[mkgmap-dev] Mkgmap dropping data
From WanMil wmgcnfg at web.de on Thu Aug 4 19:02:18 BST 2011
> Sorry, I thought MapSource usage is common on this list. The red areas > are where the map was completely rendered and you can mark them for > upload. The grey and yelllow areas are rendered only on the lowest > resolution, so they cannot marked for upload. MapSource is well known on this list but I didn't know the color code for selections. By the way: the MapSource cache can be cleared by typing <Ctrl>-G twice. > The rendering process is > broken, but mkgmap doesn't give any hint why. > NSIS is an alternative to MapSetToolkit? I'm satisfied with > MapSetToolkit, I don't think it has anything to do with the problem. Do > you have to use cgpsMapper with NSIS? NSIS is an windows installer toolkit (http://nsis.sourceforge.net/). When using the --nsis option mkgmap creates and .nsi file which can be used to build an installer containing the complete map. The installer unpacks the necessary files to a directory and performs the registry changes. So I don't see a need for any additional tool like cgpsmapper. I don't know if the MapSetToolkit is the problem but to fix your problem we need to sort out all tools which are not well known. And cgpsmapper is not well known because it's closed source. WanMil > > Zsolt > > On Wed, Aug 3, 2011 at 7:17 PM, WanMil <wmgcnfg at web.de > <mailto:wmgcnfg at web.de>> wrote: > > I am afraid but I have to ask: where's the problem? I don't see it. > > Maybe you could try to use the NSIS installer to install the map in > MapSource. This avoids to have some unknown software in the tool chain. > > By polygons I mean multipolygons and polygons. Multipolygons are > converted to normal polygons by mkgmap. > > WanMil > > > The red areas are the actually rendered areas. > > > > http://imageshack.us/photo/my-images/64/screenshot717201110719p.png/ > > > > This is the result by splitting it to 8 pieces. The eastern part is > > already OK. > > > > By polygons do you mean multipolygons? I'm sure I broke a lot of > > multipolygons by splitting the map into 36, but it seems to be > working. > > > > Zsolt > > Herrbert74 > > > > On Tue, Aug 2, 2011 at 7:52 PM, WanMil <wmgcnfg at web.de > <mailto:wmgcnfg at web.de> > > <mailto:wmgcnfg at web.de <mailto:wmgcnfg at web.de>>> wrote: > > > > > I had another go with splitter this morning and it worked out. > > But it's > > > strange. > > > If I split Hungary into 3-4 pieces, the eastern part is usually > > OK. But > > > even if I split Hungary into 33 pieces I get bad areas on the > western > > > part or near Budapest. I splitted it to 36 and it worked with > > this command: > > > java -Xmx1024m -jar splitter.jar --max-nodes=110000 > > > c:\osm\osmdata\hungary.osm.pbf > > > > What do you mean exactly with bad areas? > > Maybe you can send an image to show that. > > > > Usually for polygons it is better to select a higher > max-nodes and/or > > overlap parameter in the tile splitter because this increases the > > probability that the polygons are contained completely in the > tile data. > > > > > > > > But why is it that Eastern Hungary is OK with a million nodes? > Is it > > > because it's not so well mapped and it has less relations and/or > > errors? > > > Can I somehow rejoin the parts to have less areas? > > > > > > It would be useful if mkgmap would send a message which areas it > > > couldn't render and why. > > > > > > Another question: > > > Was I totally wrong using Osmosis to split the data or is there a > > method > > > which works with routing? To have counties is a bit nicer. I > used the > > > following (obviously breaks data): > > > call osmosis.bat --read-pbf c:\osm\osmdata\hungary.osm.pbf > > > --bounding-polygon file=c:\osm\osmdata\county_boundaries\pest.poly > > > completeWays=no completeRelations=no clipIncompleteEntities=true > > > --write-xml c:\osm\osmdata\counties\pest.osm > > > > I cannot say for sure but I would expect that using non > rectangular > > tiles is not really supported by mkgmap. > > > > > > > > 2011/8/1 Carlos Dávila <cdavilam at orangecorreo.es > <mailto:cdavilam at orangecorreo.es> > > <mailto:cdavilam at orangecorreo.es <mailto:cdavilam at orangecorreo.es>> > > > <mailto:cdavilam at orangecorreo.es > <mailto:cdavilam at orangecorreo.es> <mailto:cdavilam at orangecorreo.es > <mailto:cdavilam at orangecorreo.es>>>> > > > > > > El 01/08/2011 22:36, Zsolt Bertalan escribió: > > > > Hello > > > > > > > > I try to render a Garmin map of Hungary from OSM data with my own > > > > rules and typ file. At first I used the Tile Splitter, but > > whatever I > > > > did, I didn't manage to render a good map. > > > splitter should work fine with Hungary data and you'll get > > inter-tile > > > routing working. Can you provide the command you used to > > split the data? > > > > Then I extracted county borders (Hungary has 19 counties), > and used > > > > Osmosis with polygon borders to split the data. I managed it. But > > > > meanwhile CLC data was imported (I think this might be the > culprit) > > > > and I can't do it again. Four counties have only half the data in > > > > Mapsource. They are not big, so I don't think that the size > is the > > > > problem. > > > > I use MapSetToolkit with cgpsMapper to upload to MapSource. Here > > > is an > > > > example of how I use mkgmap: > > > > java -Xmx1024M -ea -jar mkgmap.jar --charset=latin2 --route > > > > --style-file=c:\osm\mkgmap\resources\styles\herrbert > > --family-id=5000 > > > > -c c:\osm\osmdata\counties\template.args --tdbfile > > --description="OSM > > > > Hungary" 50000.typ > > > If I remember right, -c option should be the last in you > > command line. > > > > > > > > I fI use the route option, cpreview crashes. > > > > > > > > I used osmosis with the following options: > > > > completeWays=no completeRelations=no clipIncompleteEntities=true > > > > I think this is not optimal for such a map, because I can't route > > > > between two counties (ways are split without a common node). > > But what > > > > are the best options? > > > > > > > > What am I doing wrong? If you need any more detail I'd be glad to > > > > provide it. > > > > > > _______________________________________________ > > > mkgmap-dev mailing list > > > mkgmap-dev at lists.mkgmap.org.uk > <mailto:mkgmap-dev at lists.mkgmap.org.uk> > > <mailto:mkgmap-dev at lists.mkgmap.org.uk > <mailto:mkgmap-dev at lists.mkgmap.org.uk>> > > <mailto:mkgmap-dev at lists.mkgmap.org.uk > <mailto:mkgmap-dev at lists.mkgmap.org.uk> > > <mailto:mkgmap-dev at lists.mkgmap.org.uk > <mailto:mkgmap-dev at lists.mkgmap.org.uk>>> > > > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev > > > > > > > > > > > > > > > _______________________________________________ > > > mkgmap-dev mailing list > > > mkgmap-dev at lists.mkgmap.org.uk > <mailto:mkgmap-dev at lists.mkgmap.org.uk> > > <mailto:mkgmap-dev at lists.mkgmap.org.uk > <mailto:mkgmap-dev at lists.mkgmap.org.uk>> > > > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev > > > > _______________________________________________ > > mkgmap-dev mailing list > > mkgmap-dev at lists.mkgmap.org.uk > <mailto:mkgmap-dev at lists.mkgmap.org.uk> > <mailto:mkgmap-dev at lists.mkgmap.org.uk > <mailto:mkgmap-dev at lists.mkgmap.org.uk>> > > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev > > > > > > > > > > _______________________________________________ > > mkgmap-dev mailing list > > mkgmap-dev at lists.mkgmap.org.uk > <mailto:mkgmap-dev at lists.mkgmap.org.uk> > > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev > > _______________________________________________ > mkgmap-dev mailing list > mkgmap-dev at lists.mkgmap.org.uk <mailto:mkgmap-dev at lists.mkgmap.org.uk> > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev > > > > > _______________________________________________ > mkgmap-dev mailing list > mkgmap-dev at lists.mkgmap.org.uk > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
- Previous message: [mkgmap-dev] Mkgmap dropping data
- Next message: [mkgmap-dev] Mkgmap dropping data
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the mkgmap-dev mailing list