My proposal for disputed country borders

classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|

My proposal for disputed country borders

ebel
This is my suggestion for how to map disputed/claimed borders.
https://wiki.openstreetmap.org/wiki/Proposed_features/ClaimedBorders 
(but I appear to have broken the wiki).

This proposal is simple. Map the claimed border of a country according
to another country as another regular {{Tag|type|boundary}} relation,
but add {{Tag|boundary|claimed_administrative}} +
{{Tag|claimed:admin_level||2}} (since we're nearly always dealing with
countries) Add the regular tags for a boundary relation (e.g.
{{Tag|ISO3166-1}}, {{Tag|name}}).

Then add {{Tag|according_to:XX||yes/no}} for each country that does or
doesn't claims this is the border of the subject country. If
{{Tag|according_to:XX}} is missing for an object, the value should be
assumed to be "yes" if this is {{Tag|boundary|administrative}}, and "no"
if it's {{Tag|boundary|claimed_administrative}}.

== Examples ==

=== Kosovo ===

{{Wikipedia|en|Kosovo|text=no}} has been
{{Wikipedia|en|https://en.wikipedia.org/wiki/International_recognition_of_Kosovo 
  }} recognised by about half the members of the UN, since it is de
facto  acting as a country, it's mapped in OSM {{relation|2088990}}, as
{{Tag|boundary|administative}}+{{Tag|admin_level||2}}. Kosovo was part
of Serbia, which is {{relation|1741311}}, and also
{{Tag|boundary|administative}}+{{Tag|admin_level||2}}. Serbia & Spain
don't recognise Kosovo, so I presume they view the border of "Serbia" to
be the land covered by {{relation|2088990}}+{{relation|1741311}}. We can
map this by copying the Serbia relation ({{relation|1741311}}), and
changing the members to include the larger area, then add
{{Tag|type|boundary}}+{{Tag|boundary|claimed_administrative}}+{{Tag|claimed:admin_level||2}}+{{Tag|ISO3166-1||RS}}+{{Tag|according_to:RS||yes}}+{{Tag|according_to:ES||yes}}.

We can add {{Tag|according_to:XK||yes}} to the Kosovo relation, since
(IIRC) the de facto border is what the government there claims as the
border. We can add {{Tag|according_to:RS||no}} to the Serbia relation,
which means "This is the de facto border of Serbia, and they claim it's
not the border, and the UK claims it is, and Mexico claims it isn't".

=== Crimea ===

Left as an exercise for the reader.

=== Kashmir ===

(Correct me if I'm wrong) {{Wikipedia|en|Kashmir conflict}} is mostly a
dispute between India and Pakistan, but China has claims on some parts.
Neither India or Pakistan control all of what they claim. (i) The de
facto border of India, (ii) The de facto border of Pakistan (current OSM
countries), (iii) The borders of India according to India, (iv) The
borders of Pakistan accroding to India, (v) The borders of Pakistan
according to Pakistan, (vi) The borders of India according to Pakistan.

Each of these 6 options would be mapped with a separate relation.

== Advantages ==

* Copies the same logic from multipolygons, being supported by
* 100% backwards compatible with existing scheme to map countries
* Easily readable tags that data consumers can probably deduce.

== Disadvantages ==

* Creates more relations, several extra per disputed area. This could be
unwieldy an could lead to data consistancies

== Using the data ==

=== Rendering a Map ===

To render a map of the world with the Serbian view of borders, you
import the data with `osm2pgsql`, then run a SQL query like:

DELETE FROM planet_osm_polygon WHERE boundary = 'administrative' AND
'admin_level'='2' AND tags->'claimed:by:RS' = 'no',
UPDATE planet_osm_polygon SET admin_level = '2', boundary =
'administrative' WHERE boundary = 'claimed_administrative' AND
'claimed_admin_level'='2' AND tags->'claimed:by:RS' = 'yes',

or an SQL VIEW could be used.

(Or adjust your map style appropriately to look at the
{{Tag|according_to:XX}} tag, with a reasonable default).

=== Data analysis ===

With an osm2pgsql database, you can see what areas are claimed by
country X, but not de facto controlled by it.

