Currently, any nodes that have no tags and aren't referenced by any ways or relations are discarded.
In the context of updating, retaining these otherwise useless nodes may become important. New or modified ways may reference these orphan nodes, but since they are neither features nor incorporated into ways as anonymous nodes, we have no way of knowing their location.
Since updates are non-transactional, it is also possible that new nodes could be included in one .osc file, while a way referencing them may arrive in a subsequent .osc file.
To avoid this problem, orphan nodes should be upgraded to feature nodes (at least until they are referenced by a way) by tagging them with geodesk:orphan=true. This would have the added benefit of allowing QA tools to query them (a[geodesk:orphan]).
Currently, any nodes that have no tags and aren't referenced by any ways or relations are discarded.
In the context of updating, retaining these otherwise useless nodes may become important. New or modified ways may reference these orphan nodes, but since they are neither features nor incorporated into ways as anonymous nodes, we have no way of knowing their location.
Since updates are non-transactional, it is also possible that new nodes could be included in one
.oscfile, while a way referencing them may arrive in a subsequent.oscfile.To avoid this problem, orphan nodes should be upgraded to feature nodes (at least until they are referenced by a way) by tagging them with
geodesk:orphan=true. This would have the added benefit of allowing QA tools to query them (a[geodesk:orphan]).