[Tagging] cleanup of the key natural

Martin Koppenhoefer dieterdreist at gmail.com
Fri Oct 17 13:11:26 UTC 2014


2014-10-07 15:56 GMT+02:00 Brad Neuhauser <brad.neuhauser at gmail.com>:

> Hi Martin, regarding the wiki page, I'm assuming you're talking about
> Frankthetankk's edits? I'm not seeing what the issue is. Could you clarify
> what change you see as disputable?



sorry for taking that long to respond, I've been on vacation with reduced
internet access ;-)
I think the classification system that was introduced is arbitrary and has
some issues.

Some examples:
"vegetation related" contains at least the following objects that don't fit
IMHO. Of course you can relate everything to "vegetation" (either there
will be some kind, or there won't) but that doesn't make this automatically
a logical class.

- fell (bad tag anyway, word has different meanings and seems to be used
only in a limited regional context in the way that it is defined, the
description requires "principally covered with grass" but the main
qualifier seems to be that it is a "high lying" landscape).
- moor - is characterized by the humidity and acidic soils and the
elevation level, not the vegetation
- heath (is about certain soils)
- mud (this is about the grain size of minerals and the amount of moisture)
- sand (this is about the grain size of minerals)
- stone (this is about freestanding stones! Really nothing to do with
vegetation)
- wetland (is about the amount of water in the ground)


"water related" and "mountains related": completely arbitrary, you'll find
water in the mountains, springs in the mountains, vegetation in the
mountains, vegetation in and near the water, etc.

"mountains related"
- a cave doesn't have to be in the mountains
- a cliff doesn't have to be in the mountains
- a "glacier" is water related and temperature related, but it doesn't
require mountains
- a rock can't only be found in the mountains
- a volcano isn't "mountain related" (always)

I think the inconsistencies are so significant that there is no point in
trying to "correct" these, and there is no advantage in having this
arbitrary structure instead of a long list.

cheers,
Martin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20141017/887801dd/attachment.html>


More information about the Tagging mailing list