[Talk-us] [OSM-talk] tiger polygons

[Paul Fox]

dave wrote:
 > On Tue, 2007-12-18 at 20:50 -0500, Paul Fox wrote:
 > > 
 > > i'll take a look at your code -- i've just finished the code in
 > > RoadMap to (finally -- they've been left incomplete for years)
 > > properly display fully-shaped polygons from TIGER data.  so it's
 > > somewhat on my mind, and now that RoadMap can display OSM data, i
 > > have a vested interest on that side too.
 > > 
 > > (but no promises.  :-) 
 > 
 > Do you know if the TIGER polygons share tlids with the roads?  We might
 > want to re-use some of the existing nodes that got uploaded for the
 > roads if they do.

yes, i believe so -- not only are they in the same namespace, but
the same tlid may be needed both as a line and as part of a
polygon (i.e., they're not distinct subsets of "tlid space").  at
least, i know there's code in RoadMap to fall back to a polygon's
notion of a line's layer even if the line might not be required
based on the line's own layer.  (some of those terms are probably
roadmap-specific.)

 > 
 > I still have the original .osm files and all of the things to convert
 > from tlids to the negative ids in those osm files.  I also have the
 > db-files that convert those negative ids into the in-database ids that
 > we see on the server today.  Whew!!  Enough conversion for you??!?!? :)

too much, maybe. :-)

i should qualify my offer to help with the disclaimer that i know
exactly zilch about osm internals.  all that i know, i learned from
reading xml (and binary mobile) output.

what's interesting about "negative ids"?

in what format does the data need to be in, for import?  (i should
unpack your tarball...)

paul
=---------------------
 paul fox, pgf at foxharp.boston.ma.us (arlington, ma, where it's 21.0 degrees)