[Tagging] Strange tags

[Paul Allen]

On Sun, 29 Sep 2019 at 18:25, Jan Michel <jan at mueschelsoft.de> wrote:

> On 29.09.19 17:07, Paul Allen wrote:
>


> > There are people who are VERY interested in these things.  People who
> > want to know where Munros, Donalds, Grahams, Marilyns, TuMPs, etc. are.
>
> Well... There is no documentation of these tags in the OSM wiki.
>

True.  Perhaps that means they should be documented as opposed to being
deleted upon sight.  Or perhaps something else should be done about them.

ael mentioned that these are well-known terms. I tried several
> translators and dictionaries but didn't find anything.
> A Google search only finds the wiki "List of mountains and hills
> on the British islands", but not much more related information.
>

When I tried it I got a lot of hits, but the wikipedia article seemed the
obvious
one to mention here, as it's fairly comprehensive.  I learned far more
about the
subject than I ever wanted to know from it.

These seem to be very local terms that are not used outside of Scotland
> (British Isles?).


British Isles.  With regional variations in where such terms are commonly
used.  Only
Scotland has Munros but, in the rest of the British Isles, peaks matching
the criteria of
a Munro are called Furths.

In general we oppose such local terms as keys because
> they won't be of any use outside a small area.
>

There are equivalents outside the British Isles, such as the
Eight-thousanders
(peaks over 8,000 m).

Looking at the tag history, most of these were added in a few larger
> edits in 2014 and 2015.
>

Possibly by a small number of users who were also hillbaggers.  Doesn't
really
affect whether they should be documented or deleted.  At some point we'll
have
mapped all the rivers and streams in the world, and years will go by with
nobody
mapping a river or stream - would that be grounds for deleting rivers and
streams?

What information do these terms contain exactly? What I understand from
> the Wiki page, the names are determined by three properties: The
> location (to be found from boundary relations), the height (tagged as
> 'ele') and the prominence (tagged as 'prominence').
>

If I'm reading it correctly, the location is the position of the peak.
Position, elevation
and prominence can be determined from ele=* and prominence=*, but those are
not always mapped.  The other factor, which you didn't mention, is
isolation.  However,
there's no quantitative metric of isolation, so that hits a verifiability
problem.  The nature
of the peak CAN be determined from existing tags, but only if those tags
are actually used
(and only for the types where isolation isn't a factor).   There's not much
use of
prominence=* as yet.

My very personal conclusion / opinion: These tags are undiscussed,
>

Yep.  But there are a lot of de facto tags that are undiscussed.  Sadly.

undocumented,


Yep.  But that is fixable.  If we choose to.  Even if the documentation
says "Don't use
these tags" :).


> not well-understood outside a small area


Perhaps more of a problem.  Although that could only be a problem because
of the
way they've been tagged, as opposed to the information in the tags.  See on.

and useless because they can be derived from a few documented, verifiable
> tags.
>

Only where those tags are present.  There are many peaks without an
elevation or prominence.
There are many peaks with an elevation but no prominence.

However, because I've had to do more digging to answer you, my thinking on
this has changed
a little.  These aren't really describing physical characteristics so much
as the appearance of
a particular peak on a list of peaks matched by some characteristics (not
all of which are
necessarily physical).  E.g., the Wainwrights are the 214 fells in the
English Lake District that
have a chapter in one of Alfred Wainwright's Pictorial Guides to the
Lakeland Fells.  As such,
I think these would be better handled by relations (assuming it wouldn't be
a database killer), if
they're handled at all.  They're just lists, in the same way that a cycling
route is a list of ways that
somebody decided made up a cycling route: not an inherent physical property
but an aggregation.
Except, of course, cycling routes often have signs.

-- 
Paul
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20190929/a3d18783/attachment.html>