mkgmap and relation 406175
Can someone take a look at relation 406175 (Skunk River and some related waterways in central Iowa)? With my most recent attempt at rendering with mkgmap (rev 1631) it doesn't show up at all. Prior to that it was the source of some flooding... As far as I can tell the relation is correctly defined, but maybe is is getting damaged by the splitter since it covers such a large area? I'm trying to generate a whole US map using extracts from the planet (currently using planet-100401): bzcat planet-100401.osm.bz2 | ~/osmosis-0.34/bin/osmosis -verbose --read-xml file=/dev/stdin enableDateParsing=no --bounding-box top=50 bottom=23 right=-66 left=-125 idTrackerType=BitSet completeWays=yes completeRelations=yes --write-xml file=usa-100401.osm java -Xmx4000M -jar ~/splitter.jar --cache=cache --max-areas=128 --max-nodes=800000 --mapid=70000001 --geonames-file=cities15000.zip usa-100401.osm <edit template.args> java -Xmx4096M -enableassertions -jar ~/mkgmap.jar -c template.args I'm using splitter rev 105 (the patch in rev 106 doesn't look like it affects me since I'm using a single OSM input file). -- Jeff Ollie
relation looks good. You should try to create just a single tile first to see if it's mkgmap or splitter. splitter clips ways from relations outside of the tile and this is a known problem. recent versions of mkgmap can repair in most cases by closing the multipolygon at the tile boundary. But it doesn't work in very complex situations. I see similar problems on large lakes, national park, … On 12 Apr 2010, at 9:49 , Jeffrey Ollie wrote:
On Mon, Apr 12, 2010 at 9:54 AM, Jeffrey Ollie <jeff@ocjtech.us> wrote:
Can someone take a look at relation 406175 (Skunk River and some related waterways in central Iowa)?
Actually that should be 406715.
-- Jeff Ollie _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
On Mon, Apr 12, 2010 at 12:51 PM, Apollinaris Schoell <aschoell@gmail.com> wrote:
relation looks good. You should try to create just a single tile first to see if it's mkgmap or splitter. splitter clips ways from relations outside of the tile and this is a known problem. recent versions of mkgmap can repair in most cases by closing the multipolygon at the tile boundary. But it doesn't work in very complex situations. I see similar problems on large lakes, national park, …
I just tried it with a single .osm file that was not processed with splitter and the relation is not rendered. I used the XAPI to download the extract rather than using the planet file this time around. I verified that the relation is completely included in the extract. I've put the source files and the results at <http://161.210.45.126/> if someone wants to take a look. -- Jeff Ollie
On Mon, Apr 12, 2010 at 12:51 PM, Apollinaris Schoell <aschoell@gmail.com> wrote:
relation looks good. You should try to create just a single tile first to see if it's mkgmap or splitter. splitter clips ways from relations outside of the tile and this is a known problem. recent versions of mkgmap can repair in most cases by closing the multipolygon at the tile boundary. But it doesn't work in very complex situations. I see similar problems on large lakes, national park, …
I just tried it with a single .osm file that was not processed with splitter and the relation is not rendered. I used the XAPI to download the extract rather than using the planet file this time around. I verified that the relation is completely included in the extract.
I've put the source files and the results at<http://161.210.45.126/> if someone wants to take a look.
I have one note: The wiki tells us something about tagging of multipolygons: "It is suggested to apply all tags which describe the area to the relation, and not to the ways. In many cases this may result in completely untagged ways." This means: if only one of the ways used for the relation is not completely tagged it may break the complete multipolygon. The mp 406715 does not follow this rule. I don't know if that's a problem or not but you may consider to tag the multipolygon instead of all ways. WanMil
On Thu, Apr 15, 2010 at 12:03 PM, WanMil <wmgcnfg@web.de> wrote:
The mp 406715 does not follow this rule. I don't know if that's a problem or not but you may consider to tag the multipolygon instead of all ways.
Thanks! This did seem to fix the problem for me... -- Jeff Ollie
participants (3)
-
Apollinaris Schoell -
Jeffrey Ollie -
WanMil