Re: [Tagging] natural=ridge vs natural=arete

2014-11-05 Thread Martin Koppenhoefer
2014-11-04 22:56 GMT+01:00 Friedrich Volkmann b...@volki.at:

 This discussion comes late. Both natural=ridge and natural=arete have been
 approved by voting just 2 years ago.



arguably it is not too late, there are only 450 uses of arete by now (and
17K+ ridges). Please also note that the tag natural=arete was formally
rejected (not enough votes: http://wiki.osm.org/wiki/Proposed_features/arete
)

cheers,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


Re: [Tagging] natural=ridge vs natural=arete

2014-11-05 Thread Richard Z.
On Wed, Nov 05, 2014 at 10:01:47AM +0100, Martin Koppenhoefer wrote:
 2014-11-04 22:56 GMT+01:00 Friedrich Volkmann b...@volki.at:
 
  This discussion comes late. Both natural=ridge and natural=arete have been
  approved by voting just 2 years ago.
 
 
 
 arguably it is not too late, there are only 450 uses of arete by now (and
 17K+ ridges). Please also note that the tag natural=arete was formally
 rejected (not enough votes: http://wiki.osm.org/wiki/Proposed_features/arete
 )

after two years in the wiki where it was marked as approved and active it 
would not appear as a great idea to declare the vote for invalid based on
nitpicking formalities, how many votes were missing for approval?

Would be better to make a new proposal including details of a transition path.

Another reason I don't like current arete/ridge state is that some ridges are
very long - and they may be partially arete and ridge in different segments.
Having a way that is tagged partially as natural=ridge and partially as 
natural=arete seems like a bad idea.

Richard




___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


Re: [Tagging] natural=ridge vs natural=arete

2014-11-05 Thread Martin Koppenhoefer
2014-11-05 12:23 GMT+01:00 Richard Z. ricoz@gmail.com:

 after two years in the wiki where it was marked as approved and active it
 would not appear as a great idea to declare the vote for invalid based on
 nitpicking formalities, how many votes were missing for approval?



it was 50% missing (5 people)



Another reason I don't like current arete/ridge state is that some ridges
 are
 very long - and they may be partially arete and ridge in different
 segments.



you can split them ;-)


Having a way that is tagged partially as natural=ridge and partially as
 natural=arete seems like a bad idea.




it won't be a way, it will be several ways.

cheers,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


Re: [Tagging] natural=ridge vs natural=arete

2014-11-05 Thread Friedrich Volkmann
On 05.11.2014 10:01, Martin Koppenhoefer wrote:

 arguably it is not too late, there are only 450 uses of arete by now (and
 17K+ ridges).

450 uses are quite a lot for a feature that is constantly ignored by renderers.

For the same reason, I suppose that some of the 17K+ ridges were created by
imports.

 Please also note that the tag natural=arete was formally
 rejected (not enough votes: http://wiki.osm.org/wiki/Proposed_features/arete )

http://wiki.openstreetmap.org/wiki/Proposal_process#Voting:
A rule of thumb for enough support is 8 unanimous approval votes or 15
total votes with a majority approval

It got 10 unanimous approval votes, which is obviously even better than 8.
Therefore, it clearly was approved. And, by the way, there was more than
enough time for critics to object.

-- 
Friedrich K. Volkmann   http://www.volki.at/
Adr.: Davidgasse 76-80/14/10, 1100 Wien, Austria

___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


Re: [Tagging] natural=ridge vs natural=arete

2014-11-05 Thread Friedrich Volkmann
On 05.11.2014 12:23, Richard Z. wrote:
 Another reason I don't like current arete/ridge state is that some ridges are
 very long - and they may be partially arete and ridge in different segments.
 Having a way that is tagged partially as natural=ridge and partially as 
 natural=arete seems like a bad idea.

Subtags make it even worse, because they require to split the way a couple
of times. Most renderers label every part separately; and whenever someone
searches for the ridge name, Nominatim will deliver multiple parts instead
of the whole ridge.

Therefore, it's better to tag the name on one long way, and you may make it
natural=ridge and add some additional and nameless natural=arete where you
want renderers to display arete signatures (i.e. double sided cliffs).

