Mateusz Konieczny wrote: > I am not claiming that there are quick & easy ways to reduce > complexity - but complexity has some real negative consequences.
The cycleway tagging mess has come about because a bunch of wikifiddlers keep inventing ever more spurious and unnecessary tags - highway=path, bicycle=designated, bicycle=official, and so on. That's an argument to distrust the wiki, not to give it more authority. If we were to elevate the wiki to 'MUST' status, you'd have to support a new set of tags every time ten people voted on one obscure wiki page. At least, with the current situation, any tag needs to get some sort of critical mass before clients need to worry about supporting it. FWIW, I process cycleways extensively for cycle.travel's map and routing, covering Western Europe and North America. The variant tags aren't great but they're not that much of a problem - nothing that a few ifs and tables can't solve (Lua ftw). The three serious problems I do encounter are: 1. Granular tags (notoriously highway=path) with missing information. highway=path without both access= and surface= tags is pretty much useless. 2. Differing densities of data, such as Germany with its countless highway=tracks - makes effective cartography from a single stylesheet very difficult. 3. Bad imported data, principally but not exclusively TIGER. Richard -- View this message in context: http://gis.19327.n5.nabble.com/Applicability-of-wiki-tagging-and-votes-may-should-or-must-tp5866166p5866307.html Sent from the General Discussion mailing list archive at Nabble.com. _______________________________________________ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
