[Talk-us] Import for Murray County, OK

[Mike N.]

This might get me strung up on the yardarm, but I propose
  1.)  Mapping in a place I've never been to
  2.)  Doing an import for a single county from the county's GIS data

Background: Murray County's (Oklahoma) roads were recently renamed for E911 
purposes.   This data is as-of 11/29/2010 - newer than anything likely to be 
in TIGER 2010.

The data was obtained from the county's shapefiles and processed to be OSM 
compatible. Peruse the proposed dataset in JOSM at:
http://greenvilleopenmap.info/roads5.zip
  **** Please Do not try to upload - the upload will fail in the current 
data state, and result in a messy possible partial upload ****

For a quick summary, here is an overlay of the 2 datasets, with some of the 
proposed data selected in red to stand out against the street in the aerial 
imagery.  http://greenvilleopenmap.info/StreetCompare.jpg

 The aerial imagery for this area is likely to be within 2M of actual, as 
confirmed by downloading GPS traces taken on I35.   The county GIS data 
centerline accuracy appears to be very good and much better than the 
original TIGER data.

   o  Data is public domain, per the contact in the county's GIS department
   o  An analysis shows that the only edits in the county have been to the 
Interstate (Relation / Dual Carriageway / Align/ Exits / ref) and US 
Highways (relations / refs).

Proposed work method:

 1.  Delete all TIGER roads in Murray county except 'motorway', 
'motorway_link'.
 2.  Stitch in new dataset into county lines and interstate system.
 3.  Re-add US Highway relations

Notes:
  1.  A Wiki page will be created for this import.    The data itself will 
have no source= attribute, just as the dataset roads5.osm currently exists. 
The attribution will be on the changeset itself with the source= pointing to 
the OSM wiki import page entry.
  2.   There is no attempt to maintain a link back to the original county 
GIS shapefile dataset elements - future edits will be smaller in scope, and 
will likely come from TIGER or local mappers via aerial imagery or GPS 
traces.
  3.   The address data from the county's dataset will not be imported 
because I have no way to verify its accuracy level.   It would probably be 
better than TIGER's anonymized address interpolations ranges, but would only 
be address ranges on a street, and may not correspond 100% to physical 
address locations.

 Questions -
   1.  The 'imports catalogue' 
http://wiki.openstreetmap.org/wiki/Import/Catalogue is ugly - should I just 
slap this import page somewhere under one-time imports, or just add it to 
'potential datasources' 
http://wiki.openstreetmap.org/wiki/Potential_Datasources?
   2.   The ever intriguing question of "is public domain good enough for 
OSM"?  Or, do one or more lawyers need to get involved to bless it first?
   3.   The 'Chickasaw Turnpike' toll road OSM classification is a question 
mark - I left it as secondary because it is not grade-separated carriageway, 
but I believe it qualifies as trunk.   Any other road reclassifications for 
this area?

  Thanks,

  Mike