2012/7/2 Martin Vonwald (Imagic) <[hidden email]>:
> Am 02.07.2012 um 22:09 schrieb sabas88 <[hidden email]>:
>> I'd opt for landcover system.
> +1 for landcover. IMO the tag natural should not be used for areas (yes, I know, currently it is used often for areas).
I think it is fine to use natural for areas (the very most of over 8
million features tagged with natural=* are actually areas:
http://taginfo.openstreetmap.org/keys/natural ), but I'd like to see
it used for topographical features (most of all features that are
currently tagged with natural are indeed topographical features, e.g.
lakes, bays, woods, springs, beaches, ...)
I am opposing the interpretation of "natural" as a class in contrast
to "man_made", as it is not a sufficient distinction (too few main
classes, hence it leads to exceptions and inconsequencies (everything
natural besides x, y and foo, and bar, and z, and w, and t and...
which are covered by the keys ...)).
The key isn´t perfect as previous years discussion have showed, it would be better if there where a recognized key for landcover. The key "natural" suggests that it is a topographic feature, and in some places it could be so. The key "landcover" isn´t the best name for a key either as this tag suggests an area without any cover, that is the bare rock.
The value "bare_rock" used to tag areas of uncovered badrock does have some competition.
The obvious one being the shorter "rock", that word do have too many similar uses for tagging: a single large boulder, a underwater hazard, a small skerrie in the sea, a larger steep-faced isle or even similar steepfaced hills on land. all of these being typical easily identified geogrphic objects usually tagged using the key "natural".
Another good alternative would be the value "bedrock", as it makes clear that it is a solid surface of rock, in fact the bedrock, showing up. My initial trouble with this was that it might ruin future attempts to mmake geological tags for the bedrock, even when the bedrock isn´t visible.