What Is Tukuk-OS? What This Search Engine Actually Does

From Tukuk-OS Wiki, the free knowledge base.

A full overview of Tukuk-OS: what it does, how it differs from other search engines, and why it is built on a free, open source stack.

Tukuk-OS is a web search engine I built myself, starting on one old computer at home and now reachable at tukuk.org. It indexes websites, offers image and video search, displays live NASA space data, and provides a blogging platform — all in one lightweight application with no user tracking.

What Tukuk-OS Is Not

Before explaining what it does, it matters to state what it is not. It is not a global-scale engine indexing billions of pages every day. It is not a replacement for Google on every query. And it is not a service that sells user data — because no user data is collected to sell.

Tukuk-OS is a focused, restrained search engine: a curated set of sources, high result quality for the topics it covers, and full respect for privacy. It was built to help people find answers, not to exploit their attention.

A Simple Architecture

Under the hood, Tukuk-OS is a Node.js/Express application that talks to Meilisearch for text search. The server exposes APIs under /api and /search, while the browser presents the interface. All static files — pages, themes, scripts — are served from the public directory.

This design is deliberately small. One process, one server, one index. When something fails it is easy to know where: the server log tells you within minutes. Large, complex projects take weeks to diagnose; Tukuk-OS takes minutes.

Headline Features

• Standard web search with snippets, language and site filters, and scoring that prefers matches in titles.

• Image and video search from external sources with result pagination and a "load more" button.

• NASA data: the daily picture (APOD), near-Earth asteroids, Earth views from EPIC, sounds of space, and NASA media search.

• A technical blog with original content, quality benchmarks, and structured data for search engines.

• A contact form that sends real email through Cloudflare Email Routing.

An Honest Comparison With Other Engines

It is fair to state where Tukuk-OS wins and where it loses. Compared to global search engines, Tukuk-OS is far smaller in coverage — it does not have billions of pages, and it does not track you to personalise results. If you are looking for breaking news or something very niche, a bigger engine may give you more results.

Where Tukuk-OS wins is focus and transparency. Snippets come straight from the target page, filters are clear and work instantly, and results are not influenced by whoever pays to appear. NASA pages, technical blogs, and media search all live in one place without switching between several services.

The design philosophy is also different: missing features are features. No weather widget, no news feed, no experience designed to keep you inside the app. You come, you search, you get answers, you leave — and to me that is how a search engine should behave.

Data Sources and Ethics

Much of Tukuk-OS's value comes from open sources. NASA opens almost all of its data to anyone, and Tukuk-OS takes full advantage of that permission. For image and video search I send requests to public providers and cap request rates so nobody gets an excessive load.

My crawler also reads robots.txt for every site before it walks, honouring restrictions and rate limits. A good search engine should be a courteous guest on other people's web.

The Cost Model

Tukuk-OS runs on open source software (Ubuntu, Node.js, Meilisearch) on hardware I already own. The real cost is just the annual domain and maintenance time. There is no monthly server bill because it is self-hosted, and traffic flows through a Cloudflare Tunnel without opening any port on my router.

This model is not right for every project, but for a personal search engine at moderate scale it produces near-zero cost with total control.

Who Should Use It

If you want search that is simple, free of intrusive ads, and honest about what lives on the target page, Tukuk-OS is for you. If you need minute-by-minute news coverage you may need a bigger engine — and that is fine.

Most importantly, Tukuk-OS is proof that you can build a search engine yourself today without large capital, without a team, and without selling anyone's privacy. Everything I use to build and run it is free or nearly free, and every design decision is explained on this blog.

See also

• Building an Inverted Index by Hand: My First Search Engine Code
• Meilisearch: From a Hand-Built Index to a Fast Production Engine
• My First Crawler: The Spider That Collected the World
• Tukuk-OS Is Alive: Reflections From Broken Laptop to Working Search Engine
• Tukuk-OS Complete Feature Tour: From Web Search to Space Data
• Image and Video Search in Tukuk-OS: How It Actually Works