- mkgmap-dev - The mkgmap lists

SEVERE (MapSplitter) Errors - "Area too small to split at..."
by Peter Gardner 29 Dec '13

29 Dec '13

Hi, I'm getting lots of errors like this with both mkgmap-r2889 and mkgmap-r2651: SEVERE (MapSplitter): /media/space5/my.osm.stuff.overflow/step4and5/63240021.osm.pbf: Area too small to split at http://www.openstreetmap.org/?mlat=51.53796&mlon=-0.10158&zoom=17 (reduce the density of points, length of lines, etc.) Can't set city point index for N1P 2NS (too many indexed points in division) Can't set city point index for N1P 2NR (too many indexed points in division) However it does run to completion and a gmapsupp.img of a sensible size is created. The file works on my Garmin eTrex Vista HCx and I can see and navigate to the postcodes. But some points, e.g. N1P 2NS, are missing, although others nearby, e.g.N1P 2NN, exist. This has happened since I have been adding postcodes into a map of Great Britain, using some ideas from this recipe: http://wiki.openstreetmap.org/wiki/OSM_Map_On_Garmin/Postcode_Search_On_Gar… Anyway - I don't quite know what to do now. How do I set about reducing the density of points and length of lines? Is there something I should do prior to running mkgmap e.g. some splitter options I should change? Currently I'm just doing: time splitter \ /media/space5/my.osm.stuff.overflow/step3/great-britain-latest.osm.with.rightsofway.and.postcodes.pbf \ > splitter.log time java -Xmx3072M -jar /home/dad/osm.stuff/mkgmap-r2889/mkgmap.jar \ --gmapsupp \ --style-file=/home/dad/osm.stuff/mkgmap-r2651/examples/styles/mydefault/ \ --family-id=4538 \ --max-jobs=2 \ -c /media/space5/my.osm.stuff.overflow/step4and5/template.args \ /home/dad/osm.stuff/own.garmin.images/mydefault.v3.txt Cheers, Peter

2 1

Commit: r2917: Add undocumented option --no-mergeroads to disable merging of roads
by svn commit 28 Dec '13

28 Dec '13

Version mkgmap-r2917 was committed by wanmil on Sat, 28 Dec 2013 Add undocumented option --no-mergeroads to disable merging of roads Usually this is used for debugging purposes only.

1 0

Commit: r2916: Optimize logging statements in StyledConverter
by svn commit 28 Dec '13

28 Dec '13

Version mkgmap-r2916 was committed by wanmil on Sat, 28 Dec 2013 Optimize logging statements in StyledConverter

1 0

Commit: r2915: Fix the link-pois-to-ways handling
by svn commit 28 Dec '13

28 Dec '13

Version mkgmap-r2915 was committed by wanmil on Sat, 28 Dec 2013 Fix the link-pois-to-ways handling This includes: * road speed and road class is calculated in a method that considers all modifying tags * road speed and road class is modified in a copy of the GType object. This makes the handling much easier.

1 0

mkgmap error
by Jaromír Mikeš 28 Dec '13

28 Dec '13

Hi, I am trying adopt my style files to new release 2906 ... I succeed with points and lines files but having problem with polygons file. Getting error when trying build my map: $ java -ea -Xmx6144m -jar mkgmap-r2914/mkgmap.jar --precomp-sea=/media/mira/MEGA/MAPING/sea_20131101.zip --bounds=/media/mira/MEGA/MAPING/bounds_20131101.zip --style-file=garmin_hiking_map -c arguments --remove-ovm-work-files LAND/6324*.osm.pbf smallpois.TYP Time started: Sat Dec 28 16:45:21 CET 2013 Error in style: Error: (polygons:162): Stack size is 190 Error in style: Error: (polygons:162): Stack size is 190 Could not open style null Time finished: Sat Dec 28 16:45:23 CET 2013 Total time taken: 1418ms Error in style: Error: (polygons:162): Stack size is 190 Error in style: Error: (polygons:162): Stack size is 190 Error in style: Error: (polygons:162): Stack size is 190 etc. I tried 2913 2614 and 2889 releases but all behave same. My polygons file is here: https://github.com/miramikes/garmin_hiking_map/blob/master/polygons Any idea what can be wrong? Thank you for help mira

3 4

why do we have mkgmap:road-speed-class ?
by Gerd Petermann 28 Dec '13

28 Dec '13

Hi WanMil, I fear I don't understand the meaning of this new tag. Docu says: "A value between 0 and 7. Overrides the road_speed definition in the element type definition if this tag is set." We have mkgmap:road-speed which allows to increment, decrement or override the road speed value in the element type definition. Why did you introduce this new tag? Gerd

2 5

[patch v1] link-pois-to-ways and restrictions
by Gerd Petermann 28 Dec '13

28 Dec '13

Hi all, attached is a patch that changes the handling of option --link-pois-to-ways. In trunk, this option has two effects: 1) If a POI with a tag like access=mkgmap:road-speed=1 (e.g. highway=traffic-lights) is found in a way, the way is split at the preceding and following point and the change is applied to the remaining part. The length of the effected part depends only on the distance to the next point. If the way is straight, that can be > 300m, if the POI is in an a curve, it might be < 10m. 2) If a POI with a tag like mkgmap:bicycle=no (e.g. barrier=bollard) is found in a way, splits the way and before and after the POI so that the length of the affected way segment is typically less 100m. (max 50m before + max. 50m after the POI) Next, the way is split again at the POI, so that routing over the POI is not done for the restricted vehicles. The effect is that routing to a point close to the POI might not work (don't know that for sure, the device might ignore the restriction if the target cannot be reached "legally"). I think the wanted behaviour is this: 1) The change in speed should be applied to max. 25m before and after the POI or to the connection with another way, whatever comes first. 2) A POI like a barrier should not change access to the way before or after the POI, only at the POI. The attached patch implements this changed behaviour. The type 2) POI are converted to route restrictions. The patch also changes the meaning of route restrictions. In r2907 and before, route restrictions prohibit access for pedestrians and emergency, the patched version allows them by default. If I hear no complains, I'll commit that next sunday. Gerd

5 17

Commit: r2914: changes in unit tests
by svn commit 28 Dec '13

28 Dec '13

Version mkgmap-r2914 was committed by gerd on Sat, 28 Dec 2013 changes in unit tests - modify tests that check file sizes to match a range so that they print the actual size value - change the tested values to match exactly the actual values

1 0

A bug?
by Ataro A 27 Dec '13

27 Dec '13

If i use default as base style, i've a error. Is a bug or my error? # Base styles include 'lines' from default; <finalize> # Compatibility include file include 'inc/compat_lines' from default;

3 3

Commit: r2913: Improve error message if a style file contains a second finalize section
by svn commit 27 Dec '13

27 Dec '13

Version mkgmap-r2913 was committed by wanmil on Fri, 27 Dec 2013 Improve error message if a style file contains a second finalize section

1 0