This weekend, I gave myself a wild little challenge: build a Python package in just two days.
It’s called InkCollector—a CLI tool that collects data about the Disney Lorcana TCG. I didn’t start with a grand vision—just a curious itch to make something useful and fun. And surprisingly… I pulled it off. 🎉
🧗‍♂️ What Didn’t Go So Well
At first, I did what I always do: tried to finish full features before committing. I had a bunch of tabs open, code half-done, test cases in my head, and docs that weren’t updated. Then I got distracted. Twice. Maybe more than twice.
Boom. Lost progress. No commits. Just vibes and regret.
âś… What Did Work
So I made a pivot. Instead of doing a lot, I started doing a little—and doing it often:
- One commit to tweak a function
- One commit to update the README
- One commit just to clean formatting
Tiny steps. Each one moved me forward, and I felt way less overwhelmed. Debugging got easier. The project felt lighter—like it was breathing with me instead of fighting me.
Also, I went old school and used pen and paper to make lists. ✍️
🤖 I Used AI—And I’m Glad I Did
Yup, I used AI while building this. It helped with boilerplate, untangled a few weird moments in the code, and honestly made the whole process way more fun.
Even this journal entry? Drafted with AI. I’m not above it—I’m into it.
đź’ˇ What I Learned
InkCollector isn’t perfect, but it’s alive. I shipped it, documented it, tested it, and even gave it a little polish. It’s on GitHub now, ready for feedback and improvement.
Here’s the link if Future Me wants to look back (or if someone else wants to tinker with it):
đź”— https://github.com/bertcafecito/inkcollector
Signing off for now. The package is live. I feel accomplished.
– Bert 🧙‍♂️💻
Top comments (0)