Download websites that use Luna

About Luna

Programming languages are rarely visible from the outside of a website. Server code does not announce itself in a delivered page, so when a language appears in a web technology index it is usually because something adjacent leaked into public view: a documentation site, a build artefact, a hosted editor, a project page generated by the toolchain. Luna arrives in our index labelled Augmented reality, and taking that label at face value would be a mistake about what has actually been observed.

A second complication is the name. Luna is shared by more than one product, and hardware, streaming and eyewear retail businesses have all used the word. The structured facts we hold describe something written in Haskell and released under Apache 2.0, first recorded in 2016, which fits the visual and textual programming environment rather than any of the other products that share the word. The address recorded against this entry now serves an eyewear retail technology company, so the domain no longer identifies the project those structured facts describe. Our figure counts a fingerprint match, and a fingerprint cannot separate two identically named things on its own.

The implementation language is informative in its own right. Building a language in Haskell points to a small team comfortable with type inference and functional design, aiming at expressiveness rather than mass adoption, and shipping to an audience of engineers and analysts rather than to site owners. Tooling that comes out of that world tends to be adopted in narrow pockets, to leave a public footprint made of documentation and demos rather than of production deployments, and to change names or fold into successor projects without warning.

As of October 2026 the number of hosts we associate with Luna is 1. A number produced this way should be read as detections of a public artefact, because nobody installs a programming language on a web server the way they install a content management system. What the count tracks is the visible edge of an ecosystem: the sites that publish, document or demonstrate the language, not the machines where code written in it runs.

For a researcher mapping developer tooling, the value sits in the host list rather than in the total. Pulling the matching domains out of an export lets you look at them together, check whether the pattern is one coherent project or a collision of names, and inspect what else runs alongside. A per-domain lookup cannot support that work at all, because a single site never reveals whether its match belongs to a population or stands alone.