Aggressively avoid over-engineering, restrict yourself to only whatever code is necessary to achieve your goal.

I have to repeatedly tell myself this when building out prototypes. It’s surprisingly difficult to not get lost in the weeds doing something not immediately necessary like setting up CI/CD or modularizing components.

I’ve always wanted to write a chess engine.

I agree with the premise, but this website makes me feel very stupid / unskilled. 2 months for a basic x86 kernel with keyboard and mouse drivers, elf loading etc? I don’t think 2 months would be enough for me to even partially complete one of the subtasks. Unless this is 2 months of 8-hour 5-day weeks but, no, even then I doubt I could do it.

Upvote for recommendation to make a GBA game. GBA is seriously a joy to dev for.

I guess, this is different from automating a part of your workflow every so often, in that you don’t yet know the surrounding context. It is the point to just dick around with some new context for a bit.

And if you get bored of it, i.e. you’ve learned enough about the new context, then you can just archive the repo and move on, because you don’t actually need it as a solution.

I always write toy software. I derive joy from writing in G’MIC.

I love this site !!

I have always wanted to make something but its just so time consuming. I also dont have ideas for something that isnt complex. …and I refuse to use an llm to do it, for me thats fake and not real learning.