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
we have integrated OpenFreeMap in osmapp.org, but there is an issue with type encoded in the id field in the MVT tiles. The last number in the id should encode 0=node, 1=way and 4=relation.
It is decoded as 1->way, thus way/1601837931, and osmapp brings us to this URL osmapp.org/way/1601837931
same issue is with almost all active POIs on the map.
Expected result:
In reality it is a node, so it should end with 0. If we use the default layer Basic from Maptiler, it works correctly.
.
Relevant issues / related code:
Maptiler also had a long time issues with wrong encoding IDs, I believe they were using openmaptiles internally, but later switched to thier own impl (planetiler?) and that needed fixes to the id encoding too.
Luckily it's a single change, OFM uses planetiler and planetiler uses this system
Yep that's the new behavior as of #826 - the vector tile feature IDs are osm ID * 10 + {1 for OSM nodes, 2 for OSM ways, 3 for OSM relations, 0 for any other source} so you can tell whether the feature came from a node, way or relation. You can set --feature-source-id-multiplier=false if you want to disable the behavior and just set vector tile feature IDs to the original ID of the osm element.
Hi,
we have integrated OpenFreeMap in osmapp.org, but there is an issue with type encoded in the
id
field in the MVT tiles. The last number in theid
should encode 0=node, 1=way and 4=relation.Example
16018379311
way/1601837931
, and osmapp brings us to this URL osmapp.org/way/1601837931Expected result:
In reality it is a node, so it should end with
0
. If we use the default layer Basic from Maptiler, it works correctly..
Relevant issues / related code:
Maptiler also had a long time issues with wrong encoding IDs, I believe they were using openmaptiles internally, but later switched to thier own impl (planetiler?) and that needed fixes to the id encoding too.
(same issue in indoor= stack: indoorequal/indoorequal#49)
The text was updated successfully, but these errors were encountered: