> Lähettäjä: Dane Springmeyer wrote: > Jukka, > A couple things: > > 1) Can you file this as an issue here: > http://code.google.com/p/mapnik-utils/issues/list > That way we can spare the mapnik list from the gory details > of nik2img alpha code :) > > 2) Can you upgrade to the SVN trunk version of nik2img.py? > http://mapnik-utils.googlecode.com/svn/trunk/desktop/nik2img/n > ik2img.py > > 3) Have you tried keeping your OSM data in epsg:4326?
Hi, 1) I am not sure about what to file as an issue after Jon Burgess suggested adjusting extents in the mapfile. I feel that there is a real issue when the extents of data really are wrapping around the world like when having the whole planet file in spherical mercator. In that case the projectec bounding box gets wrong values, at least if requested in UTM 35N. I will do a set of new trials some day, at least: - by having data in 900913 but adjusting the extents to the Finnish data set to avoid wrapping - by having data in 4326 and using correct extents - by having data in 32635 and using correct extents. I will create some test maps and check if I'll get all the OSM vectors on it, if the bounds of the resulting map are the same that I used in the request and if I can overlay the maps in some GIS software with other maps. 2) That version did not change things a lot. Main issue is probably due to wrapping extents. 3) That did have an effect. While 900913 lead to vectors missing from the west, 4326 lead to vectors missing from the north :) I start to feel that nik2img is going to lead to a working system for creating projected Mapnik maps which can be georeferenced accurately with, for example, world files. Perhaps there could be even an option in nik2img.py for creating world files or MapInfo tab files automatically? -Jukka Rahkonen- _______________________________________________ Mapnik-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/mapnik-users