== See also ==

* [[Proposed features/Mapping disputed boundaries]]
* [[Proposed_features/DisputedTerritories|Previous (abandoned)
proposal]] on mapping disputed territories.


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: My proposal for disputed country borders

Joseph Eisenberg
This proposal has several problems:
1) Too many new relations, up to 180 per border or whatever the number of independent states has reached.

2) OSM is for “real, current” data
- Claimed borders are not real.
- Many old claims have never been officially surrendered

3) “Don’t map your local legislation”
- legislation in country X has no jurisdiction in country Y

These last two points are in the good mapping practice wiki page and 2) is on the main page for new mappers
On Wed, Nov 28, 2018 at 6:43 AM Rory McCann <[hidden email]> wrote:
This is my suggestion for how to map disputed/claimed borders.
https://wiki.openstreetmap.org/wiki/Proposed_features/ClaimedBorders
(but I appear to have broken the wiki).

This proposal is simple. Map the claimed border of a country according
to another country as another regular {{Tag|type|boundary}} relation,
but add {{Tag|boundary|claimed_administrative}} +
{{Tag|claimed:admin_level||2}} (since we're nearly always dealing with
countries) Add the regular tags for a boundary relation (e.g.
{{Tag|ISO3166-1}}, {{Tag|name}}).

Then add {{Tag|according_to:XX||yes/no}} for each country that does or
doesn't claims this is the border of the subject country. If
{{Tag|according_to:XX}} is missing for an object, the value should be
assumed to be "yes" if this is {{Tag|boundary|administrative}}, and "no"
if it's {{Tag|boundary|claimed_administrative}}.

== Examples ==

=== Kosovo ===

{{Wikipedia|en|Kosovo|text=no}} has been
{{Wikipedia|en|https://en.wikipedia.org/wiki/International_recognition_of_Kosovo
  }} recognised by about half the members of the UN, since it is de
facto  acting as a country, it's mapped in OSM {{relation|2088990}}, as
{{Tag|boundary|administative}}+{{Tag|admin_level||2}}. Kosovo was part
of Serbia, which is {{relation|1741311}}, and also
{{Tag|boundary|administative}}+{{Tag|admin_level||2}}. Serbia & Spain
don't recognise Kosovo, so I presume they view the border of "Serbia" to
be the land covered by {{relation|2088990}}+{{relation|1741311}}. We can
map this by copying the Serbia relation ({{relation|1741311}}), and
changing the members to include the larger area, then add
{{Tag|type|boundary}}+{{Tag|boundary|claimed_administrative}}+{{Tag|claimed:admin_level||2}}+{{Tag|ISO3166-1||RS}}+{{Tag|according_to:RS||yes}}+{{Tag|according_to:ES||yes}}.

We can add {{Tag|according_to:XK||yes}} to the Kosovo relation, since
(IIRC) the de facto border is what the government there claims as the
border. We can add {{Tag|according_to:RS||no}} to the Serbia relation,
which means "This is the de facto border of Serbia, and they claim it's
not the border, and the UK claims it is, and Mexico claims it isn't".

=== Crimea ===

Left as an exercise for the reader.

=== Kashmir ===

(Correct me if I'm wrong) {{Wikipedia|en|Kashmir conflict}} is mostly a
dispute between India and Pakistan, but China has claims on some parts.
Neither India or Pakistan control all of what they claim. (i) The de
facto border of India, (ii) The de facto border of Pakistan (current OSM
countries), (iii) The borders of India according to India, (iv) The
borders of Pakistan accroding to India, (v) The borders of Pakistan
according to Pakistan, (vi) The borders of India according to Pakistan.

Each of these 6 options would be mapped with a separate relation.

== Advantages ==

* Copies the same logic from multipolygons, being supported by
* 100% backwards compatible with existing scheme to map countries
* Easily readable tags that data consumers can probably deduce.

