) - never be afraid to delete old, obsolete stuff from your project. You can't reason about your code if it's just a big mess.
Comments
Log in with your Bluesky account to leave a comment
2] Mess with geometry on the fly. The engines make it seem very static - import mesh from a 3D program and that's it. But be aware that you can also construct meshes on the fly or mess with vertices/UVs in scripts and in shaders. It's fun 😊
4] Connect google sheets - some things are better edited with google sheets, especially helpful with multiple collaborators, for things like translations or game design parameters (e.g. weapon stats). Connecting to google sheets is easy using their Javascript API
5] Measure before Optimizing - some things are fast, some things are slow, without profiling how can you be really sure you've made things faster? see also: premature optimization is the root of all evil.
6] Make it work, than make it fast - connected to the last point, try to first validate your game design ideas, before spending time optimizing their code. You might cut them entirely and that time would be wasted.
7] Your CPU is very, very fast. If you're working with structs, tightly packed in memory, you can easily update thousands of items per frame, without perf. issues. If you iterate over a list of class instances, scattered around in memory, the CPU will be mostly idle waiting for the memory.
Comments
4] Connect google sheets - some things are better edited with google sheets, especially helpful with multiple collaborators, for things like translations or game design parameters (e.g. weapon stats). Connecting to google sheets is easy using their Javascript API