Local Wikipedia filters geocoded articles to what's within a half-hour walk

What the site does

Local Wikipedia, a project hosted at edjefferson.com that recently surfaced on Hacker News's front page, takes Wikipedia's corpus and narrows it to a single question: what is near me? According to the site's description, it shows only articles about things within a half-hour walk of your current position.

The tagline, "The Encyclopedia For Local People," signals the conceit. Instead of the sum of all human knowledge, you get a hyperlocal slice of it: the streets, buildings, monuments and other places around wherever you happen to be standing.

How the filtering works

The site produces none of its own content. Everything on it is pulled directly from Wikipedia, restricted to geocoded pages, meaning articles that carry coordinate data. Many Wikipedia entries about physical places include such coordinates in their metadata, and that is the property the project leans on: filter by distance from the reader, then render the ordinary article.

Two thresholds keep the result set usable. If fewer than 25 geocoded articles fall within the walking radius, the site widens its net to the closest 25 things within roughly 6 miles. At the other extreme, if more than 500 articles qualify, only the closest 500 are included. In other words, there is a guaranteed floor of 25 entries and a hard ceiling of 500, so the page stays populated without becoming overwhelming.

Location stays with the user

The project also makes a specific privacy claim: your location stays in your browser, and links you share do not include it. That matters for a tool whose entire premise is knowing where you are. Handling position on the client side means the site can do its job without accumulating a server-side record of where visitors are, and you can send someone a link without disclosing the spot you were reading from.

The site notes it has no affiliation with Wikipedia itself. It is a reuse of openly licensed content, Wikipedia's text being published under Creative Commons terms that permit this kind of republishing, rather than an official product.

Caveats

Coverage is only as good as Wikipedia's geocoding. The fallback radius of about 6 miles exists precisely because some places have fewer than 25 geocoded articles nearby, while the 500-item cap suggests dense, well-documented cities can produce far more than a comfortable browsing list. A reader in a city centre and a reader in the countryside will get very different experiences from the same tool, through no fault of the interface.

The half-hour-walk framing also raises questions the description does not answer. How walking time is converted into a geographic radius, and whether roads, rivers or terrain factor in, is not specified, so the figure is best read as the project's intuitive way of expressing proximity rather than a precise routing calculation.

Why it matters

The interesting idea here is not new content but a new lens on existing content. Wikipedia already holds an enormous amount of geographically anchored information, yet almost nobody browses it geographically. By re-slicing the corpus around the reader's position, Local Wikipedia turns a general reference work into a neighbourhood guide, a way of discovering that an unremarkable building on the corner has a history.

It is also a small case study in client-side design: a location-based service that does its filtering in the browser and shares nothing downstream. And it demonstrates how much structured metadata quietly sits inside Wikipedia, available to anyone willing to build an interface around it. If the approach catches on, expect more experiments that treat the encyclopedia less as a destination and more as a dataset.