== Disadvantages ==

* Creates more relations, several extra per disputed area. This could be
unwieldy an could lead to data consistancies

== Using the data ==

=== Rendering a Map ===

To render a map of the world with the Serbian view of borders, you
import the data with `osm2pgsql`, then run a SQL query like:

DELETE FROM planet_osm_polygon WHERE boundary = 'administrative' AND
'admin_level'='2' AND tags->'claimed:by:RS' = 'no',
UPDATE planet_osm_polygon SET admin_level = '2', boundary =
'administrative' WHERE boundary = 'claimed_administrative' AND
'claimed_admin_level'='2' AND tags->'claimed:by:RS' = 'yes',

or an SQL VIEW could be used.

(Or adjust your map style appropriately to look at the
{{Tag|according_to:XX}} tag, with a reasonable default).

=== Data analysis ===

With an osm2pgsql database, you can see what areas are claimed by
country X, but not de facto controlled by it.

== See also ==

* [[Proposed features/Mapping disputed boundaries]]
* [[Proposed_features/DisputedTerritories|Previous (abandoned)
proposal]] on mapping disputed territories.


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: My proposal for disputed country borders

Yuri Astrakhan-2
In reply to this post by ebel
Rory, thanks for tackling this. You might want to re-upload your proposal to the wiki, as it does appear to be borked at the moment.

I think we should not store undisputed territories in the same relation as the disputed ones.  Lets just store the disputed regions as individual relations, e.g. Kurill's islands, Kosovo, Crimea, etc., and each of them will have two or more "claimed_by_XX" tags.  Crimea for example would have "claimed_by_RU=AM;BE;..." (alphabetized), and "claimed_by_UE=*".  The * means everyone else who is not listed in the other "claimed_by_...". This way the list is small enough to be manageable, and we do not have hundreds of yes/no flags.

This will make it very easy to consume, by simply subtracting disputed territories from the official country's claims based on the target audience.

Example:   drawing a map of Europe for someone in France would draw Ukraine without alterations (FR is part of the "claimed_by_UE=*" wildcard), and Russia as "Russia minus Crimea" because FR is not listed in the "claimed_by_RU" tag.


On Tue, Nov 27, 2018 at 4:43 PM Rory McCann <[hidden email]> wrote:
This is my suggestion for how to map disputed/claimed borders.
https://wiki.openstreetmap.org/wiki/Proposed_features/ClaimedBorders
(but I appear to have broken the wiki).

This proposal is simple. Map the claimed border of a country according
to another country as another regular {{Tag|type|boundary}} relation,
but add {{Tag|boundary|claimed_administrative}} +
{{Tag|claimed:admin_level||2}} (since we're nearly always dealing with
countries) Add the regular tags for a boundary relation (e.g.
{{Tag|ISO3166-1}}, {{Tag|name}}).

Then add {{Tag|according_to:XX||yes/no}} for each country that does or
doesn't claims this is the border of the subject country. If
{{Tag|according_to:XX}} is missing for an object, the value should be
assumed to be "yes" if this is {{Tag|boundary|administrative}}, and "no"
if it's {{Tag|boundary|claimed_administrative}}.

== Examples ==

=== Kosovo ===

{{Wikipedia|en|Kosovo|text=no}} has been
{{Wikipedia|en|https://en.wikipedia.org/wiki/International_recognition_of_Kosovo
  }} recognised by about half the members of the UN, since it is de
facto  acting as a country, it's mapped in OSM {{relation|2088990}}, as
{{Tag|boundary|administative}}+{{Tag|admin_level||2}}. Kosovo was part
of Serbia, which is {{relation|1741311}}, and also
{{Tag|boundary|administative}}+{{Tag|admin_level||2}}. Serbia & Spain
don't recognise Kosovo, so I presume they view the border of "Serbia" to
be the land covered by {{relation|2088990}}+{{relation|1741311}}. We can
map this by copying the Serbia relation ({{relation|1741311}}), and
changing the members to include the larger area, then add
{{Tag|type|boundary}}+{{Tag|boundary|claimed_administrative}}+{{Tag|claimed:admin_level||2}}+{{Tag|ISO3166-1||RS}}+{{Tag|according_to:RS||yes}}+{{Tag|according_to:ES||yes}}.

