[OHM] Advice on modelling historical boundaries & where to keep them?

[Jo Walsh]

A case recently came up for the DWG in which an effort to model
historical administrative boundaries is interfering with geocoding
results from Nominatim, and data deletion is resulting. The problem
seems in part to lie in the use of admin_level to tag historical
boundaries, without a clearly understood means of parsing date ranges
for boundary validity. I see some wiki notes on historic boundaries
here, but they seem partial and outdated in nature:

http://wiki.openstreetmap.org/wiki/Tag:boundary%3Dhistoric

It's not purely a research question as there can be decent business
value in historical boundary data, cf the self-supporting Vision of
Britain Through Time", whose underlying data is unfortunately caught up
in the issue of Ordnance Survey derived works, in part.
http://www.visionofbritain.org.uk/ and that project did some
investigatory work in Sweden and Estonia, I believe; I've lost touch
with Humphrey Southall, the founder, since i moved on from research
support work. 

Anyway, it seems to me as if OHM would be the perfect place to redirect
people with a burning desire to map historical abstractions, in the same
way that opengeofiction.net is an ideal outlet for people who must map
fictional places which don't belong in the main OSM dataset. Does this
make sense and is OHM as it stands a stable and long term resource for
such mappers?

I would appreciate any advice or opinions on the topic of historic
boundaries in OSM from those who have been thinking about it for a lot
longer than me,


- Jo
-- 
  Jo Walsh
  metazool at fastmail.net