Proposal:Natural=rock cleanup

From OpenStreetMap Wiki
Jump to navigation Jump to search
natural=rock cleanup
Proposal status: Approved (active)
Proposed by: fkv
Tagging: natural=rock
Applies to: node area
Definition: cleanup definition and usage of natural=rock (A remarkable rock or group of rocks with connection to the ground.)
Statistics:

Rendered as: a rock symbol, e.g. an upright isosceles triangle
Draft started: 2013-07-09
RFC start: 2013-07-09
Vote start: 2014-08-12
Vote end: 2014-08-26

Abstract

There are more than 100000 uses of natural=rock, but the documentation is vague and unusable for both mappers and application developers. This proposal aims to clean up the wiki page and to enable a subsequent cleanup of the data.

Current status

According to the Wiki page, natural=rock

  • "has been suggested for mapping areas of rocks and small skerries that can be used for navigation or that might be a potential danger for navigation". No statement on actual usage.
  • was used by an import for what would be tagged natural=bare_rock nowadays.

Similar tags

  • natural=cliff indicates a line where you can fall down. There seems to be no consensus on whether natural=cliff may be set on a node, but steepness would matter even in that case.
  • natural=bare_rock This indicates rock surface. It does not make a statement on form or significance.
  • natural=bedrock This proposed tag is synonymous to natural=bare_rock and is somehow deprecated by the approval of the latter.
  • natural=stone A rock that is "not attached to the underlying terrain".

The definition for natural=rock should refrain from overlapping with these tags. Therefore, natural=rock should not be used for rock surface, neither should it be used for glacial erratics and alike.

On the other hand, there are still many rocks where those tags don't apply. E.g.:

  • "rocks and small skerries that can be used for navigation or that might be a potential danger for navigation", as in the current Wiki text
  • unique rocks and rock formations that are historically, touristically or otherwise significant, but which cannot be tagged as natural=stone because they are at least partly attached to the bedrock.

What about tourism=attraction?

tourism=attraction alone does not bear any physical information, i.e. what this actually is. However, natural=rock + tourism=attraction are a useful combination wherever rocks are of touristical interest.

Cleanup the data

The biggest issue are probably the objects that are tagged natural=rock where it should be natural=bare_rock or other. We should:

  1. change natural=rock to natural=bare_rock on all objects from the french import
  2. find out other imports where lots of natural=rock come from, and what natural=rock stands for in those cases. Update them too if necessary.
  3. Use an heuristic approach on the remaining natural=rock objects. E.g. change areas with an extent of at least 100 m to natural=bare_rock, and do the same when extent is at least 30m and no name or *ref* is set. Don't touch any nodes.

As these edits are subject to the Mechanical_Edit_Policy, they cannot be part of this proposal, but should be discussed in the (local) mailing lists instead.

In the thread "convert imported natural=rock areas to bare_rock" in the tagging mailing list it has been suggested that many areas imported as natural=rock deserve manual remapping anyway.

Suggested definition for natural=rock

A remarkable rock or group of rocks, with at least one of them firmly attached to the underlying bedrock.

Examples

node 2316453536
<-- No, this is not a natural=stone !
Residents made a bet with more than 100 French Soldiers that they could not move that rock. The Frenchmen pushed and pulled and were working with poles, but the rock did not move a millimeter.

Voting

  • I approve this proposal I approve this proposal. --Fkv (talk) 11:27, 12 August 2014 (UTC)
  • I approve this proposal I approve this proposal. -- HillWithSmallFields (talk) 13:54, 12 August 2014 (UTC)
  • I oppose this proposal I oppose this proposal. - Sorry but "with at least one of them firmy attached to the underlying bedrock" is impossible to verify except in some rare situation of precise geological survey Mateusz Konieczny (talk) 06:03, 13 August 2014 (UTC)
  • I approve this proposal I approve this proposal. -- Math1985 (talk) 14:25, 15 August 2014 (UTC) Thanks for doing this. The description might be a bit vague, but the pictures clearly show what is intended.
  • I approve this proposal I approve this proposal. --Neuhausr (talk) 14:48, 15 August 2014 (UTC)
  • I approve this proposal I approve this proposal. --Yoko99 (talk) 15:36, 15 August 2014 (UTC)
  • I approve this proposal I approve this proposal. --Hedaja (talk) 16:05, 16 August 2014 (UTC)
  • I approve this proposal I approve this proposal. -- Please note though that "remarkable" is a curious word to use in English since it implies that the feature is amazing or unexpected or suchlike, a more emotive term than needed here. I think it can be trivially changed to "notable" or suchlike, but I think that's minor enough that we can ignore that for now. --Danstowell (talk) 11:06, 18 August 2014 (UTC)
  • I approve this proposal I approve this proposal. --Sarchittuorg (talk) 11:08, 18 August 2014 (UTC)
  • I approve this proposal I approve this proposal. --Unixasket (talk) 05:42, 20 August 2014 (UTC)
  • I approve this proposal I approve this proposal. ...but change to "notable" as per Danstowell. --Gormo (talk) 07:14, 20 August 2014 (UTC)
  • I approve this proposal I approve this proposal. Still somewhat uncomfortable about restricting the mappable rocks to "remarkable" (or "notable") rocks, but overall the proposal is sound. I'm hoping for inclusive usage of the tag in practice. --Tordanik 13:31, 20 August 2014 (UTC)
  • I approve this proposal I approve this proposal. --CMartin (talk) 16:12, 20 August 2014 (UTC)
  • I approve this proposal I approve this proposal. --Saxonyking (talk) 20:23, 20 August 2014 (UTC)
  • I approve this proposal I approve this proposal. --Hamgom95 (talk) 19:00, 30 August 2014 (UTC)

=> This proposal was approved with 13:1 valid votes and 1 late vote. Two users request a change in wording ("notable" instead of "remarkable").

Too late votes

  • I oppose this proposal I oppose this proposal. -- As said by Mateusz Konieczny distinction with natural=stone on the ground is hard and probably of no interest to most mappers (rock:attached=yes better ?). It has been approved for surface while it conflicts with natural=bare_rock and Current (as of 02/2015) description isn't what was in the original proposal. sletuffe (talk) 16:27, 14 February 2015 (UTC)