We can add {{Tag|according_to:XK||yes}} to the Kosovo relation, since
(IIRC) the de facto border is what the government there claims as the
border. We can add {{Tag|according_to:RS||no}} to the Serbia relation,
which means "This is the de facto border of Serbia, and they claim it's
not the border, and the UK claims it is, and Mexico claims it isn't".

=== Crimea ===

Left as an exercise for the reader.

=== Kashmir ===

(Correct me if I'm wrong) {{Wikipedia|en|Kashmir conflict}} is mostly a
dispute between India and Pakistan, but China has claims on some parts.
Neither India or Pakistan control all of what they claim. (i) The de
facto border of India, (ii) The de facto border of Pakistan (current OSM
countries), (iii) The borders of India according to India, (iv) The
borders of Pakistan accroding to India, (v) The borders of Pakistan
according to Pakistan, (vi) The borders of India according to Pakistan.

Each of these 6 options would be mapped with a separate relation.

== Advantages ==

* Copies the same logic from multipolygons, being supported by
* 100% backwards compatible with existing scheme to map countries
* Easily readable tags that data consumers can probably deduce.

== Disadvantages ==

* Creates more relations, several extra per disputed area. This could be
unwieldy an could lead to data consistancies

== Using the data ==

=== Rendering a Map ===

To render a map of the world with the Serbian view of borders, you
import the data with `osm2pgsql`, then run a SQL query like:

DELETE FROM planet_osm_polygon WHERE boundary = 'administrative' AND
'admin_level'='2' AND tags->'claimed:by:RS' = 'no',
UPDATE planet_osm_polygon SET admin_level = '2', boundary =
'administrative' WHERE boundary = 'claimed_administrative' AND
'claimed_admin_level'='2' AND tags->'claimed:by:RS' = 'yes',

or an SQL VIEW could be used.

(Or adjust your map style appropriately to look at the
{{Tag|according_to:XX}} tag, with a reasonable default).

=== Data analysis ===

With an osm2pgsql database, you can see what areas are claimed by
country X, but not de facto controlled by it.

== See also ==

* [[Proposed features/Mapping disputed boundaries]]
* [[Proposed_features/DisputedTerritories|Previous (abandoned)
proposal]] on mapping disputed territories.


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: My proposal for disputed country borders

Andy Townsend
In reply to this post by Joseph Eisenberg
On 27/11/2018 23:01, Joseph Eisenberg wrote:
This proposal has several problems:
1) Too many new relations, up to 180 per border or whatever the number of independent states has reached.

It's a concern (I've made similar points about languages in the past) but in this case I don't think that there will be _that_ many new border relations.  To take Ukraine as an example, there will only I suspect be one extra relation, with a very large number of "according_to:XX" tags as almost everyone supports Ukraine's border claim, but it doesn't match the current line of control.  You might get up to a dozen in some areas (perhaps around the South China Sea, also UK with or without Chagos Archipelago, Falklands, Gibraltar etc.), but I suspect not many more than that.


2) OSM is for “real, current” data
- Claimed borders are not real.
- Many old claims have never been officially surrendered

It's true that verifiability is an issue here (the problems with some historic claims were mentioned in a previous thread) but in many cases there really isn't an argument about _where_ the border is, only _what_ the status of the thing within it has, and (taking Ukraine as an example again) I'd suggest that the statement "Ukraine claims that Crimea is part of Ukraine" is very verifiable.


3) “Don’t map your local legislation”
- legislation in country X has no jurisdiction in country Y
Where the wiki says "Don't map your local legislation" it's again just making a point about verifiability.  "legislation in country X has no jurisdiction in country Y" doesn't seem to exist in the OSM wiki at all; so I guess that you're just saying that _only_ de facto boundaries should be in OSM?  That's nearly where we are now, except that we do have a border for e.g. Western Sahara, and attempts have been made to map the claims between India, China and Pakistan (but not currently the resulting claimed country borders).

On Wed, Nov 28, 2018 at 6:43 AM Rory McCann <[hidden email]> wrote:
This is my suggestion for how to map disputed/claimed borders.
https://wiki.openstreetmap.org/wiki/Proposed_features/ClaimedBorders
(but I appear to have broken the wiki).
(text snipped)
== Examples ==


Just a thought - how about uploading some examples as a test to the dev server (or elsewhere) to allow people to experiment with the data?  It's often easier to see potential problems once you're actually trying to process the data rather than in the abstract.

