Large-scale mass import of place nodes

Namaste all!

Recently, while contributing data following up my latest trekking trip in Dolpa, I noticed a large-scale mass import of place nodes carried out in numerous change sets by user WojtekK [1]. Me and another contributor have discussed a change set [2] but we get no response yet from WojtekK.

The resource obviously is a GIS data set provided by the Office for the Coordination of Humanitarian Affairs OCHA Nepal[3].

AFAIK this import was

(a) not discussed before with the community nor is it documented anywhere.

(b) The license of the data seems to be not compatible with the OpenStreetMap Open Database License, as the terms of use [4] require „Compilation or creation of derivative works from these Materials is subject to the inclusion of credit to the source of the Materials used (which may or may not be OCHA)“.

(c) The import introduces countless place node duplicates. Furthermore the spatial quality of the place nodes is poor. Positional errors of hundreds of meters are not uncommon. The reason for this is obvious: The node positions has been derived from labels on 20+ years old printed topographic maps published by the Survey Department of Nepal.

(d) The import use tags derived from metadata of the source shape file which are mainly undocumented or pointless, see i.e. https://www.openstreetmap.org/node/5669411272

As this import significantly does not comply with the guidelines [5], I have asked the data working group to execute an immediate revert of the corresponding change sets.

Best regards
geow

[1] https://www.openstreetmap.org/user/WojtekK/history

[2] https://www.openstreetmap.org/changeset/59598339

[3] https://data.humdata.org/dataset/settlements-in-nepal

[4] https://www.humanitarianresponse.info/en/applications/data/page/terms-use

[5] https://wiki.openstreetmap.org/wiki/Import/Guidelines

This was disputed with Nepali community and described in OSM Wiki page Project Ilam.
Please take a look and ask Kathmandu Living Labs stuff before starting revert.
Due to the fact, that nepali gov. offices are using OSM data I don’t think this actions would be helpful to anybody. Positions of Villages are manually modified and naming corrected according to the nepali grammar.
Take a look in discussions in this forum.
Thanks.

Hi,

Geow, as marek kleciak mentioned there had been a discussion with the OSM Nepal community before the start of the OSM Ilam project. The work done has been documented in the wiki. However, I do not recall any large scale importations during this project maybe marek can clarify.

Regarding the OCHA data I am not clear if you are refering to data uploaded right after the Earthquake as OCHA was fully aware of licensing policies of OSM before making any data available for upload to OSM.

I agree that duplicate node and positional accuracy is a big problem but it might be better to try and create an OSM Nepal effort to clean this data rather than remove them completely. Please give us time to further understand the dataset. I will also put this forward in the OSM Nepal page in FB.

The revert has been implemented last night by the DWG and the user WojtekK has received a 0-hour-block. Remakably he had been blocked 5 years ago due to an undiscussed import: https://www.openstreetmap.org/user_blocks/2038

There is no documentation nor reference to any import procedure of place nodes at the Ilam site https://wiki.openstreetmap.org/wiki/Project_Ilam_specification

I am not aware of any recent discussion of this specific import, executed two weeks ago. You might provide a link?

I’m sure that the import was well-intentioned, but it was definitely ill-considered and violates the import guidelines substantially. A database cluttered with misplaced and duplicated place nodes is not helpful to anybody.

Undoubtedly this import needs to be reconsidered and improved prior to performing a reload.

Best regards
geow

I forwarded this discussion to the KLL because of my summer holidays. KLL should fix it with DWG.
Best regards!
Marek

I have informed user geow that I ended my all OSM activities because of him and his removing of data in Nepal. I delete my OSM account and wish to say my apologise to all Nepali. Your country is one of the most beautiful in the world. Be proud to be Nepali!

Hmm, the undiscussed import was removed by the OSM Data Working Group because it does not comply with the guidelines. I’m sure that the import of user WojtekK was well-meaning, but it was definitely ill-considered and poorly crafted, i.e. it introduces countless place node duplicates with poor positional accuracy, see #1. AFAIK neither WojtekK nor KLL have made efforts to solve the problems before attempting another import.

That’s a shame. :roll_eyes: I would be very happy if you would find your way back to constructive work for OSM. :slight_smile:

+1 I had the privilege to spend some months in Nepal on trekking tours, so most of* my contributions for OSM are based on ground knowledge.* Maybe that’s why I don’t like bad imports. :wink:

Best regards
geow