2018-05-28 02:00 skrev Warin:
[ ... ]
Actually there are large areas that have flats/apartments with the
ground floor as shops ... so the suggested
landuse=residential and landuse:secondary=retail is what is happening
on the ground in some places.
Random thoughts:
Does this really buy anything compared to overlapping polygons (except,
obviously, where the polygons are coincident, in which case it saves a
polygon)?
I’m a bit nervous that primary/secodary/tertiary is a bit too fuzzy. In
a city, primary is street level, secondary is above, while in the
forest, primary is the forest and secondary is what’s on the floor?
With linear objects, such as streets, we gladly chop them into umpteen
pieces to put the right tags on each piece (this bit is no-parking, this
stretch is a pedestrian zone, those 50 meters are one-way, all of them
are named "Main Street"). Do we want to do that for polygons as well
(extreme case: for every point on the surface of the planet, there is
exactly one polygon that covers that point and holds *all* the tags that
apply at that point) or do we want to work with overlapping polygons?
_______________________________________________
Tagging mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/tagging