Best Regards,

Andy

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: My proposal for disputed country borders

Johnparis
In reply to this post by ebel
Thanks for this, Rory. I'll add it as a comment to the active proposal (https://wiki.openstreetmap.org/wiki/Proposed_features/Mapping_disputed_boundaries).

I don't think the notion of "according_to" is viable unless it is restricted to the two disputing parties. (Three-way disputes can be simplified into three two-way disputes.)

I'll also look at your proposed tag (boundary=claimed_administrative) vs. my proposed tag (boundary_claim=administrative). Not sure whether there's any significant difference in implementation, but would like your thoughts on that.

I plan to rethink my proposal along these lines. One problem I now see with my examples, for instance, is that it provides relations for SADR (de facto and claimed) and MA (de facto and claimed), but not MA as seen by SADR. So you can't draw a global map from these relations from the viewpoint of SADR and its supporters. (You can do one from the viewpoint of MA, because SADR doesn't exist in that world, but that's a special case.)

Finally, is there some reason you want to create a competing proposal? I don't have any knowledge of two competing proposals being discussed at the same time; would they be followed by two votes? I thought the idea was to reach consensus.

John

On Tue, Nov 27, 2018 at 10:43 PM Rory McCann <[hidden email]> wrote:
This is my suggestion for how to map disputed/claimed borders.
https://wiki.openstreetmap.org/wiki/Proposed_features/ClaimedBorders
(but I appear to have broken the wiki).

This proposal is simple. Map the claimed border of a country according
to another country as another regular {{Tag|type|boundary}} relation,
but add {{Tag|boundary|claimed_administrative}} +
{{Tag|claimed:admin_level||2}} (since we're nearly always dealing with
countries) Add the regular tags for a boundary relation (e.g.
{{Tag|ISO3166-1}}, {{Tag|name}}).

Then add {{Tag|according_to:XX||yes/no}} for each country that does or
doesn't claims this is the border of the subject country. If
{{Tag|according_to:XX}} is missing for an object, the value should be
assumed to be "yes" if this is {{Tag|boundary|administrative}}, and "no"
if it's {{Tag|boundary|claimed_administrative}}.

== Examples ==

=== Kosovo ===

{{Wikipedia|en|Kosovo|text=no}} has been
{{Wikipedia|en|https://en.wikipedia.org/wiki/International_recognition_of_Kosovo
  }} recognised by about half the members of the UN, since it is de
facto  acting as a country, it's mapped in OSM {{relation|2088990}}, as
{{Tag|boundary|administative}}+{{Tag|admin_level||2}}. Kosovo was part
of Serbia, which is {{relation|1741311}}, and also
{{Tag|boundary|administative}}+{{Tag|admin_level||2}}. Serbia & Spain
don't recognise Kosovo, so I presume they view the border of "Serbia" to
be the land covered by {{relation|2088990}}+{{relation|1741311}}. We can
map this by copying the Serbia relation ({{relation|1741311}}), and
changing the members to include the larger area, then add
{{Tag|type|boundary}}+{{Tag|boundary|claimed_administrative}}+{{Tag|claimed:admin_level||2}}+{{Tag|ISO3166-1||RS}}+{{Tag|according_to:RS||yes}}+{{Tag|according_to:ES||yes}}.

We can add {{Tag|according_to:XK||yes}} to the Kosovo relation, since
(IIRC) the de facto border is what the government there claims as the
border. We can add {{Tag|according_to:RS||no}} to the Serbia relation,
which means "This is the de facto border of Serbia, and they claim it's
not the border, and the UK claims it is, and Mexico claims it isn't".

=== Crimea ===

Left as an exercise for the reader.

=== Kashmir ===

(Correct me if I'm wrong) {{Wikipedia|en|Kashmir conflict}} is mostly a
dispute between India and Pakistan, but China has claims on some parts.
Neither India or Pakistan control all of what they claim. (i) The de
facto border of India, (ii) The de facto border of Pakistan (current OSM
countries), (iii) The borders of India according to India, (iv) The
borders of Pakistan accroding to India, (v) The borders of Pakistan
according to Pakistan, (vi) The borders of India according to Pakistan.

