There are at least two sets of GNIS data that were imported separately. I know of at least the populated places (original import scripts here: http://beej.us/osm/cityosm/, no wiki that I know of.) and the POI type features (I don't know if and where the original import information is preserved for those). That may be the source of the confusion regarding the unique identifier.
I support the notion of preserving lineage through a unique identifier *as long as that identifier is still meaningful*. (TIGER for example no longer maintains the same identifiers imported with the 2005 data). Also, a look at the GNIS metadata page may help resolve some of the discussion topics in this thread: http://geonames.usgs.gov/domestic/metadata.htm On Thu, Aug 22, 2013 at 1:02 AM, Bryce Nesbitt <[email protected]> wrote: > > > On Wed, Aug 21, 2013 at 11:59 PM, Bryce Nesbitt <[email protected]>wrote: > >> >> gnis:feature_id >> gnis:id >> GNIS_ID >> >> and, in fact, several others less widely used. >> > > tag gnis appears to have feature ID's also. > I'd be in favor of deleting EVERYTHING in the gnis: space (except id, ele, > fcode). Consolidate any 6 or 7 digit ID into gnis:feature_id. > > _______________________________________________ > Imports mailing list > [email protected] > http://lists.openstreetmap.org/listinfo/imports > > -- Martijn van Exel http://oegeo.wordpress.com/ http://openstreetmap.us/
_______________________________________________ Imports mailing list [email protected] http://lists.openstreetmap.org/listinfo/imports