-- 
Friedrich K. Volkmann   http://www.volki.at/
Adr.: Davidgasse 76-80/14/10, 1100 Wien, Austria

___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


Re: [Tagging] natural=ridge vs natural=arete

2014-11-05 Thread Mateusz Konieczny
This doesn't matter in this particular case, because natural=ridge and
natural=arete were approved at the same time.


It is about futureproof solution - new values may appear and break existing
data consumers. Adding subtags would not cause problems like this.

That's why we have a wiki with descriptions.

Is is better to have general tags that are usable without checking wikis.
Also, definition of arete is fuzzy by itself (see post by Alan Trick Most
of the things that the Wikipedia refers to as aretes are coloquially
called ridges here)

Your argumentation is based on practical issues, but essentially your
considerations are merely theoretical.

No, I am using OSM data in two projects
- map of cycleway-related data, currently in planning stages
( https://github.com/mkoniecz/bicycle_map_of_Krakow )
- contributing to Default style
( https://github.com/gravitystorm/openstreetmap-carto/ )

I can assure you that cascading tag scheme is way easier to support,
at least in my opinion.


2014-11-06 6:38 GMT+01:00 Friedrich Volkmann b...@volki.at:

 On 05.11.2014 07:19, Mateusz Konieczny wrote:
  And it is not just because with the second solution new values
  for main tag will quickly appear (see building=*).

 This doesn't matter in this particular case, because natural=ridge and
 natural=arete were approved at the same time.

  With second
  scheme there is much smaller pool of people that really understand how
  main_tag should be processed and tagged.
 
  For example I have enough general knowledge to implement support for
  natural=ridge (everybody knows what it means), but with natural=arete it
  would require at least some learning about specialist terms. Currently I
  have no idea is this tag is even correctly spelled - Wikipedia defines
 arete
  as term meaning virtue or excellence. - and ridge related article
 is
  titled arête. I also have not enough knowledge to decide whatever
  something is ridge or arête, is it clear term or something fuzzy.

 That's why we have a wiki with descriptions. When you find a description
 fuzzy or misleading, please improve it.

  Yes, I can learn about it - but I worry about the same happening for more
  things. I am NOT interested in learning about how to recognize different
  different power tower types, I want to tag just power=tower and leave
  further classification to power enthusiast that will use subkeys so
  rendering power towers on my map will be easy to implement.

 Your argumentation is based on practical issues, but essentially your
 considerations are merely theoretical.

 As a data consumer, e.g. when you want to render power towers in a 1:25000
 map, you are well advised to have a look at the subtags. Maybe there are
 power tower types which represent power towers so small and unimportant
 that
 you find it better to omit those on your 1:25000 map.

 --
 Friedrich K. Volkmann   http://www.volki.at/
 Adr.: Davidgasse 76-80/14/10, 1100 Wien, Austria

 ___
 Tagging mailing list
 Tagging@openstreetmap.org
 https://lists.openstreetmap.org/listinfo/tagging

___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


Re: [Tagging] natural=ridge vs natural=arete

2014-11-04 Thread Mateusz Konieczny
I think that natural=arete should be rather subtag of natural=ridge
(natural=ridge; ridge=arete).
It is opening way for next specialized tags - what will make using data
significantly harder.

2014-11-04 13:58 GMT+01:00 Richard Z. ricoz@gmail.com:

 Hi,

 following some discussions on github (1) and talk-at (2) I have tried
 to clarify the definition of natural=ridge in the wiki

 http://wiki.openstreetmap.org/w/index.php?title=Tag:natural%3Dridgediff=1104725oldid=998905

 Not sure if this is good enough, personaly I would prefer a single ridge
 key with additional subkeys denoting properties such as gentle,sharp, cliff
 ridges.

 1. https://github.com/gravitystorm/openstreetmap-carto/issues/1097
 2.
 https://lists.openstreetmap.org/pipermail/talk-at/2014-November/007058.html

 Richard

 ___
 Tagging mailing list
 Tagging@openstreetmap.org
 https://lists.openstreetmap.org/listinfo/tagging

___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


Re: [Tagging] natural=ridge vs natural=arete

2014-11-04 Thread Martin Koppenhoefer
2014-11-04 13:58 GMT+01:00 Richard Z. ricoz@gmail.com:

 personaly I would prefer a single ridge
 key with additional subkeys denoting properties such as gentle,sharp, cliff
 ridges.