Each of these 6 options would be mapped with a separate relation.

== Advantages ==

* Copies the same logic from multipolygons, being supported by
* 100% backwards compatible with existing scheme to map countries
* Easily readable tags that data consumers can probably deduce.

== Disadvantages ==

* Creates more relations, several extra per disputed area. This could be
unwieldy an could lead to data consistancies

== Using the data ==

=== Rendering a Map ===

To render a map of the world with the Serbian view of borders, you
import the data with `osm2pgsql`, then run a SQL query like:

DELETE FROM planet_osm_polygon WHERE boundary = 'administrative' AND
'admin_level'='2' AND tags->'claimed:by:RS' = 'no',
UPDATE planet_osm_polygon SET admin_level = '2', boundary =
'administrative' WHERE boundary = 'claimed_administrative' AND
'claimed_admin_level'='2' AND tags->'claimed:by:RS' = 'yes',

or an SQL VIEW could be used.

(Or adjust your map style appropriately to look at the
{{Tag|according_to:XX}} tag, with a reasonable default).

=== Data analysis ===

With an osm2pgsql database, you can see what areas are claimed by
country X, but not de facto controlled by it.

== See also ==

* [[Proposed features/Mapping disputed boundaries]]
* [[Proposed_features/DisputedTerritories|Previous (abandoned)
proposal]] on mapping disputed territories.


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: My proposal for disputed country borders

ebel
(Some of this is on the wiki)

On 28/11/2018 06:39, Johnparis wrote:
> I don't think the notion of "according_to" is viable unless it is
> restricted to the two disputing parties. (Three-way disputes can be
> simplified into three two-way disputes.)

This is why I like according_to:XX=yes/no. It allows for many options,
not just 2.

> I'll also look at your proposed tag (boundary=claimed_administrative)
> vs. my proposed tag (boundary_claim=administrative). Not sure whether
> there's any significant difference in implementation, but would like
> your thoughts on that.

boundary=claimed_administrative say "this is a boundary and it's of this
type".

> I plan to rethink my proposal along these lines. One problem I now see
> with my examples, for instance, is that it provides relations for SADR
> (de facto and claimed) and MA (de facto and claimed), but not MA as seen
> by SADR. So you can't draw a global map from these relations from the
> viewpoint of SADR and its supporters. (You can do one from the viewpoint
> of MA, because SADR doesn't exist in that world, but that's a special case.)

I suspect "Country X thinks country Y doesn't really exist" is very
common, surely it must happen when a region secedes. I suspect Turkish
Republic of North Cyprus doesn't exist as a country according to Cyprus
or Italy, or.... Taiwan isn't a separate country according to China.
Kosovo isn't a separate country according to Serbia. etc.
"according_to:XX=no" can be useful here.

> Finally, is there some reason you want to create a competing proposal? I
> don't have any knowledge of two competing proposals being discussed at
> the same time; would they be followed by two votes? I thought the idea
> was to reach consensus.

Yes this can appear a little snarky, that's not my intent. I half
heartely suggested this idea 2 years ago (
https://lists.openstreetmap.org/pipermail/tagging/2016-May/029211.html 
), but abandoned it.

As you know, recent events mean OSM should have an answer to this. We'll
talk and discuss and surely we can come to consensus.


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Both proposals for disputed country borders

Andy Townsend
(previous conversations heavily edited)

On 28/11/2018 10:49, Rory McCann wrote:

