[OSM-dev] Mapnik dist-upgrade broke my rendering

Jon Burgess jburgess777 at googlemail.com
Sat Apr 12 22:24:33 BST 2008

On Sat, 2008-04-12 at 22:44 +0200, Frederik Ramm wrote:
> Hi,
> > The mapnik SVN tree has been broken for OSM rendering for a little
> > while. Tom just committed a fix today which resolves this problem[1].
> Well it made my problem a little bit different - The numbers drawn
> along the street axis are now integers and not floats it seems:
> http://www.remote.org/frederik/tmp/broken.png
> But broken it is still. 

That looks like what I got prior to Tom's fix. Are you sure you are
running the latest SVN version of Mapnik?

> Could anybody with a current, working Mapnik setup try the following
> Python script for me and tell me if they get the same as above, or
> something different? (I use the osm.xml map file from the OSM SVN
> tree, modified only the db access parameters, and I have a plain
> postgresql into which I imported data using osm2pgsql without extra
> parametes.)
> If it works for everyone else then there must be a bug in my setup...
> if others also get the strange result then the problem must be
> somewhere in the map file or data structure... Mapnik itself does
> work, at least the rundemo.py does the right thing.

I ran your script against the main Tile Postgres DB and got the result: 


The only change I made was to adjust the projection line to match the
"osm2pgsql -m" projection

prj = Projection("+proj=merc +a=6378137 +b=6378137 +lat_ts=0.0
+lon_0=0.0 +x_0=0.0 +y_0=0 +k=1.0 +units=m +nadgrids=@null +no_defs

Either you've still got a problem with Mapnik or there is some bad
interaction with the latest osm2pgsql code. The main tile DB was
imported using a version built a couple of weeks ago which does not have
Kleptog's latest changes.


More information about the dev mailing list