Remove deprecation warnings for is_in tag in the DRC

[clairedelune]

The deprecation warnings associated to the "is_in" related tags for more than a year now lead some remote mappers to remove valuable information and created a lot of confusion among mappers in the Democratic Republic of the Congo since this information is still actively collected on the ground. If it is still in wide use in the country it is mainly because of the absence of sufficiently accurate, official and usable sources of administrative limits despite existing efforts of the OSM and GIS communities. I'm therefore requesting for a country exception to be applied to the current validator rule.

[stoecker]

Wouldn't it be better to add lower accuracy boundary information and step by step increase the accuracy instead of reviving a method which has been obsoleted over a decade ago?

[clairedelune]

In the DRC, admin_levels 7 and 8 are not always continuous polygons (independently from water surfaces, but rather to population displacements from traditional chiefdoms). In several parts of the country these administrative entities are made of multiple disconnected areas and some of these limits are linked to on-going deadly conflicts. In such cases, displaying lower accuracy limits doesn't sound like an adequate solution. Moreover for admin_level 8, the only publicly available dataset still represents entities as individual points, from which polygons cannot yet be extrapolated in a satisfying way.
For health limits that are more continuous and less prone to serious conflicts, we are currently in the process of adding lower accuracy boundary information where available to be improved progressively. However despite the participation of numerous stakeholders for over a decade, the national coverage of the health area limits (health_level=8) is yet to be achieved.
I don't think this tag should be maintained forever but at this point, it is still answering some of the current needs in country.