Hi Stefano, the problem with your file cardoso.osm is that the multipolygon is incomplete. With this data mkgmap tries to guess the shape of the mp and that fails in this case. I don't know how exactly you created the file, when I use JOSM to download an area from Overpass api the multipolygon is complete. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Stefano Nannetti <stefano.nannetti@gmail.com> Gesendet: Mittwoch, 24. April 2019 14:49 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: [mkgmap-dev] Strange behaviour in multipolygon Processing a small area in my region, I noticed a strange behaviour in the rendering of one multipolygon. In the original OSM map (see image A.jpg, bounds are 43.9978 to 44.0135, 10.2989 to 10.3228) the area along the street to "Cardoso" is a strip of empty land (untagged) sorrounded by a forest. This forest is a large and complex multipolygon tagged with natural=wood. Thus, the strip of empty land is defined by lines which are the outer members of the multipolygon (see file cardoso.osm). When processed with mkgmap, the result is somehow "inverted": the map shows the strip of land as wood and the area around as empty, both in Basecamp and in my eTrex 20x (see image B.jpg). To test the issue I created a simple style (see files polygons and points) containing only the file "version", the file "points" which defines only the rendering of places and the file "polygons" which defines only the rendering of the areas tagged with natural=wood. I cannot figure out if the problem lies in mkgmap or in the incorrect mapping of the area in OSM. By the way, the area is correctly shown in many online maps deriving from OSM data and also in the OpenMTBmap which I downloaded and installed in Basecamp. I'm definitely stuck. Any suggestion on how to solve the problem?