Luquan in Hebei, China is missing

Avatar
  • updated
  • Not a bug

In OSM, Luquan (3009751) is a subarea in Shijiazhuang (3009732), but the admin area doesn't appear in OSM boundaries tree.

Pinned replies
Avatar
Magnus
  • Answer
  • Not a bug

OSM-Boundaries supports multiple extracts of OpenStreetMap's planet-extracts. The current latest fully imported version is 2021-02-01 and in that extract China does exist. The latest version that's partially ready is 2021-03-15, China exists there as well. It actually exists in 2021-03-08 as well, but there it has the wrong admin-level which causes issues with the tree view. The fact that it has the wrong admin-level has nothing to do with OSM-Boundaries, this is how the data looked in OpenStreetMap at that point in time.

Since OSM-Boundaries supports multiple extracts, you can choose another DB and look into that data set to see how it looks there.

Relation 3009732 does exist in 2021-03-15 for example, though 3009751 doesn't exist there. Relation 3009751 hasn't existed since database 2020-09-07. This is because the data is broken in OpenStreetMap in one way or another. If you for example visit https://tools.geofabrik.de/osmi/?view=areas, type in Luquan into the search box and then select Luquan District (second alternative) you will get a map. If you zoom into the pink circles you can see that the relation isn't properly closed. Since it's not a proper polygon we can't display it as such. You can actually see it at Openstreetmap.org as well, if you know where to look.

We import polygon data according to the rules mentioned in the Documentation, but for the import process to work the source data must also be correct.

OpenStreetMap is crowd sourced with thousands of active users. There are a lot of new issues created each day. If we were to fix all these errors ourselves we would probably need more than ten full timed employees doing just that. The best we can do is to help show the issues in OpenStreetMap and provide the historical versions for download.

Avatar
Magnus
  • Answer
  • Not a bug

OSM-Boundaries supports multiple extracts of OpenStreetMap's planet-extracts. The current latest fully imported version is 2021-02-01 and in that extract China does exist. The latest version that's partially ready is 2021-03-15, China exists there as well. It actually exists in 2021-03-08 as well, but there it has the wrong admin-level which causes issues with the tree view. The fact that it has the wrong admin-level has nothing to do with OSM-Boundaries, this is how the data looked in OpenStreetMap at that point in time.

Since OSM-Boundaries supports multiple extracts, you can choose another DB and look into that data set to see how it looks there.

Relation 3009732 does exist in 2021-03-15 for example, though 3009751 doesn't exist there. Relation 3009751 hasn't existed since database 2020-09-07. This is because the data is broken in OpenStreetMap in one way or another. If you for example visit https://tools.geofabrik.de/osmi/?view=areas, type in Luquan into the search box and then select Luquan District (second alternative) you will get a map. If you zoom into the pink circles you can see that the relation isn't properly closed. Since it's not a proper polygon we can't display it as such. You can actually see it at Openstreetmap.org as well, if you know where to look.

We import polygon data according to the rules mentioned in the Documentation, but for the import process to work the source data must also be correct.

OpenStreetMap is crowd sourced with thousands of active users. There are a lot of new issues created each day. If we were to fix all these errors ourselves we would probably need more than ten full timed employees doing just that. The best we can do is to help show the issues in OpenStreetMap and provide the historical versions for download.

Avatar
alexander misel
Quote from Magnus

OSM-Boundaries supports multiple extracts of OpenStreetMap's planet-extracts. The current latest fully imported version is 2021-02-01 and in that extract China does exist. The latest version that's partially ready is 2021-03-15, China exists there as well. It actually exists in 2021-03-08 as well, but there it has the wrong admin-level which causes issues with the tree view. The fact that it has the wrong admin-level has nothing to do with OSM-Boundaries, this is how the data looked in OpenStreetMap at that point in time.

Since OSM-Boundaries supports multiple extracts, you can choose another DB and look into that data set to see how it looks there.

Relation 3009732 does exist in 2021-03-15 for example, though 3009751 doesn't exist there. Relation 3009751 hasn't existed since database 2020-09-07. This is because the data is broken in OpenStreetMap in one way or another. If you for example visit https://tools.geofabrik.de/osmi/?view=areas, type in Luquan into the search box and then select Luquan District (second alternative) you will get a map. If you zoom into the pink circles you can see that the relation isn't properly closed. Since it's not a proper polygon we can't display it as such. You can actually see it at Openstreetmap.org as well, if you know where to look.

We import polygon data according to the rules mentioned in the Documentation, but for the import process to work the source data must also be correct.

OpenStreetMap is crowd sourced with thousands of active users. There are a lot of new issues created each day. If we were to fix all these errors ourselves we would probably need more than ten full timed employees doing just that. The best we can do is to help show the issues in OpenStreetMap and provide the historical versions for download.

Thank you for your detailed explanation. I fixed that boundary on OSM.