‘ve found myself drawn to a particular challenge: properly mapping the Rhondda Valleys in South Wales. It’s a task that’s proving both fascinating and frustrating in equal measure, but one that’s teaching me invaluable lessons about the complexities of local geography.

My approach has been quite straightforward - I’ve chosen to focus primarily on Ystrad, the area I know most intimately. This isn’t just because it’s familiar territory, but because understanding one place thoroughly seems like the best foundation for expanding outwards. Every quirk I discover, every mapping technique I develop, and every local peculiarity I uncover in Ystrad becomes a tool I can apply to neighbouring areas.

One of the most surprising aspects of this mapping project has been discovering just how much the “official” maps get wrong. I’ve taken to photographing street name signs and house numbers wherever I go - a habit that’s revealed some genuinely baffling discrepancies between reality and what appears on both Ordnance Survey maps and Google Maps.

The most striking example I’ve encountered is Bryn Terrace in Ystrad. This is a very real street, with real houses, real residents, and real post delivered daily. Yet somehow, it’s completely absent from both OS and Google mapping data. Walking down it, photographing the street sign, and checking my GPS coordinates left me questioning whether I was experiencing some sort of cartographic twilight zone. How does an entire street simply not exist in the digital world when it’s so clearly part of the physical one?

The house numbering systems throughout the Rhondda have provided another source of bewilderment. What might seem like straightforward sequential numbering often reveals itself to be anything but. Streets that appear to follow odd/even patterns suddenly throw in a completely out-of-sequence number, or entire ranges seem to have been skipped altogether.

These aren’t just historical quirks either - they’re active addresses that confuse delivery drivers and visitors alike. Mapping them accurately requires patient detective work, often involving conversations with locals who can explain the historical reasons behind these peculiarities. I recently found out about the note attribute, so I’m hoping to detail these nuances as I go along.

Each mapping session in teaches me something new about the challenges of accurately representing the real world in digital form. The lessons learned here - how to handle unusual address schemes, when to trust local knowledge over official sources, how to photograph evidence effectively - all prove invaluable when I venture into less familiar parts of the Rhondda.

What started as a simple desire to improve the map of my local area has evolved into a broader appreciation for the complexity of place. Every street name sign photographed, every GPS point recorded, and every correction made feels like a small victory in the ongoing battle to make OpenStreetMap truly representative of how people actually live and navigate their communities.

Our valley have decades of industrial and social history embedded in their geography, and much of that complexity isn’t captured in existing mapping data. Streets like Bryn Terrace might be missing entirely, whilst others are misnamed or incorrectly positioned.

By continuing this methodical approach, hopefully mastering Ystrad first, then expanding outwards with confidence - I’m hoping to contribute to a more accurate and useful map of this historically significant area of the world. After all, if we can’t get the basic building blocks right, how can we expect the map to serve the people who actually live here?

Onwards.