[OSM-talk] OSMF makes a political decision where should be a technical solution?
victor.shcherb at osmand.net
Thu Nov 22 18:53:34 UTC 2018
I followed many topics for the last 3 days about the Ukraine/Crimea and I
would like to propose another look to all known issues.
First of all, I would like to thank everybody for staying calm and
considering all views for this *non-trivial* problem. Originally, I'm from
Belarus (former USSR) and currently I live for a long time in the
Netherlands, so I hope I can express my point of view objective but also
explaining the gist of the issue. Even though I'm leading OsmAnd mobile
development, I will speak solely from my perspective on this question.
*STATEMENT 1. *There is no ToTG principle for artificial objects such as
boundaries + Boundaries are always driven by one or another authorities.
*1.1 *To clarify that principle we can go the lowest level, city or suburb
boundaries. it is very clear that it is impossible to identify on the
ground whether city-suburb has ended and another has started.
*1.2 *Of course, we can clarify or build it that knowledge from milestone,
flags or fence. Though we have different Mapping for fence, milestone and *we
shouldn't mix it with country borders.* Following that principle, it will
be hard to build polygons cause we always could miss data in between or it
could be very incomplete from the Ground knowledge
*1.3 *Most of the data today is coming from authorities. Local
municipalities provide that public data, so admin_level of lower level
*1.4 *There is no ToTG to identify a country, unless you go and do a voting
on the special piece of terriritory. I think you can find lots of places
like Kurdistan where people would say that they belong to country that
doesn't exist or not listed in UN. Country is an entity that coexist with
other countries and other countries should acknowledge it and acknowledge
their borders (especially for neighbor countries).
*1.5 *Fence or any physically present object which could be verified by
ToTG doesn't make border legitimate and will be very likely removed from
admin_level relations doesn't matter if it looks or claimed by
non-authorized entity a special territory if it contradicts to 99.9%
perception of other mappers.
With this point I'm trying to say, that hiding a solution behind ToTG
principle we are raising even more questions than we had.
*STATEMENT** 2. *There is no right decision unless we clarify what exactly
data and how it should be organized.
*2.1 *Moving objects or making special statements about concrete objects
will drive to a mess. It is obvious that we better focus on Proposal and
clarify how to deal with data rather than changing the map itself.
*2.2 *Every mapper should be able to make the right decision himself
following the wiki documentation. If it is not possible than the
documentation or rules are not complete or not correct. We should not block
people that do mistakes in understanding whether their logic correct or not
especially if it is a significant number of people.
*STATEMENT** 3. *Any decision about current Relations in OSM will be
political and it will only evolve confrontation between local editing
groups. I believe OSMF should not take any political decisions in such
I truly believe that DWG/ OSMF didn't want to make any political decision
and only correct the data and make it consistent which makes perfect sense
from top level view unless you don't bother with the real situation itself.
Unfortunately it is not possible to solve political question and don't get
dragged into politics.
*3.1 *Hidden political decisions are bad. Why this decision is political?
First of all, there is a small politics involved anyway DWG is elected by
OSMF members and DWG made that decision. In case people are against it,
they can vote for different DWG group and next year the situation around it
will be changed again and again. The problem remains here anyway, what if
Ukrainian community to vote will be smaller than Russian or what if votes
can be based on various connections between people, so we'll get into a
*3.2 *THE WORST PART: We evolve confrontation between 2 big group of
mappers which were always welcome in OSM but as of today they read OSM
rules differently and the war of edits begins. In case we want to keep both
group of mappers because they INDIVIDUALLY support different regions, we
need to find a solution for both of parties.
In that case I would actually support idea of deleting all country
boundaries to avoid this question completely.
*3.3 *DWG/OSMF should minimize any political impact with all possible
technical solutions. I truly believe that this question could be solved
with proper tagging mechanism and even more it could be solved on the level
Google Maps solves it, by having different maps for locales UK_en.
Everything should be considered and estimated in order to minimize
reputation and diplomatic damage.
*STATEMENT** 4. *Decisions should be focused on providing the most value to
A) users/editors B) Applications -> to users in the end.
*4.1 *openstreetmap.org website solves at most 5-10% use cases where OSM
data There is not much critical that some features are not supported on it
and it is first of all made to be useful for mappers where data through
applications could reach much bigger audience and make mappers happier to
impact on the world by small changes.
*4.2 *Any decision about Ukraine osm-relation doesn't satisfy End-User
application. For example, in OsmAnd we would like to display different maps
for different regions and as of today we are forking country relations and
doing lots of manual hard fork and all that is coming anyway from OSM but
there is no way to contribute it back. We don't have clear algorithm to
represent correct boundaries from Chinese Government perspective and in the
end we don't represent any boundary correct at all which makes that data
almost useless for End-User application.
*4.3 *OSM should focus on providing as much and as relevant data as
possible and solution will come naturally. If we focus to much what we can
display on openstreetmap.org website we will bound ourselves of what
feasible/not-feasible to implement on it where it is not necessary. Give
right amount of data to the application and give an indicative algorithm
for most typical use-cases and problems and it will be sufficient.
For a sake of edit-wars some features could be disabled on the main
website, so the data question will be leading.
*4.4 *Making political risky decisions on the data side will transfer the
legitimate problems to End-User applications which they wouldn't know how
to solve correctly, basically it will have an amplifier effect.
*STATEMENT 5. *Making a decision about Ukraine creates a very nasty
precedent which should lead to other changes and raise another questions.
*5.1 *Ukraine government doesn't control EAST part of Ukraine (even though
there are some relations with citizens who live on that part). *So in that
case Ukraine border should change again.*
*5.2 *South Cyprus doesn't control North Cyprus
also contain statement about UN recognition which could be applied to
Ukraine and then Ukraine should contain Crimean and Russia shouldn't
*5.3* Admin levels for China disputed territories are not consistent.
Somehow Country level *contains less than all administrative child
relations under it *(https://www.openstreetmap.org/relation/4052315 goes
beyound country level)
Other questions as well. So, it means DWG is not creating a rule but rather
changing 1 local object which should be done by local community (in that
case it should be ukranian community as Ukraine refers to it).
I don't want to enumerate possible solutions cause I think if we spend
enough time on it and invite people from different communities, we could
find it easily.
*P.S.* I hope we do not take that question as something that will die out
in a week or in a month. Disputed territories were a painful part for OSM
for a long time and the pain is only getting bigger. I totally support
decision OSMF stay away from politics but it looks like if you don't get
involve enough, then politics will start involving you in a different
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the talk