[Tagging] Feature Proposal - RFC - natural=bare_rock

[Johan Jönsson]

Steve Bennett <stevagewp at ...> writes:
> IMHO there are some subtle differences between these concepts:
> surface=rock
> landuse=rock
> natural=rock
> 
> The first to me suggests that the ground beneath some other feature,
> like a path or a park, is rock. surface=* is almost always a
> supporting tag, rather than a tag by itself.
> 
> The second is a bit odd, but would imply an area that is not used for
> anything because it's rocky - perhaps some kind of barren wasteland.
> 
> The third describes a geological feature that is useful as a landmark.
> There are trees over there, there are rocks over here.
> 
I agree, and further more, the word rock can mean a lot of things like skerries
and boulders. That is why the proposal is on bare_rock instead. An alternative
could be bedrock.

Regarding the first concept you mention: the ground in a feature. It could be of
bare_rock, in Sweden we have some cliff bathes that is some kind of beach with a
rock surface. I guess there could be roads on bare rock on some places in the
world, where the surface tag could come in use. Beach and road with subtag of
surface is probably to prefer over natural=bare_rock. But if there is no other
good tag for the area then you can use the landcover tag of natural=bare_rock,
instead of leaving it blank.

Regarding the second concept: landuse=rock that could be landuse=quarry
http://wiki.openstreetmap.org/wiki/Quarry

Regarding the third concept of geological landmarks. To get a more lively map
with nice landmarks there probably should be more tags like hillock,
stone_pillar, monolith, cliff, plateau, hill. The more detailed tagging on these
hills could use natural=bare_rock, natural=cliff, natural=scree for the parts
with rock surfaces and other tags for the vegetated parts. 
In the same way as the old abutters tag is the description of the terrain useful
to orient yourself: there are trees over there, there are bare_rock over here.

/Johan Jönsson