This came up because when I documented "cycleway:lanes"[1], I initially wrote that bicycle:lanes is an alternative (which for me, it is) but a user complained that it's "cycleway:lanes" which denotes this and I've decided to remove this after a while after some back and forth.
The reason why I see "bicycle:lanes" as an alternative to "cycleway:lanes" is that "cycleway=lane" is an artifact when per-lane access wasn't invented back in the day (akin to how "cycleway=opposite" became obsolete by "oneway:bicycle=no" for one-way traffic, my proposal to deprecate "busway=lane" because of the overlap with "bus:lanes" and I also consider "cycleway=share_busway" to also be deprecated) and most likely never would have existed had bicycle:lanes been invented first. Case in point, I don't see any purpose in using "cycleway:lanes" at all when "bicycle:lanes" exist and use at most "cycleway:<side>" to satisfy QA like SC. (I know that bicycle lanes was brought up during the proposal but given my stance on "busway=lane", I would actually advise against "cycleway:lanes" if I were an active OSM user during the proposal.) The problem is that the opposite party mentioned about potential customer-only bicycle lanes and that the only way to be sure is with "cycleway:lanes". I countered this with two problems: What about motorcycles (since there is no widespread "motorcycleway" tag) and are tags like "customer" and "private" even defined for "access:lanes" (which IMO should gets its own page instead of redirecting to "access")? The idea with the latter is that ":lanes" never went in depth with the "access" key and although "access:lanes"-exclusive values were invented, they were after "*:lanes" was defined. [1] https://wiki.openstreetmap.org/wiki/Key:cycleway:lanes _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
