Coastline inside coastlline

By chance, I found a closed natural=coastline way inside another closed natural=coastline way. It had been in the database for 9 years.

Curiously, the coastline tool of OSMI doesn’t detect this as a coastline issue, but this could obviously cause problems for people trying to consume OSM coastline data.

If you have a planet file loaded up, and like writing queries, making sure there aren’t any more of these would be helpful!

It is not detected in OSMI because it is not wrong. The Caspian Sea is completely surrounded by land.

Many years back when the softare behind the OSMI coastline check was developed we still mapped the great lakes in North America with “natural=coastline”…