Re: [mkgmap-dev] Maintaining generated sea polygons
On Thu, May 20, 2010 at 10:57:54PM -0700, NopMap wrote:
Therefore my question: Would it be possible to save the generated sea
polygons to disk in osm format so an intact set of polygons can be
re-used with future osm data? That way it could be ensured that future
updates will have the same, correct sea polygons and not be randomly
broken by temporary inconsistencies in the osm data.
I've found this command useful:
osmosis --rx finland.osm.bz2 --tf accept-ways natural=coastline
--used-node --wx finland-coastline.osm
The resulting file is small enough to be loaded to JOSM. I have also
done something similar with boundary=administrative.
Storing the image files and using them as a seperate map layer is not a
solution as older Garmin devices are really cumbersome with layers or
cannot handle multiple layers at all - and the map is for an older
device.
Sorry, I can't help there. A couple of months ago, I tried merging some
additional coastline to finland.osm, but it did not work for me. I did
not investigate further, but instead decided to patch the data (move the
end nodes of the coastline in the map extract to the tile boundary,
creating ugly vertical or horizontal lines). Like this:
bzip2 -dc $OSM_BZ2|
perl -e \
'my $del=0;
while(){
$del=1 if (/relation.* version=1.* user=usm78-gis/);
s/(node id=28954644.*lat=)60\.51564/$159.326172/;
s/(node id=29193143.*lon=)24\.12826/$119.072266/;
print unless $del;
$del=0 if m|/relation|;
}' $OSM
This one also gets rid of generated, slightly erroneous landuse
multipolygons in Russia. I used to use a similar technique for omitting
incomplete coastline sections (sea islands), but I fixed that by editing
the Geofabrik cutting polygon. It took a few iterations, but Frederik
Ramm was patient with me.
The coastline of Finland seems to be pretty static; I can't remember
when I last saw a generate-sea warning. There has been a tempest in
teapot regarding converting lakes away from natural=coastline. When I
get around to it, I should convert the last big lake (Päijänne) to
natural=water multipolygons.
Marko
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Re: [mkgmap-dev] Maintaining generated sea polygons
Moin, NopMap schrieb am 21.05.2010 07:57: For three weeks I have been unable to grab a planetfile with correct coastline for central europe - they are broken at about the same speed we can fix them. So even though I was able to build a really nice map a while ago, all update attempts were broken. I always use the same tile borders for my maps. So when the generation of a single tile fails, I can use the last usable version of this tile for getting a complete mapset. Gruss Torsten ___ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Re: [mkgmap-dev] Maintaining generated sea polygons
Hi, I don't know if this is the same problem as Nop described, but lately I found that on my map some sea areas were gone in the Netherlands: see http://img338.imageshack.us/img338/9992/errorcoastline.jpg On osm it looks ok: http://www.openstreetmap.org/?lat=52.817lon=4.955zoom=10layers=B000FTF Im using the geofabrik abstract from europe to compile my benelux map, Any idea how to fix this? I tried another abstract (planet-benelux) but this gave the same results. Is it caused by the splitter? Is it an error in de osm data? In the mkgmap args i use generate-sea: no-mp,extend-sea-sectors,close-gaps=1000 and it never went wrong except for the last two weeks. Nop wrote: Hi! I have been playing around with generated sea polygons. The algorithm works remarkably well - but I have been experiencing trouble due to inconsistent use of natural=coastline in the data. There was an invalid use of natural=coastline in the planetfile, that caused a whole tile in the Alps to be flooded. So I tracked it down, fixed it and waited for the next planetfile. This time another tile near Berlin was flooded. I tracked it down, noted that another mapper had already fixed it and waited for the next planetfile. This time another tile south of Würzburg was flooded... - you note the pattern? :-) For three weeks I have been unable to grab a planetfile with correct coastline for central europe - they are broken at about the same speed we can fix them. So even though I was able to build a really nice map a while ago, all update attempts were broken. Therefore my question: Would it be possible to save the generated sea polygons to disk in osm format so an intact set of polygons can be re-used with future osm data? That way it could be ensured that future updates will have the same, correct sea polygons and not be randomly broken by temporary inconsistencies in the osm data. Storing the image files and using them as a seperate map layer is not a solution as older Garmin devices are really cumbersome with layers or cannot handle multiple layers at all - and the map is for an older device. Any ideas? Nop -- View this message in context: http://gis.638310.n2.nabble.com/Maintaining-generated-sea-polygons-tp5082865p5082865.html Sent from the Mkgmap Development mailing list archive at Nabble.com. ___ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev ___ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
[mkgmap-dev] Fedora packaging for mkgmap
I'm currently working on packaging mkgmap for Fedora[0]. I'm using revision 1647, as it was the most up-to-date when I was doing the packaging. I'd welcome any recommendations or advice on which revision to choose for the initial packaging, or any other comments. [0] https://bugzilla.redhat.com/show_bug.cgi?id=592645 -- Bruce Cowan br...@bcowan.me.uk ___ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Re: [mkgmap-dev] Fedora packaging for mkgmap
On 21.05.2010 19:59, Bruce Cowan wrote: I'm currently working on packaging mkgmap for Fedora[0]. I'm using revision 1647, as it was the most up-to-date when I was doing the packaging. I'd welcome any recommendations or advice on which revision to choose for the initial packaging, or any other comments. [0] https://bugzilla.redhat.com/show_bug.cgi?id=592645 The newest, last two month were bugfixes only. ___ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Re: [mkgmap-dev] Debugging generated sea polygons
On Fri, May 21, 2010 at 07:34:55PM +0200, Minko wrote: On osm it looks ok: http://www.openstreetmap.org/?lat=52.817lon=4.955zoom=10layers=B000FTF Someone mentioned that the sea polygons on the Mapnik layer are not updated at once. Mapnik is using old data that comes from a complex process, as far as I understood. Im using the geofabrik abstract from europe to compile my benelux map, Any idea how to fix this? I tried another abstract (planet-benelux) but this gave the same results. Extract the natural=coastline from your map extract or the individual tiles and check it in JOSM, using the Validator plugin. Pay attention to the corners or special looking lines or points. You may also find my wayselector plugin useful. I wrote it for selecting the mainland coastline (there were quite a few natural=coastline islands and puddles around it on the Finnish coast, and I wanted to tag the puddles as natural=water). Is it caused by the splitter? Is it an error in de osm data? In the mkgmap args i use generate-sea: no-mp,extend-sea-sectors,close-gaps=1000 and it never went wrong except for the last two weeks. Pay attention to the mkgmap warnings. Enable logging and check the log file. See http://www.polkupyoraily.net/osm/ for an example (osm2img.sh, logging.properties, logging.ignore). Marko ___ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
