[Tagging] Cycling infrastructure routes (was Re: cyclist profiles - was:Feature Proposal - RFC - value 'basic_network' - cycle_network?)

[Minh Nguyen]

Vào lúc 17:58 2021-12-01, stevea đã viết:
> There really are two distinct, mathematically unique "sets of things" that are different from one another:  a "road way" which is a real thing with a name, and signposts along it which denote that.  OSM models the former as a way with a name=* tag and might model (does so in the case of certain bicycle and hiking routes/networks) these as a or some "signposted node route(s)/network(s)" (and there seems to be a method of doing that everybody agrees upon, now denoted with network_type:node_network, perhaps others).
> 
> If there is a Main Street with 1st Avenue, 2nd Avenue, et cetera as perpendiculars to it, let's say the city didn't get around to signposting "Main Street" at 6th Avenue and 7th Avenue, but did again at 8th and 9th Avenues.  Main Street didn't cease to be Main Street around 6th and 7th.  Yet, OSM can properly model both of these distinct "sets of things" in the real world:  as a way with a name=* tag, and as a relation of nodes representing signposts, where importantly, there is a "skip in the data" because "there is a skip in reality."  There can be both, OSM can model both.  Yes, the latter are rare, the former are much more common, but so what?

I was thinking of gaps in physical infrastructure, not just gaps in 
signposting. At least in my dialect of American English, words like 
"bikeway" and "trail" can informally refer to either highway=cycleway or 
my hypothetical route=cycleway, gaps and all.

For example, the Simon Kenton Trail [1] in Springfield, Ohio, has an 
on-road gap on one street. [2] Such gaps are gradually being closed as 
dedicated bike paths are built, but for now cyclists just cope with them 
by riding on the shoulder. There happens to be a state bike route 
concurrent with the SKT, but there's also sufficient signposting to 
ignore that designated route and instead follow the SKT itself as a 
route, despite the decades-long detour.

As the state bike route system becomes more well-known, people may stop 
following trails by name like this. If the gap isn't closed by then, it 
might make sense to delete the route relation and transfer the 
wikidata=* tag onto each individual highway=cycleway way (ignoring the 
highway=residential ways). Or, we could retag the existing route 
relation as something else, like route=cycleway.

[1] https://www.openstreetmap.org/relation/1237291
[2] https://www.openstreetmap.org/way/19007995

-- 
minh at nguyen.cincinnati.oh.us