On 5/12/2020 10:58 PM, Paul Johnson wrote:
On Tue, May 12, 2020 at 9:37 PM brad <bradha...@fastmail.com <mailto:bradha...@fastmail.com>> wrote:OK, but it seems redundant to me. A trail/path get tagged as a path. There's a trailhead and a sign, it gets a tagged with a name. Why does it need to be a route also? Same reason all 0.11 miles of I 95 in Washington DC is part of a route. It's part of a route.
Yes but that's *part* of a route, a route relation with many other members. Brad's asking about single-member route relations. My understanding, still evolving, is that tagging conventions originally developed for long-distance walking routes -- thing like osmc:symbol, colour, distance, network -- are sometimes applicable to shorter trails, including those that are only a single highway=path/footway. Mappers reading the wiki page for osmc:symbol will be told that this tag is only to be used with route relations. Some mappers who want to add a symbol to a single-highway trail might tag osmc:symbol directly on the highway anyway (Taginfo shows 2924 instances of this) and some might create a single-member route relation. Another thing to consider -- for vehicle roads we have a many-tiered hierarchy from motorway down to track, which assists in routing and rendering. Paths and footways have no such hierarchy, so adding them to a relation along with the relation-specific tags is one technique mappers have used to call out trails of greater importance. Finally there's the issue of software and rendering support. Waymarked Trails, as Kevin mentioned, only supports route relations. I believe other hiking map renderers work similarly. Of course this is not how OSM is "supposed" to work -- structuring data for a particular renderer or software -- but nonetheless it is a factor in how people map. Jason
_______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
