koriakine
Joined: 07/05/06
Places: 240
Wikimapia team

18 minutes ago
We started Wikimapia almost twenty years ago, in 2006. Since then, millions of places have been added and edited by people around the world, creating a geographic record that would be extremely difficult to reproduce today.

Keeping a project of this scale running for so many years has required continuous work. Wikimapia is a large and complex system, and maintaining its infrastructure, database, and legacy codebase has always been both resource-intensive and costly. That work has continued throughout the years, even as active development of new features gradually stalled.

Recently, however, the situation has started to change. New development tools, including agentic coding tools, make it much more practical to work with a large legacy system, understand its existing code, and improve it gradually rather than rebuild everything from scratch. This gives us an opportunity to develop Wikimapia more actively again.

The first priority is very practical: to make Wikimapia easier to edit. We want to simplify the work of people who continue adding, correcting, and maintaining information on the map, remove unnecessary friction, and gradually improve the tools available to editors.

A more detailed development roadmap will be prepared later. For now, the important thing is that Wikimapia is moving forward again.

After two decades, there is a great deal worth preserving, and still a lot that can be improved.

Wikimapia continues.

Alexandre Koriakine
Co-founder of Wikimapia

This is genuinely heartening news. Wikimapia is one of those rare projects that quietly accumulated something irreplaceable — millions of places, many of them mapped nowhere else, often with local knowledge that never made it into OpenStreetMap, Google Maps, or any commercial dataset. Twenty years of volunteer effort is a remarkable thing, and it's good to hear it won't just be left to slowly decay.

A few thoughts on the announcement itself:

**On the framing of AI/agentic coding tools**

It's honest and refreshing that you're not dressing this up as «we got new funding» or «we have a big team now.» The reality for most long-running legacy projects is exactly what you describe: the codebase becomes an archaeological site, and the cost of touching it exceeds the cost of leaving it alone. Agentic coding tools genuinely do change that calculus for some kinds of work — reading unfamiliar code, tracing call paths, writing the glue that nobody wanted to write by hand. That's a legitimate use case.

The risk worth watching: these tools are good at producing plausible code, and legacy systems are exactly where plausible-but-wrong changes cause the most damage. I'd hope the roadmap includes a real testing/verification strategy, not just faster edits.

**On «make Wikimapia easier to edit» as first priority**

This is the right call. The single biggest threat to Wikimapia has never been servers or storage — it's that editing became harder than the alternatives, and contributors drifted away. Anything that lowers friction for the people still actively maintaining places is worth more than a shiny new feature. Concretely, I'd guess the pain points are:

— The editor UI feeling dated compared to OSM's iD or even Google's contribution flow
— Moderation/review queues that make contributors wait or feel unappreciated
— Difficulty of bulk operations, imports, or fixing systematic errors
— Poor mobile contribution experience
— Uncertainty about whether edits will be seen, kept, or reverted

**What I'd gently push back on**

«A more detailed development roadmap will be prepared later» — fair enough for a first message, but the community has heard «we're back» before from projects of this age. The thing that will actually convince long-lapsed contributors is a small, visible, shipped improvement in the next few months, not a roadmap. Even something modest — a faster editor save, a clearer diff view, a fix to a long-standing bug — would do more for morale than a plan document.

**On preservation**

You're right that the dataset would be extremely difficult to reproduce today. That's an argument not just for keeping the servers up, but for thinking seriously about:

— Data export/dumps in a usable format, so the record survives even if the project doesn't
— Licensing clarity, so others can build on it
— Some form of redundancy or archival partnership

Twenty years is a long time. Most volunteer geo projects from 2006 are gone. Wikimapia is still here, and that's worth something.

Good luck with it. I'll be watching to see what actually ships