Tukuk-OS Wiki

Tukuk-OS Is Alive: Reflections From Broken Laptop to Working Search Engine

From Tukuk-OS Wiki, the free knowledge base.

Tukuk-OS Is Alive: Reflections From Broken Laptop to Working Search Engine
TypeEncyclopedia article
SubjectStory
AuthorFevian Donald
Published28 October 2026
Revised28 October 2026
LicenseFree to read & cite

An honest reflection after Tukuk-OS started working: what was hardest, what I would do differently, and what keeps me writing code.

Advertisement

One night I opened tukuk.org from my phone outside the house, typed a question, and watched results appear within a few dozen milliseconds. Right then I remembered the laptop that died years ago — the flickering screen, the fan growing quieter, and the despair of losing my only learning tool.

Why I Started

I did not start with the goal of building a search engine. I started with the goal of never stopping learning. Every step — Linux, HTML, JavaScript, Node.js — was completed because I wanted to see one more result tomorrow. When those steps began to add up, I realised I had the skills to build something bigger than a tutorial.

Tukuk-OS was not an idea I planned from day one. It was born out of frustration: other search engines felt too full of ads, too much tracking, and too few answers. I wanted search that was simple, honest, and mine.

The Hardest Part

Honestly, the hardest part was not writing code. It was maintenance. A server running 24 hours needs logs, backups, monitoring, and a plan for when everything fails at three in the morning.

I once had a night where the NASA API failed, the browser showed broken images, and the contact form stopped sending email — all at the same time. That night taught me to build systems that fail gracefully: honest error messages, retry buttons, and fallbacks that kick in automatically.

What I Would Do Differently

If I started over, I would write tests first. At first I relied on "manual testing in the browser", which worked until I had too many routes to remember. Automated testing changed how I add new features: I can now add a route and know confidently that old routes are not broken.

I would also write blog content from the start. I put off writing until the system was ready, while actually writing helps me think about the system more clearly. Writing about what I build exposes weaknesses I did not notice while writing the code.

What Stays

Three things I will not give up:

  • Honesty in errors. When an external service fails I tell the user. No deception, no confusing blank screens.

  • Trust in open data. NASA, USGS, the World Bank — open data makes Tukuk-OS meaningful without selling anyone's information.

  • Ownership. My code, my domain, my index, my backups. The freedom to change, shut down, or rebuild any time is the real motivation behind this project.

What Comes Next for Tukuk-OS

A good search engine is never really finished. My backlog is long, and I list it here not as a promise but as a principle: every improvement must make search more useful or the system more resilient, not merely add features.

Among the near-term plans is improving Malay-language coverage in the index — most of my sources are English today, and Malay readers deserve equally good results in their own language. I also want to add advanced search modes: exact phrase matching, term exclusion, and finer date filtering.

On the infrastructure side I want to move monitoring beyond a simple liveness check to real metrics — p95 response time, error rate, and result counts — so I can see slow degradation before it becomes a visible problem. And I want to test restore procedures regularly, because an untested backup is just hope stored in gzip format.

What I look forward to most, however, is content. A search engine without valuable content is an empty pipe. Every new article, every snippet improvement, and every added data source makes that pipe more useful to the people who rely on it.

If You Are Just Starting

My advice: do not wait for a new laptop, faster internet, or a perfect mood. Start with one empty file today. Learn one command. Write one page. Hook it up to a server. Repeat tomorrow.

The journey from a broken laptop to a living search engine is not a story about talent. It is a story about consistency — line by line, day by day — until one day you open your work from your phone and realise you have arrived. And the most exciting part: you can keep building from here.

Advertisement

See also

Categories: Story · Reflections · Tukuk-OS
Advertisement