>
>
> On 28/11/2018 06:39, Johnparis wrote:
>
>> ... Finally, is there some reason you want to create a competing
>> proposal? I don't have any knowledge of two competing proposals being
>> discussed at the same time; would they be followed by two votes? I
>> thought the idea was to reach consensus.
>
> Yes this can appear a little snarky, that's not my intent. I half
> heartely suggested this idea 2 years ago (
> https://lists.openstreetmap.org/pipermail/tagging/2016-May/029211.html 
> ), but abandoned it.
>
> As you know, recent events mean OSM should have an answer to this.
> We'll talk and discuss and surely we can come to consensus.
>
Personally at this stage I'd suggest keeping things simple, and
https://wiki.openstreetmap.org/wiki/Proposed_features/ClaimedBorders is
certainly simpler than
https://wiki.openstreetmap.org/wiki/Proposed_features/Mapping_disputed_boundaries 
.  Some of the definition in Mapping_disputed_boundaries is useful, but
some might need clarifying (I'm thinking of the "De facto borders" and
"Default borders" definitions, which if I read correctly is just
defining some of what we currently do).  I suspect that the idea of
"OSM-designated borders" might meet some opposition, even though we
effectively have those in a few places.  The former Western Sahara is an
example; attempts to claim that Morocco has control over all of the
former Western Sahara have been reverted, but attempts to better map the
wall (which is visible from imagery) have been welcomed.

As I understand it, the broad requirements of both proposals are similar
- in a case with two claims over an area you'd get either two extra
relations (in the case of Crimea) or only one (Serbia/Kosovo).  The new
relations will be large (relations the size of a country, however you
define that, tend to be) and may be subject to breakage as current
admin_level=2 relations are (those are monitored by e.g.
https://forum.openstreetmap.org/viewtopic.php?pid=728103#p728103 ).

It'd be good to get something into the OSM data that represents these
claims so that we can provide fully worked examples to people who want
to render particular sets of borders, the same as we do for people who
want to create maps in a particular language.

Best Regards,

Andy


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Both proposals for disputed country borders

Johnparis
Thanks, Andy.

I'm currently working on refining my proposal. You're right that some of it is just to define what we already do. I've had a few private conversations with others to get feedback, as well as the feedback here and on the discussion page.

I expect to have Version 1.2 posted, with examples on the development server, this weekend.

Take good care,

John


On Sat, Dec 1, 2018 at 1:15 PM Andy Townsend <[hidden email]> wrote:
(previous conversations heavily edited)

On 28/11/2018 10:49, Rory McCann wrote:
>
>
> On 28/11/2018 06:39, Johnparis wrote:
>
>> ... Finally, is there some reason you want to create a competing
>> proposal? I don't have any knowledge of two competing proposals being
>> discussed at the same time; would they be followed by two votes? I
>> thought the idea was to reach consensus.
>
> Yes this can appear a little snarky, that's not my intent. I half
> heartely suggested this idea 2 years ago (
> https://lists.openstreetmap.org/pipermail/tagging/2016-May/029211.html
> ), but abandoned it.
>
> As you know, recent events mean OSM should have an answer to this.
> We'll talk and discuss and surely we can come to consensus.
>
Personally at this stage I'd suggest keeping things simple, and
https://wiki.openstreetmap.org/wiki/Proposed_features/ClaimedBorders is
certainly simpler than
https://wiki.openstreetmap.org/wiki/Proposed_features/Mapping_disputed_boundaries
.  Some of the definition in Mapping_disputed_boundaries is useful, but
some might need clarifying (I'm thinking of the "De facto borders" and
"Default borders" definitions, which if I read correctly is just
defining some of what we currently do).  I suspect that the idea of
"OSM-designated borders" might meet some opposition, even though we
effectively have those in a few places.  The former Western Sahara is an
example; attempts to claim that Morocco has control over all of the
former Western Sahara have been reverted, but attempts to better map the
wall (which is visible from imagery) have been welcomed.

As I understand it, the broad requirements of both proposals are similar
- in a case with two claims over an area you'd get either two extra
relations (in the case of Crimea) or only one (Serbia/Kosovo).  The new
relations will be large (relations the size of a country, however you
define that, tend to be) and may be subject to breakage as current
admin_level=2 relations are (those are monitored by e.g.
https://forum.openstreetmap.org/viewtopic.php?pid=728103#p728103 ).

It'd be good to get something into the OSM data that represents these
claims so that we can provide fully worked examples to people who want
to render particular sets of borders, the same as we do for people who
want to create maps in a particular language.

Best Regards,

Andy


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging