[OSM-talk] Do we care if its forest or wood? Natural world mapping ...

[David Lynch]

On Mon, Jul 20, 2009 at 11:59, Tom Chance<tom at acrewoods.net> wrote:
> On Monday 20 Jul 2009 17:08:30 Andrew Ayre wrote:
>> I've been adding the national forests in Arizona, and the Wikipedia
>> definition doesn't fit too well. There are areas here that are inside an
>> administrative boundary called a National Forest where the trees are
>> very sparse - 10s of meters between them. Elsewhere in the forest the
>> trees are dense but it is a gradual transition from sparse to dense that
>> could take 50 miles or more to travel through.
>
> The point is that we won't ever find a useful correspondence between real "out
> there in the world" uses of "Forest" and "Wood" (which are already very
> inconsistent), everyone's individual perceptions of the difference, dictionary
> / encyclopedia / professional definitions, and the reality of the slightly
> chaotic OSM tagging.
>
> The division of landuse and natural, forest and wood, is utterly pointless.
>
> Hence my proposal to only use natural=wood, and allow further tags to
> designate the type of tree, whether it's used for commercial logging, etc.

IMO, national forests fall into a third category, which neither your
proposal nor current tagging covers - land which is designated by
government as a forest which is preserved or managed under special
restrictions. The trees don't necessarily stop at the boundary, and it
is possible that there are areas within the boundary which aren't
covered in dense trees. It probably needs some kind of boundary=____
tag.

I'm also thinking that deprecating both landuse=forest and
natural=wood might be a good idea if this goes forward. Replace it
with natural=trees, which is just as self-explanitory, and which (to
this particular mapper) sounds like a better fit for small clumps of
<10 trees than "wood."
-- 
David J. Lynch
djlynch at gmail.com