[OSM-talk] landcover=trees [was Re: OpenStreetMap Carto v2.33.0 release]

moltonel 3x Combo moltonel at gmail.com
Mon Aug 17 14:35:37 UTC 2015

On 16/08/2015, Mateusz Konieczny <matkoniecz at gmail.com> wrote:
> 2015-08-16 15:27 GMT+02:00 Greg Troxel <gdt at ir.bbn.com>:
>> landuse=forest does not imply the area is completely tree covered.
> Note that in typical usage it means exactly this. Maybe original intention
> was for that tag was to mean something else - but it is not changing how
> it is used by most mappers.

If only there was a good way to assert "typical usage", we might have
managed to standardise on it and solve the problem by now.

The landuse=forest definition of "area where trees are grown for
commercial purposes, expected to be covered by trees by default but
often also natural=scrub for about a decade after logging" is fairly
typical too. FWIW, that's the definition I've been using in my
mapping, as all the others (managed, named, size, etc) seemed very

If everyone adhered to my POV we wouldn't have a problem (sarcasm).
But I've given up hope of that happening, so the next best thing IMHO
is the landcover=trees reboot, which isn't perfect but which we can
hopefully agree on.

On 16/08/2015, Greg Troxel <gdt at ir.bbn.com> wrote:
> One could argue that natural=trees is a synonym for landcover=trees.

That'd work for me as well, it actually sounds much better. But
pragmatically I prefer to follow the more popular tag, unless I see
some strong consensus for natural=trees elsewhere.

More information about the talk mailing list