Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-10 Thread Marko Mäkelä
> I will update http://www.polkupyoraily.net/osm/ as soon as I get it > right (hopefully today). Done. My osm2img.sh script is already there, and the gmapsupp.img will be soon. Actually the gmapsupp.img will be from an older version of the script that did not adjust the coastline at the NW,

Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-10 Thread Marko Mäkelä
10.03.2010 14.32.39, Clinton Gladstone wrote: > On Wed, Mar 10, 2010 at 8:09 AM, Marko Mäkelä > wrote: > >> If I include the Swedish, Russian and Estonian coastlines to reach >> the tile borders, then mkgmap will take forever to generate the sea. >> Maybe I should try with a grossly simplif

Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-10 Thread Clinton Gladstone
On Wed, Mar 10, 2010 at 8:09 AM, Marko Mäkelä wrote: > If I include the Swedish, Russian and Estonian coastlines to reach the > tile borders, then mkgmap will take forever to generate the sea. Maybe > I should try with a grossly simplified coastline (just straight lines > to the tile borders). I

Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-09 Thread Marko Mäkelä
09.03.2010 22.35.00, Clinton Gladstone wrote: > I have been using Osmosis 0.31.1. At least with this version, > disabling date parsing made a huge difference. It could be that more > recent releases are more efficient in this regard. I am using 0.34, and it does not seem to have problems with

Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-09 Thread Clinton Gladstone
On Mar 9, 2010, at 20:20, Marko Mäkelä wrote: >> I regularly take the entire Geofabrik extract of Europe and extract a >> bounding-box containing all of Germany. > > Have you been able to make it utilize all cores? I have a dual-core computer, > and java is eating only one CPU. Apparently, the b

Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-09 Thread Marko Mäkelä
On Tue, Mar 09, 2010 at 06:51:44PM +0100, Clinton Gladstone wrote: > You should disable date parsing (enableDateParsing=no); this > considerably speeds up Osmosis. > > See further discussion here: > > http://wiki.openstreetmap.org/wiki/Osmosis/Examples > > I regularly take the entire Geofabrik e

Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-09 Thread Clinton Gladstone
On Tue, Mar 9, 2010 at 3:47 PM, Marko Mäkelä wrote: > > When I added > --bb completeWays=yes left=19.1162109375 right=31.5966796875 > bottom=59.4140625 top=70.0927734375 > before the --wx in the chain, I had to interrupt it after letting it > run several minutes. You should disable date parsing (

Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-09 Thread Marko Mäkelä
08.03.2010 11.04.54, Marko Mäkelä kirjoitti: > There is some flooding in the east border, as expected. I would like > to fix it by patching finland.osm.bz2 from Geofabrik as follows: I downloaded all truncated coastlines and a section of the Swedish, Russian, and Estonian natural=coastline to "s

Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-09 Thread Marko Mäkelä
09.03.2010 15.05.04, Marko Mäkelä kirjoitti: > 08.03.2010 11.04.54, Marko Mäkelä kirjoitti: > Osmosis can merge my separately downloaded coastline to > finland.osm.bz2 at a bearable speed (less than 10 minutes), but > splitting with completeRelations=yes completeWays=yes took way too > long

Re: [mkgmap-dev] Finland with almost-working generate-sea

2010-03-08 Thread Marko Mäkelä
08.03.2010 09.06.40, Marko Mäkelä wrote: > 08.03.2010 00.43.17, Clinton Gladstone wrote: >> I can confirm the same for a map of Germany. With 1597, everything >> was (for perhaps the first time) perfect. With 1598, all coastline >> tiles are flooded. > > Same here: I got a completely flooded