[Imports] Importing Arkansas data

[Mike N]

On 3/28/2011 3:29 PM, Josh Doe wrote:
> Hmm, also is it just me or does it seem like everything that he's
> imported consists of nodes without any tags? I looked through a few of
> the changesets and didn't see any ways,

   This part is OK - in terms of an upload, nodes must be present to 
attach ways to.  In a large upload, nodes will be placed first.

   The problem has already been noted here; empty nodes will be visible 
for a period, tempting them to be deleted.


 > and most disturbingly no tags at
 > all on any of the nodes.

   Having no tags on the nodes is actually good - as we learned in the 
TIGER dataset; the individual nodes don't need to be tagged to learn the 
source.  Only the changeset itself need be tagged to provide 
attribution, or possibly just the ways.