Hi, PJ Houser wrote:
1) Why are relations preferred for bike routes?
Because a stretch of way can belong to different routes at the same time, and this is difficult to model without relations.
2) In the database, how do relations apply to ways? The attributes associated with a relation - how are they tied to ways? Will routing software use the attributes in a relation to determine if a way is suitable?
In the central database, no assumptions about this are made whatsover. The software that builds the routing graph from OSM data would have to decide this. I think most people assume that relation attributes provide a default that can be overridden by way attributes but YMMV.
3) In Portland, Oregon, we have an interconnected series of unnamed bike boulevards - how should we split these into relations?
I'll pass on that.
4) Do shapefiles maintain relations?
I know of no shapefile generator that does. It would be possible to write one and maybe someone has, without my knowledge. The osm2shp from OSM's SVN certainly ignores relations.
Bye Frederik -- Frederik Ramm ## eMail [email protected] ## N49°00'09" E008°23'33" _______________________________________________ Talk-us mailing list [email protected] http://lists.openstreetmap.org/listinfo/talk-us
