[Talk-us] Possible method for identifying major US cities

Nathan Edgars II neroute2 at gmail.com
Sun Oct 24 22:46:30 BST 2010


On Sun, Oct 24, 2010 at 5:39 PM, Anthony <osm at inbox.org> wrote:
> In my experience, the Tiger CDPs in Florida are really bad (for
> instance, the CDP that my house is listed as being in is completely
> wrong).  I'd imagine the Tiger CDPs of principal cities are probably
> somewhat better, but I'd want a little bit more information before
> saying that marking Tiger CDPs as cities is a good thing.  Personally,
> I'd rather we just deleted them (at least in Florida, and without them
> being hand-checked by someone familiar with the area).

I hate the CDP boundaries too, and have been slowly phasing them out
in favor of landuse polygons in the Orlando area. I've already marked
them all as boundary=census (meaning that, as a pleasant side effect,
they don't render).

But most of the CDPs also have place nodes from GNIS. (The one that is
a 'principal city' but doesn't, The Villages, has a landuse polygon:
http://www.openstreetmap.org/?lat=28.9101&lon=-81.9802&zoom=14&layers=M)
These would remain even if the boundaries are deleted in the future.



More information about the Talk-us mailing list