Backround

About 6 years ago now, the Ireland OSM community had a bunch of online and face-to-face discussions. There was a desire to have a common campaign, rather than everyone just paddling their own canoe, mapping old boundaries, addding 110KV monster pylons, plotting the holy stones of Clonrickert, or whatever you are having yourself.

Why Buildings?

And so #osmIRL_buildings was born. It took lots of months to pull together. There was a discussion document put out, and lots of decisions and guidance via videos, long conversations on Telegram, and frequent issues discussed on the mailing lists. The task was designed with a few things in mind. Firstly, the community recognized that compared to other territories, we had relatively small levels of completion of buildings. Secondly, a prominent academic had stated that the Irish Government knew more about the number and condition of cattle than it did about buildings. Thirdly, a lot of citizen science projects were trying to collate and capture where derelict and disused buildings were located in cities, with the hope that they might be repurposed for housing. Fourthly, there was a National Planning Framework launched in 2018 that concluded that the spaces for the next 1 million people to live in could not be sprawl outside of Ireland’s cities and towns. There were other reasons too, but those are the ones I recall, so apologies to all those other reasons and their proponents. Nevertheless, all of the ones I mention here could have been addressed by the creation of a fully open spatial dataset of the buildings on the island, and not what passes for open data by data.gov.ie.

Reservations

Of course there were detractors; some mappers worried about the threat of being inundated by the glibness of millions of “building”=”yes” objects. One man in Kilkenny was worried that the climate would have changed by the time the task would finish. He might yet be right.

Kick Off

The #osmIRL_buildings campaign began with gusto and with a small number of regular contributors, county-level tasks on a task manager. What we aimed to do was make a base-map, in other words a very foundational level map, with correct building shapes and all buildings attempted. We had lots of good banter we got going and mapped several counties, Kilkenny, Sligo and Carlow being first across the line.

My Break

Four years into it I stopped making regular contributions, but egged everyone else on and promised to come back. And I kept my promise this year initially hittting StreetComplete, and leveraging the fresh new buildings and completing quests and adding addresses out on the ground. However, that led me to realise that the basic mapping wasn’t very accurate. That ranged from wrong guesses at terraces, to whole buildings being missing and lots of buildings left with the “yes” and “house” tags.

Current Offerings in the tasks

So this drew me back to JOSM, where I started to see the horror of the full picture. This included tasks mapped where existing buildings left by earlier mappers were left “as-is”. Where mappers added objects without trying to offset (so you are left with 2+ offsets in a small area). Although the tasks were very specific about “attempt all buildings” and “provide the correct shape” that’s miles away from what is now being done.

How it was meant to be

I won’t speculate here about the drop in standards of what we are doing, but I will point out that long-range campaigns bear their own special risks, and if we had a time machine I would use it to ensure mappers gets breaks and avoid burnouts where the collective quality drops to this standard.

If you are a craft mapper we need your help in validating and mapping these tasks according to their instructions. Please help us come together and improve our mapping quality.