On Tue, 8 Nov 2011 13:43:29 -0800 (PST), GerdP wrote:

I fear I don't understand why elevation lines are a problem for version 181

Because they are added with node IDs above the IDs used by OSM data. With the increased MAX_ID splitter r181 needs more memory than 32 bit Java can provide under Windows to process this data. Additionally, srtm2osm is creating contour data with IDs starting at 2*10^9. This is hitting the maximum of 2^31 allowed for node IDs in splitter r181.

so maybe my version will not work with that as well. What do I have to test to verify it?

Add some data with node IDs above 2*10^9 to the normal OSM data and check if the splitter needs more than about 1,5GB of RAM to process it. Or even better, use srtm2osm to create some "real" elevation data and split it together with the OSM data for the same region. Michael