+1
or the subkey variant Mateusz has offered.

cheers,
Martin
___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


Re: [Tagging] natural=ridge vs natural=arete

2014-11-04 Thread Friedrich Volkmann
On 04.11.2014 14:04, Mateusz Konieczny wrote:
 I think that natural=arete should be rather subtag of natural=ridge
 (natural=ridge; ridge=arete).

This discussion comes late. Both natural=ridge and natural=arete have been
approved by voting just 2 years ago. And I think that there's nothing wrong
with them. Whether to use subtags is mainly a matter of taste.

-- 
Friedrich K. Volkmann   http://www.volki.at/
Adr.: Davidgasse 76-80/14/10, 1100 Wien, Austria

___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging


Re: [Tagging] natural=ridge vs natural=arete

2014-11-04 Thread Mateusz Konieczny
Whether to use subtags is mainly a matter of taste.

No. Lets say that there is something with four main values that are
noticeable for general public and several subtypes, important for
specialists.

For data consumers interested in just four values version with subtags
is vastly easier to use cascading tagging scheme.

main_tag=first_value
main_tag=second_value
main_tag=third_value
main_tag=fourth_value

with subkeys, ignored by this general purpose renderer
first_value=a
first_value=b
(...)
fourth_value=z

Scheme with placing everything into main tag is much harder to support:

main_tag=a
main_tag=b
main_tag=c
(...)
main_tag=z

And it is not just because with the second solution new values
for main tag will quickly appear (see building=*). With second
scheme there is much smaller pool of people that really understand how
main_tag should be processed and tagged.

For example I have enough general knowledge to implement support for
natural=ridge (everybody knows what it means), but with natural=arete it
would require at least some learning about specialist terms. Currently I
have no idea is this tag is even correctly spelled - Wikipedia defines
arete
as term meaning virtue or excellence. - and ridge related article is
titled arête. I also have not enough knowledge to decide whatever
something is ridge or arête, is it clear term or something fuzzy.

Yes, I can learn about it - but I worry about the same happening for more
things. I am NOT interested in learning about how to recognize different
different power tower types, I want to tag just power=tower and leave
further classification to power enthusiast that will use subkeys so
rendering power towers on my map will be easy to implement.

On the other hand I am interested in cycling infrastructure and tag it to
absurd detail. But I am using

[amenity=bicycle_parking; bicycle_parking=stands]
[amenity=bicycle_parking; bicycle_parking=wall_loops]
not
[amenity=bicycle_stands]
[amenity=wall_loops]

-

[oneway=yes; oneway:bicycle=no]
not
[oneway=yes;bicycle:no]

And yes, I am considering usability of oneway=yes;bicycle:no
and natural=arete as similar. natural=arete in fact is even worse
as in the first case everything starting from ; sign may be
safely dropped what allows adding new details without breaking
more general rendering, and once somebody changed natural=ridge
to natural=overly_specific_value it is impossible to recover
information that it is ridge, without manually adding support to new tags.

So I will tag ridges as natural=ridge (also arêtes and other subtypes),
I will not use natural=arete as user of OSM data and I consider existence
of this tag as a really poor idea.

Maybe it is matter of taste but I have yet to see arguments why flat
tagging
is superior.

2014-11-04 22:56 GMT+01:00 Friedrich Volkmann b...@volki.at:

 On 04.11.2014 14:04, Mateusz Konieczny wrote:
  I think that natural=arete should be rather subtag of natural=ridge
  (natural=ridge; ridge=arete).

 This discussion comes late. Both natural=ridge and natural=arete have been
 approved by voting just 2 years ago. And I think that there's nothing wrong
 with them. Whether to use subtags is mainly a matter of taste.

 --
 Friedrich K. Volkmann   http://www.volki.at/
 Adr.: Davidgasse 76-80/14/10, 1100 Wien, Austria

 ___
 Tagging mailing list
 Tagging@openstreetmap.org
 https://lists.openstreetmap.org/listinfo/tagging

___
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging