On 08/31/2011 06:03 AM, Martin Koppenhoefer wrote:
2011/8/31 Greg Troxel<[email protected]>:
I thought the issue was that there are two distinct concepts:
boundaries, where there is some legal distinction and a precise edge
place names, which have more or less indistinct boundaries.
just because they have no legal status does not mean there aren't
distinct limits. Usually / often there are. There can be natural
limits (cliffs, rivers, lakes, woods, ...) and man_made limits (big
streets and motorways, railways, ...), the limits might also be soft
or social / cultural / ethnic / economic / structural / typological /
historic, ...
It is very typical is to have/just one clear edge/ (e.g. shantytown is
definitely "west of the railroad tracks"). The rest is a matter of
perspective and interpretation, which is not a very firm foundation for
mapping.
That leaves:
1. registered administrative boundaries (as legally defined: often
co-incident with an existing boundary (e.g. the City and County
and State borders are all the same)).
2. non-administrative places as nodes (with indistinct or at least
undefined edges)
3. non-administrative places as areas (which I think is more rare
than this discussion has been suggesting).
The main goal should be so the name search box gets you to the right area?
We want to capture all the alternative names, so that when someone
searches with OSM they get a better result than commercial maps?
_______________________________________________
Tagging mailing list
[email protected]
http://lists.openstreetmap.org/listinfo/tagging
