I have mapped perhaps tens of thousands of miles of bike routes in OSM. Yes, really. I don't do this sort of "apply the name of the route to the element track/path." We shouldn't.
Zeke's example is excellent and is a good reason for "route element naming" to be "case by case" rather than there be a "one-size-fits-all" approach, which simply cannot work for all cases. There are other examples (let's stick with bicycle routes) where it won't be a residential street, but a path which is specifically unnamed, but IS part of a route relation (which DOES have a name), as well as (almost?) every combination that can be thought of. If the path is unnamed, LEAVE it unnamed, even if it is part of one or more route relations. Indeed if it is part of one route relation, you might think you should name it, though, you shouldn't. If it is part of MORE than one route relation, you might think there is ambiguity and shouldn't name it, but Zeke presents us with an example where "the dominant" (I paraphrase what I think he means) route name DOES (happen to) influence the naming of the element way, but that is very much a "local rules apply" situation. In fact, I am of the strong opinion that ONLY "local rules apply" and OSM must strive to "name" things (like this, in relations) on a case-by-case basis. Because one-size-fits-all doesn't work in our real, messy world. The world is a messy place, meaning there can be and are inconsistencies when it comes to applying OSM "rules" for applying specific tags to specific features in our map. Yes, we should strive to apply (name=*) tags as correct as possible, verifiable on-the-ground and where applicable, both to ways correctly and to relations correctly. This might mean that a way is named, or not. It might mean that a named way is different than the route relation it belongs to (or even multiple route relations). Anything that can be imagined is likely possible in our messy real world, and our tagging should reflect our real world, rather than we should strive to apply artificial rules, like "route names should be applied to tracks/paths" (in a route relation) — UNLESS it is clearly known that they ARE named (or are named something else). My apologies of all of this is confusing, it's not meant to be, but the world is messy. Whew: what's the short version? Don't name elements of a relation after the named relation. (There may even be exceptions to this, like UNLESS the element really IS known by that name independently of the route — but I'm trying to keep the short version somewhat simple). So, I'm adding my +1 (to DaveF and Peter) that "the 'names' should be removed from these 'unnamed' things." Generally speaking and absent something else going on. This almost seems like it shouldn't have to be said, but, there: I said it. "Hard and fast rules" are seldom a good idea in OSM, as the world is messy. Sure, let's try to be consistent and complete, but let's not invent reasons for tagging things when there isn't a good one to do so. _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
