Peter Körner wrote:

Theoretically, my meridian gap could be caused by osmosis, by osm2pgsq,
or by any of their underlying libraries. My gut feeling though is that
it's caused by a rounding error in osmosis or by an accuracy error in
its underlying routines.

Can't you check this by looking at the resulting .osm files?

For the Manaus example, the .osm files say

<?xml version='1.0' encoding='UTF-8'?>
<osm version="0.6" generator="Osmosis 0.36">
  <bound box="-90.00000,-77.00000,0.00000,-76.00000" 
origin="http://www.openstreetmap.org/api/0.6"/>

and

 <bound box="-90.00000,-78.00000,0.00000,-77.00000" 
origin="http://www.openstreetmap.org/api/0.6"/>

respectively.

However,

grep 'lon="-77' S-77--76.osm
  <node id="249411355" version="1" timestamp="2008-02-26T19:17:30Z" uid="308" user="MichaelCollinson" 
changeset="209621" lat="-1.333333" lon="-77">
  <node id="249411565" version="1" timestamp="2008-02-26T19:17:40Z" uid="308" user="MichaelCollinson" 
changeset="209621" lat="-2" lon="-77">
  <node id="249420393" version="4" timestamp="2010-05-28T09:38:41Z" uid="77114" user="LLAQWA" 
changeset="4829133" lat="-1.916667" lon="-77">
  <node id="249598742" version="1" timestamp="2008-02-28T19:43:08Z" uid="308" user="MichaelCollinson" 
changeset="223447" lat="-0.333333" lon="-77">
  <node id="414231264" version="1" timestamp="2009-06-01T21:09:15Z" uid="113186" user="Peru ruteable" 
changeset="1398408" lat="-12.12796" lon="-77"/>

grep 'lon="-77"' S-78--77.osm returns nothing.

Hence, it's not osm2pgsql that ate up the missing parts.

Still, this is now weirder than before. If osmosis is selecting
everything inside the bounding box including the box itself, there
should be thousands of nodes here. And if osmosis is selecting
everything inside the bounding box but not the box itself, there
should be nothing. Those five nodes above don't fit either theory.

I put the .osm files at www.xs4all.nl/~oracle/S-77--76.osm.bz2
and www.xs4all.nl/~oracle/S-78--77.osm.bz2 anyway. They were
extracted from the 100825 planet file.

Z




_______________________________________________
osmosis-dev mailing list
[email protected]
http://lists.openstreetmap.org/listinfo/osmosis-dev

Reply via email to