Tools & StackMay 15, 20264 min read

The Dumb Simple System I Use to Not Lose My Mind

Two days a week to code. One kid who doesn't care about my sprint velocity. This is what actually works.

An organized detailing shelf with labeled bottles and folded microfiber towels, an Audi Q7 behind

I used to lose track of everything.

Oil change dates. What polish I tried on the hood. Whether I'd actually applied that ceramic coating or just meant to. All of it floating around in my head or buried in some text thread from 2023.

Then I started writing things down.

Not in Notion. Not in some fancy car management app. Just a single note per car. Mileage, date, what I did. That's it. Turns out complexity is the enemy when you have about zero free time.

2 days
Per week I get to write code
3 tasks
The hard daily limit
1 file
The entire system, plain text

My Stack

For VaultBay - the SaaS I'm building - I keep the stack equally dumb.

Frontend
Next.js 14
I'm not trying to be clever here. It works, the documentation is solid, and I can move fast when Wednesday rolls around.
Backend
Supabase
Handles auth, database, and storage. I don't want to manage servers. That's not what I'm good at and it's not where I want to spend my time.
Hosting
Vercel
Push to main, it deploys. That's the whole CI/CD pipeline.
Payments
Stripe
Obviously.

That's it. No microservices. No Kubernetes. No GraphQL layer I have to maintain at 11pm after the kid finally sleeps.

What Actually Stuck

Nobody talks about this, but the tools don't matter as much as you'd think. What matters is having a system you actually stick to.

I tried Jira. I tried Linear. I even tried a Trello board at one point because someone on Twitter swore by it.

None of them stuck.

What stuck: a plain text file called TODO.md. The whole routine fits on one hand:

  1. Open it every Wednesday morning.
  2. Write down the three things I'm doing today.
  3. Do them.
  4. Close the laptop and be a dad.

I know how that sounds. But hear me out.

When you're coding two days a week, you don't have the luxury of managing a complex workflow system. The overhead of keeping a project management tool updated often exceeds the benefit it provides.

Time-Boxing Everything

I started time-boxing everything.

Not in a Pomodoro, productivity-app way. In a "this task gets two hours and then I move on" way.

Pro Tip

Give every task a hard time box, then move on whether it's perfect or not. The box is what kills the rabbit hole before it starts.

This sounds obvious but it completely changed how I build. Before, I'd get lost in a rabbit hole - debugging some weird edge case, refactoring code that was already working, adding features nobody asked for.

Now I ask one question before starting anything: does this directly serve the person using VaultBay?

If yes, it gets time. If no, it goes on the list.

The list is long. That's fine. It'll get there when it gets there.

What I'm NOT Using

Just to be clear about the constraints I'm operating under:

This isn't a recommendation. It's just the reality of someone working alone with two days a week and a baby who thinks 5am is a reasonable wake-up time.

The Point

You don't need the perfect stack. You need to start.

โ€” me, to past me

I wasted months trying to optimize my setup before I shipped anything. The "perfect" development environment doesn't exist. What exists is good enough to ship, iterate on, and improve over time.

VaultBay is far from done. It's rough around the edges. But it's real and people are using it.

That's better than the perfect app that only exists in my head.

If you're building something too, I'd genuinely like to hear about it.

๐Ÿ“„ Free Guide

The Garage System

Everything in this article, organized into a complete system. Free - just your email.

Get the free guide