Where I’m at

• I need to get much much better to writing tests (units, logic, integration). Starting with test-driven development.
• Still working through my personal crypto project, I’ve pretty much finished the placeholder UI for now. More or less we’ve got to figure out the data, and also the logic of the buy and sell functionality.

What I learnt

• Not so much learnt, but making things that you want to do mindless or effortless is going to be the key to doing the right things.
  • e.g. reviewing content when I create a new study journal entry is a great example of this.

What I did

• Started to write unit tests for my personal crypto project
• Created a zsh alias for what the powershell scripts are currently doing

What I reviewed

The content that I reviewed today was: Problem Solving: My own thoughts

• I really like how identifying the problem is broken into: context, issue and why do I care?. Makes sure that I have not gone overboard on trying to correct a minor issue.
• Breaking down the problem into smaller ones by asking questions, is a good way of getting to the atomic level which means you can start to write code.
• Writing psuedocode is a great way of starting to visually break up the problem, and get thoughts down on paper.
• Writing tests is a great way of starting, you write the test of what should happen, and then write code that works to that.
• I think using an LLM to get a starting prototype of the code first, even if it looks ugly, I’ll go through and refine it. That’s the best use of an LLM to help code at the moment.
• And then leave it so I can reflect, and then optimse it after it works according to the tests.

Also I added another comment on 30 July, and I tried to express it as trying to get to the atomic level. In other words, trying to break down the problem enough that you can write code for what psuedocode or documentation you have written to.