You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jul 24, 2021. It is now read-only.
Reporter: namuori [Submitted to the original trac issue database at 3.36pm, Friday, 7th September 2012]
The following relations fail to appear in Nominatim search despite being in OSM data for quite some time. Please add them to the database - these are admin_level=4 (equivalent to provinces) areas, so having them missing yields incorrect search results for a large part of South Korea. I initially thought the relations were in the database, but were not being classified correctly. It turns out they were just missing there all this time.
Click on the Nominatim error link to see the attempt to find the relation ending in a simple error page.
Author: lonvia [Added to the original trac issue at 4.03pm, Saturday, 8th September 2012]
I only checked the first two. The boundaries look really strange right about here: http://osm.org/go/55kcIAft
You'll see that the boundary ways double back on themselves. Fix that and the relations will show up in Nominatim correctly. The other examples most likely have similar problems.
Author: namuori [Added to the original trac issue at 4.48am, Monday, 10th September 2012]
Replying to [comment:1 lonvia]:
I only checked the first two. The boundaries look really strange right about here: http://osm.org/go/55kcIAft
You'll see that the boundary ways double back on themselves. Fix that and the relations will show up in Nominatim correctly. The other examples most likely have similar problems.
That defect is now fixed. Inspecting the rendered map visually, I also found a similar defect along Sejong-Chungcheongbuk-do border, and another one along Daejeon-Chungcheongnam-do. If these three defects are all that's causing the problem, the error should disappear soon, I suppose. If it doesn't, I will need a better tool to check for bad polygons because OSM Inspector website has failed to pick out any of these defects, hence the visual check.
Author: namuori [Added to the original trac issue at 4.33pm, Tuesday, 11th September 2012]
After some careful manual checking of the ways and finding one more defect on the polygon, Nominatim now shows all the affected relations. This ticket will be closed.
Reporter: namuori
[Submitted to the original trac issue database at 3.36pm, Friday, 7th September 2012]
The following relations fail to appear in Nominatim search despite being in OSM data for quite some time. Please add them to the database - these are admin_level=4 (equivalent to provinces) areas, so having them missing yields incorrect search results for a large part of South Korea. I initially thought the relations were in the database, but were not being classified correctly. It turns out they were just missing there all this time.
Click on the Nominatim error link to see the attempt to find the relation ending in a simple error page.
Gyeonggi-do
http://www.openstreetmap.org/browse/relation/2306392
Error Nominatim: http://nominatim.openstreetmap.org/details?osmtype=R&osmid=2306392
Incheon
http://www.openstreetmap.org/browse/relation/2297419
Error Nominatim: http://nominatim.openstreetmap.org/details?osmtype=R&osmid=2297419
Daejeon
http://www.openstreetmap.org/browse/relation/2349984
Error Nominatim: http://nominatim.openstreetmap.org/details?osmtype=R&osmid=2349984
Sejong
http://www.openstreetmap.org/browse/relation/2349795
Error Nominatim: http://nominatim.openstreetmap.org/details?osmtype=R&osmid=2349795
Chungcheongbuk-do
http://www.openstreetmap.org/browse/relation/2327258
Error Nominatim: http://nominatim.openstreetmap.org/details?osmtype=R&osmid=2327258
Chungcheongnam-do
http://www.openstreetmap.org/browse/relation/2327259
Error Nominatim: http://nominatim.openstreetmap.org/details?osmtype=R&osmid=2327259
The text was updated successfully, but these errors were encountered: