Henning Scholland wrote
Do you mean that splitter should generate the list only if the user gives a name (path) for it? Yes and only one list with all problematic objects.
OK, I agree that we need only one list. Should splitter try to add a name for each object so that we see something like way:101580 # Rue Docteur Bordier .... rel:47796 # Netherlands This would blow up the file and increase run time a little bit, but could be done when splitter runs the MultiTileProcessor. Another possibly interesting information would be to know why the object is a problem case. We have two very different cases: a) a polygon crosses two or more tiles b) a polygon leaves a tile, and no other tile contains the outer nodes I think this is also what Chris would like to know when we looks at the "way x is missing y nodes" messages. Henning Scholland wrote
These planet-xxx files are only meant for debugging. I think we can completely drop the creation. Or do you plan to use it for something else? No, only for debugging. Could also be removed completly.
OK, I comment the lines that produce the planet-xxx files. Gerd -- View this message in context: http://gis.19327.n5.nabble.com/splitter-r239-tp5736433p5736512.html Sent from the Mkgmap Development mailing list archive at Nabble.com.