On Wednesday 20 September 2017, Yuri Astrakhan wrote:
> Christoph, a valid point. Yet the duplicate would allow finding many
> of these errors, rather than leaving wp-only to go bad due to
> changing nature of the WP articles.

Actually no - you can find the errors just as well without adding the 
wikidata tags to OSM as after doing so.

The perpetuation of errors is one of the primary reasons why mechanical 
edits are often not considered favorably in OSM.

> As for sameness argument - lets 
> try to work on them on a case-by-case basis.

I don't really want to argue this here - as said i have no objections 
against having wikipedia/wikidata tags as references for 'related 
features' but many treat these references under the assumption that 
they indicate identical real world concepts exist on both ends - or 
even worth: they might think a lack of identity is an indication for a 
factual error in the data on one side.  I wanted to point out that this 
is a fundamentally incorrect assumption.

Don't assume such cases are just a freak anomaly - they are not.  OSM 
and wikidata are two very different projects which developed in very 
different contexts.  Just another example: For most cities and larger 
towns (at least in Germany) there exists an admin_level 6/8 unit with 
the same name and most of these seem to have a single wikidata item 
while in OSM we have two separate concepts for the populated place 
(place=city/town) and the administrative unit (boundary relation with 
boundary=administrative + admin_level=6/8).

Christoph Hormann

talk mailing list

Reply via email to