On Sun, Aug 16, 2015 at 1:27 AM, Friedrich Volkmann <[hidden email]> wrote:
> On 03.08.2015 00:55, Daniel Koć wrote:
>> I have just discovered that while landcover=trees has no Wiki page, it's
>> quite established tag (I wouldn't say "popular" here, because it's just
>> about 1% of forest/wood uses) and we could officially define as a generic
>> tag for trees areas, when it's not clear for the mapper if it's natural or
>> not ("forest" vs "wood").
> No, because the landcover=* key is just nonsense. There is no definition for
> that key. What does landcover mean? Vegetation? Soil? Atmosphere? Buildings?
> Ocean? Everything we map is landcover in some respect. You could use
> landcover=motorway instead of highway=motorway, and landcover=playground
> instead of leisure=playground. The landcover key matches all and nothing.
It's true that this key is mainly supported by a very small group. I
wouldn't say it is "established". The taginfo stats are relatively
stable and 760 users is peanuts 5 years after the proposal in OSM
(11/2010). One reason is probably because it's not rendered in main
But another reason is that most of the constributors draw such
polygons from the aerial imagery and they don't care about the
difference between land "use" and land "cover". When they see a piece
of grass or forest, they just want to draw a single polygon and use a
single tag for "grass" and "forest". Live must be easy for the
contributors. And promoting a new key which may replace or be combined
or partially overlap the old "landuse" will be extremely confusing for
the non-experts. And this will be endless, like the "footway vs path"