[Imports] Planning to import Indonesia's village boundaries
osm at imagico.de
Sun Apr 16 14:42:19 UTC 2017
On Sunday 16 April 2017, Ivan Garcia wrote:
> An extract of the original SHP files(for Riau archipielago province
> for example) can be found here:
> https://github.com/capiscuas/osm-playground , also it contains the
> converted .osm file that would be imported to OSM after comparing
> with the current data in place already.
We don't just need an example, we need the complete data to be able to
assess its quality.
What can be said from your example area is that at least in parts the
data is pretty inaccurate. The admin units are apparently defined as
ending at the coastline but their position does not match the coastline
very well either in OSM or in reality, examples:
Especially for small island areas like here where the vast majority of
the geometry data is outer costline boundaries it would be much more
valuable for OSM to invest the effort into generating accurate
geometries and only use the admin unit name and internal boundaries
from the external data set.
Experiences from elsewhere indicate that when importing inaccurate
administrative boundaries the mismatch between the boundary geometries
and the defining observable features like the coastline tend to be
eternalized. Mappers do not usually synchronize geometries later when
mapping accuracy is improved.
> UNICEF and the owner of the data:BPS statistics agency share a
> COUNTRY PROGRAMME ACTION PLAN(CPAP). The email answer from BPS giving
> explicit permision to UNICEF to do the necesary changes in OSM using
> their data should be enough. There are precedents in OSM imports with
> explicit permission through emails being sufficient. Correct if I am
> wrong and please provide what sort of document/signature/email
> response shall we ask BPS to make it official then. Thanks.
You can find some helpful information on:
The basic requirement is that the data needs to be either
* published under a compatible license independent of use in OSM
* have explicit permission granted by the owners of the data to the OSMF
to use and distribute the data like for example in the template from
the above page.
> > * you do not in any way discuss what administrative boundary data
> > exists in OSM at the moment for Indonesia, from what sources this
> > is and how you intend to combine this with the data you plan to
> > import.
> The current boundaries of Indonesia can be displayed in real time
> here: http://overpass-turbo.eu/s/ooe
I know how i can find out what boundary data exists in OSM for
Indonesia. What i want to know from you is how you assess this data
and how you plan to combine the data to be imported with the existing
data. You cannot import admin_level 5-7 while ignoring admin_level 4
and lower (and despite that there is also existing data for the higher
levels in some parts).
And you should not assume you will handle this on a per case basis as
you move forward with the import - Indonesia is a big country and you
need to have a solid and robust plan before engaging in such a fairly
> > * you write you were asked to do the import but you don't write who
> > asked you to do this and what exactly the task is you were asked to
> > perform (like for example if you were instructed to get certain
> > data into OSM or to make sure administrative boundaries for
> > Indonesia are mapped in OSM).
> Yantisa (which is in CC) from HOT Indonesia contacted me to help them
> converting the SHP files into .osm and import those missing parts
> into the OSM server.
So you will be performing not only the data preparation but also the
actual import work?
More information about the Imports