About

In first person · Updated 2026

I got into this because I wanted to know why the internet broke every time I unplugged a cable. That is a stupid reason to start, and it turned out to be the only reason that mattered.

My name is Mohamad Dafinsa Tauladani Setiawan, and I go by Dafinsa. I build web systems, deploy them to servers I operate myself, and keep the network underneath them alive. That combination is usually split across three people. I ended up doing all three because each layer kept pulling me into the next one.

If all you want are my job titles, the homepage already lists them and there is no reason to repeat them here. What follows is why I do this work and how I think about it, which is the part a resume cannot say.

I started with network engineering. My vocational school years were spent learning why packets go where they go, why subnet masks matter, and how to break a router in exactly the wrong way so that I would have to learn how to fix it. The first server I ever set up was a disaster. Apache refused to start for reasons I did not understand, the firewall blocked everything including me, and I spent three days learning that "connection refused" and "connection timed out" mean very different things.

That kind of failure is addictive if you are the sort of person who needs to know why. I kept pulling on the thread. The machines we networked were serving web pages, so I wanted to know who wrote them. First a static HTML file. Then PHP. Then databases, sessions, authentication, and everything else a modern web application assumes you already know.

The part that surprised me, and that still shapes how I work today, is how much the two halves feed into each other. Writing the application made me a better operator, because I knew what the server was actually doing. Running the server made me a better developer, because I had to watch my own code fail under load and learn from it.

Every layer I learned made the layer below it less mysterious and the layer above it more interesting.

Boring beats clever

The cleverest solution in the room is usually the one nobody can debug at three in the morning. I try to write code and configure systems that a tired version of me can still read six months later. That usually means picking the option I have already seen work over the one I want to try for the first time. I have never once regretted that choice.

If it is not monitored, it is not running

A server that returns HTTP 200 is not necessarily healthy. A cron job that runs without logging is a rumor. I treat observability as part of the build, not something to bolt on after launch. Logs, health checks, and alerts are the difference between finding out from a dashboard and finding out from a user.

Write it down

Every time I have had to recover a system from memory alone, I have regretted not writing the runbook first. Documentation is not paperwork. It is a gift to whoever inherits the thing when you are not around, and if you are the only person running it, that someone is going to be a future version of you.

I am interning at Global Media Data Prima, where most of my time goes into keeping production systems stable and learning how a real operations team moves. In parallel I maintain a small self-hosted environment for my own projects: a Proxmox host, a handful of containers, reverse proxying through Nginx, certificates issued automatically, DNS and traffic routed through Cloudflare.

This site is one of the things running there. I am rebuilding it piece by piece, mostly to give myself a place to write that is not someone else's platform and does not disappear when the platform does.

I am also working through the last stages of the BNSP Junior Network Engineer certification material, because passing the assessment and actually understanding the material turned out to be two different things.

I read a lot of network engineering writing, which surprises nobody. I also read about how other people design and write software, which surprises more people. The two are closer than they look: both are about understanding how things fail before they fail.

When I am not at a terminal I am usually writing about what I just learned, mostly because explaining it is how I find out whether I actually understood it. Some of that ends up in the writings section. Some of it stays in a drafts folder forever.

I recently started learning Go, partly because I wanted a compiled language in my toolkit and partly because I missed the feeling of a compiler telling me I was wrong. It is going slowly, and that is fine.

Contact

If any of this resonates

I am open to freelance, contract, and full-time work. Email is the best way to reach me, and I usually reply within a day.