On Jun 4, 2010, at 8:32 AM, Artem Pavlenko wrote: > On 4 June 2010 16:05, Michal Migurski <[email protected]> wrote: > On Jun 2, 2010, at 7:32 AM, Dane Springmeyer wrote: > > > My first guess would be that Mapnik is unable to reproject the osm data > > into UTM because the osm layers extents are not valid for UTM. > > Out of curiosity, what does it mean that the OSM layers are invalid for UTM? > > I think Dane meant that you can't re-project the whole world using UTM. > Transverse Mercator projections are only valid for a particular region > (zone). If there's very long way spanning multiple UTM zones, re-projecting > it wont produce anything sensible. I'm still not sure what's the best way to > handle this in Mapnik, perhaps we can clip to the valid region (it's not > going to be rectangular shape. though) before sending coordinates to proj4 or > use something more robust than proj4. Any other ideas most welcome!
Right, makes sense. It's a bit weird since I think a way needs to be very far out of a zone to be invalid, but at low zoom levels this might come up. Clipping seems like it'd be the right thing if done post-projection, since I know the UTM grid can get pretty far out of square in higher and lower latitudes. -mike. ---------------------------------------------------------------- michal migurski- [email protected] 415.558.1610 _______________________________________________ Mapnik-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/mapnik-users
