Skip to content
This repository has been archived by the owner on Jul 24, 2021. It is now read-only.

Closed ways tagged as buildings are not rendered #3932

Closed
openstreetmap-trac opened this issue Jul 23, 2021 · 8 comments
Closed

Closed ways tagged as buildings are not rendered #3932

openstreetmap-trac opened this issue Jul 23, 2021 · 8 comments

Comments

@openstreetmap-trac
Copy link

Reporter: Pyrokrat
[Submitted to the original trac issue database at 10.17am, Monday, 25th July 2011]

Mapnik keeps on refusing to render way 122186232 (and also others in this area) as a building while it renders the adjacent way 122184832, which is tagged identically, correctly. Osmarender shows it correctly.

Rendered correctly:
http://www.openstreetmap.org/browse/way/122184832

Invisible:
http://www.openstreetmap.org/browse/way/122186232

Both ways result from splitting a multipolygon (large block of buildings) into smaller ways (individual buildings) thus obsoleting the multipolygon relation. The relation itself was deleted manually during this process.

@openstreetmap-trac
Copy link
Author

Author: Ldp
[Added to the original trac issue at 8.03pm, Monday, 25th July 2011]

http://lists.openstreetmap.org/pipermail/talk/2011-July/059522.html ?

OSM isn't rendering tiles at the moment. Tiles from the backup static cache may be 10 days old. Your edits are less than 10 days old. Let's wait for the main server to return to service.

@openstreetmap-trac
Copy link
Author

Author: Pyrokrat
[Added to the original trac issue at 6.04am, Tuesday, 26th July 2011]

This behavior is observed since Jul-19, and I see all of my newer edits to show up almost immediately (except for this recent server outage). See my initial question on OSM help here

http://help.openstreetmap.org/questions/6422/invisible-buildings

I suspect that this is not related to server availability.

@openstreetmap-trac
Copy link
Author

Author: Ldp
[Added to the original trac issue at 1.24pm, Thursday, 15th September 2011]

Most likely, something has gone wrong during the application of diffs, and the building geometry for way 122186232 isn't actually in the database for mapnik to render.

Jon, can you confirm this theory?

@openstreetmap-trac
Copy link
Author

Author: JoshD
[Added to the original trac issue at 2.53pm, Friday, 13th January 2012]

This building and others nearby seem to render just fine now.

@openstreetmap-trac
Copy link
Author

Author: Pyrokrat
[Added to the original trac issue at 2.10pm, Wednesday, 8th February 2012]

Even after cleaning local tile caches, the reported building still doesn't render on Mapnik. Maybe you checked on Osmarender when closing this defect. If desired, I can attach current screenshots.

@openstreetmap-trac
Copy link
Author

Author: amm
[Added to the original trac issue at 10.30pm, Wednesday, 30th January 2013]

This sounds very much like bug #4525 in osm2pgsql.

A way that is part of a multi-polygon is removed from the database as its own polygon. However, if the way is taken out of the multi-polygon it is not reprocessed and thus remains deleted from the rendering tables.

There is a fix for it, but I am not sure what it does to performance and so I'd like to get an opinion from other osm2pgsql developers if this is the best way to do it, before applying the patch.

@openstreetmap-trac
Copy link
Author

Author: Pyrokrat
[Added to the original trac issue at 7.14am, Thursday, 31st January 2013]

The buildings in question were in fact a multipolygon before I broke them down into individual building shapes (ways).

@openstreetmap-trac
Copy link
Author

Author: Andy Allan
[Added to the original trac issue at 6.46pm, Monday, 9th December 2013]

If amm is correct, then this wasn't an issue with the stylesheets and is in any case fixed.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

No branches or pull requests

1 participant