October 2, 2026
Why I'm Writing Again
A blog is a public notebook. Here's what I plan to put in mine and why I stopped treating writing as something separate from building.
I've been building software for eight years. In that time I've shipped a credit platform for 4,400 personnel, a set of Web3 games with live on-chain data, a handful of Chrome extensions and VS Code extensions that real people install, and more failed experiments than I care to count.
I've written almost none of it down.
That was a mistake.
Why write
Two reasons.
The first is selfish. Writing forces clarity. Half the bugs I've fixed, I fixed while trying to explain them to someone else. Half the architectural decisions that turned out well, turned out well because I'd already written the reasoning down somewhere and could see the holes.
The second is that most of what I've learned is not written anywhere in a form I recognize. There are a thousand tutorials on "how to use RabbitMQ." There are almost none on "here's what happens when you run a credit engine on RabbitMQ for 4,400 users and the bank API takes 40 seconds to respond on a Tuesday."
That second kind of knowledge only exists in the heads of people who've actually done the work. If I don't write it down, it disappears when I move on.
What this blog is
A public notebook.
Not a tutorial site. Not a thought-leadership channel. Not a content marketing play. Just a place where I write down things I figured out, in enough detail that future-me (and maybe someone else) can use them.
Three kinds of posts, roughly:
- Things I built how and why, with the parts that went wrong left in. The Nafisagold credit engine. The Chrome extension that isolates cookies per tab. The scroll restoration workaround that took three attempts to get right.
- Things I learned the specific mechanics. How D1 differs from R1 and when it matters. Why Lottie players break during Nuxt hydration. What idempotency keys actually buy you in a payment system.
- Things I think opinions on the industry that I'd defend in an argument. The specialist-vs-generalist debate. Why tools matter more than frameworks. Why "software is a means, not an end" is not a platitude.
What this blog isn't
- It's not on a schedule. Posts appear when I have something to say, not to hit a quota.
- It's not SEO-first. I'll write titles that are honest, not ones optimized for search.
- It's not comprehensive. If you're looking for a complete guide to X, this is the wrong place. If you're looking for one specific hard-won detail about X that nobody else has written down, maybe not.
Why now
Two things converged.
First, I finally have enough distance from the earlier work to see what was actually interesting about it. When you're in the middle of shipping, everything feels equally urgent. A year later, you can tell which decisions held up and which didn't.
Second, I've been watching the AI-era conversation about engineering and realizing I disagree with most of it. The "specialist is dead" takes, the "just learn to prompt" takes, the "10x engineer" takes they're not wrong exactly, but they're not right either. I have specific things to say about this and no good place to say them except here.
So this is the place. First real post comes next week.
If you want to follow along, I post shorter notes on Telegram. The code for everything I write about is on GitHub.