I've been thinking a lot today about how motivation to persist with gamedev is so hard. Like, really hard. The time it takes to go from "idea" to "fun thing" when you are learning is just so long. It is hard to be patient.

But I just have to keep taking incremental steps forward.

On the other hand, I've also been reading "Game Feel" by Steve Swink, which has been really eye-opening. I've even been learning things from that book that are relevant for my daily work in data and analytics, which is great.

I hope I can work on the prototype tomorrow, but realistically not.

I think once I add sounds I want to set up some kind of experience and level-up mechanic, as well as a loot and inventory system, to get familiar with that stuff. I need to think about different unique enemy types and a boss, too.

Further, I was researching some sound effects and music. I find it so hard to decide on the aesthetic of the game. The style; everything from pixel art vs vector, color palette, and so forth. I can really appreciate the value of a professional artist in the project.

I also spent an hour working on a prototype for voice-activated control of projectiles, which was interesting, but ultimately the API for speach-to-text was too slow. I came up with a clever idea (I think) to mask the delay with an animation and some effects, but dropped it for now.

On my mind today was applying architectural diagrams (from i.e. data engineering and similar spaces) and design-thinking to indie gamedev. It seems like this is something that a lot of people would benefit from as I go through tutorials and courses online.

It's kind of hard to decide when I'll be "done" with this first prototype. In the meantime I'm trying to just churn up as many ideas as I can to work on next. In general, I'm learning a lot. It might be interesting to just proceed genre-by-genre, I don't know.

I don't think that's realistic, so i'll just be happy if I can add a few sounds and background music. Maybe if I'm lucky scoring and saving, I don't know. I think next week I'll focus almost solely on UI and saving / data visualisation / analytics / stats.

Work is pretty crazy right now, and my wife will be on call the next 3 days. I'm thinking I won't get much time to work on the prototype. It would be great if I could finish the full loop though from start to scoring and game over before sunday.

I've mastered the animation nodes and feel quite comfortable with TileMapLayer. It's interesting how a lot of things you can do with animations that you wouldn't necessarily associate with an animation. For instance disabling shapes and hiding sprites. I haven't done anything with sound yet.

It's also interesting how the LLM will always try code-first approaches for things, like adding nodes to groups or altering properties, when it's easier and more maintainable to do it in the UI. It will never suggest the UI either unless it's literally the only way, and then it's often wrong.

So far that is working well but it can make me frustrated. It gives me hints if I'm way off though.

I think the frustration is a good sign. LLMs take away your right to struggle with certain problems and you need that back if you want to learn to some extent.

I find that I can read and understand all of the code that the LLM like Claude generates, but it is hard for me when I get an idea to know where to start and which code, functions, or syntax to use.

To address this I've instructed Claude to withhold the answer and make me guess 3 times, first.

In the prototype, I now have a decent targeting schema where the user can switch between manual and auto-targeting. I even implemented a shader to highlight the targeted enemy without new sprites.

Data-driven elements are key, and it's something that I think I want for my games to be "unique". I want proper data visualization theory to be applied to the game UI and data... as well as lots of (useful!) statistics and export functionality.

I'm really excited for that.

The goal is to get a first game that is "playable" cold turkey. Meaning I can start it up, load a save file, play a level to try and score (and have it feel fun).

Then I can focus on learning enhancement elements like UI, sound, animations, particles, shaders, and also data-driven elements.