- mkgmap-dev - The mkgmap lists

Showing labels next to POIs on Garmin Map
by viorap＠web.de 22 Jul '22

22 Jul '22

4 3

Merge tiles to gmapsupp based on poly files
by Daniel Vogelbacher 21 Jul '22

21 Jul '22

Hi all, the regular workflow for building maps is something like: 1. get osm pbf data for the region 2. split data using splitter.jar into rectangular tiles 3. build each tile 4. merge all map tiles into single gmapsupp.img But from my understanding, this raises issues on region boundaries when multiple regions (multiple gmapsuppXYZ.img files) are used on device. The data for an extract is usually defined by a .poly file, so everything outside of this polygon is not included. But splitter.jar splits not at polygon bounds but by rectangles of different sizes. So some tiles for a map contains not the full osm data for this rectangle because the data was already cut off during osm data extraction. For example, building the map for Germany and France by two separate osm extracts leads to missing data on the polygon bounds for these tiles. And I assume this breaks routing across maps. A better approach would be to split and build from the whole world (or at least Europe), then merge all tiles for Germany into gmap_germany.img and all tiles from France into gmap_france.img. The tiles on country borders must be included in both .img files but because they have the same map-id, the device should be able to take care of duplicated map tiles. The area of tiles can be defined by the same .poly file used in the regular workflow, but now as a selector for tiles instead of osm data. So because I don't see anything in the docs mentioning this workflow, I assume this is not how it works :-) What is the recommended workflow to build multiple maps and providing proper routing across multiple gmapsupp.img - or is this just not possible with Garmin devices? -- Best regards / Mit freundlichen Grüßen Daniel Vogelbacher

2 3

Raise an error when levels and dem-dists mismatch
by Daniel Vogelbacher 14 Jul '22

14 Jul '22

Hi all, I've debugged the past two days why a custom map won't be recognized by my Garmin Oregon 700 device. The issue was the dem configuration. My style uses this level mapping: levels = 0:24, 1:22, 2:20, 3:18, 4:16 While my dem distribution line was: dem-dists=3312,6624,9936,13248,16560,19872,23184,26496 So there are only 5 levels defined, but the distribution maps to 8. (I copied the dem-dists line from Freizeitkarte-style without thinking about it...) The build process was successfully, but the map was never listed in the device map selection. After fixing the level mapping in the style/option file, the map gets listed. So may I ask to add a check/exception to prevent building maps when there is a mismatch between levels vs. dem-dists? Thank you! -- Best regards / Mit freundlichen Grüßen Daniel Vogelbacher

2 4

Commit r4905: Stop execution and report error when number of DEM dist values is higher than the levels.
by svn commit 13 Jul '22

13 Jul '22

Version mkgmap-r4905 was committed by gerd on Wed, 13 Jul 2022 Stop execution and report error when number of DEM dist values is higher than the levels. http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4905

1 0

Error while using splitter-r652
by viorap＠web.de 07 Jul '22

07 Jul '22

2 1

Using OSM card on garmin / summit names/height
by viorap＠web.de 04 Jul '22

04 Jul '22

2 2

Osmium -Tool
by Thomas Morgenstern 02 Jun '22

02 Jun '22

Hi, maybee this is not the right forum, but I assume there are many spezialists regading OSM , reading it. I search for a ready to use compiled version for the OSMIUM- Tool. I found only a few hints, how to compile it using LINUX. But I am not able to compile it for myself. Has anybody such a OSMIUM.EXE and is willing to share it ? If so, contact please thomasmorgenstern(a)web.de Regards Thomas

2 1

Some observations
by Andrzej Popowski 31 May '22

31 May '22

Hi, I have tweaked my maps recently and tripped on some problems. Nothing serious, but a bit troublesome. First I noticed, that tweak for better looking roundabouts not always works. In default style there is something like that: junction=roundabout & (highway=tertiary | highway=tertiary_link) [0x0c road_class=1 road_speed=1 resolution 24 continue] junction=roundabout & (highway=tertiary | highway=tertiary_link) [0x10804 resolution 21] I have noticed, that sometimes I get roundabout partially covered by line 0x10804. Maybe mkgmap is splitting roundabout and a part of it is written into img after line 0x10804? Sorry, I have reworked my style and can't provide an example now. Then I got problems with --make-opposite-cycleways. Looks like this option is not executed at finalize step, but the created line gets full processing in style. If there exist somewhere a roundabout with opposite cycle line, then default style will create double lines 0x10804. Safe version of style would be: mkgmap:synthesised!=yes & junction=roundabout & (highway=tertiary | highway=tertiary_link) [0x10804 resolution 21] Then I have found that following doesn't work: if (mkgmap:synthesised!=yes) then junction=roundabout & (highway=tertiary | highway=tertiary_link) [0x10804 resolution 21] end Result: Mkgmap version 4902 Time started: Sun May 29 21:47:50 CEST 2022 SEVERE (global): Error in style: Error: (lines:15): Invalid rule, expression cannot be indexed: $mkgmap:synthesised!=yes SEVERE (global): Could not open style But this one works: if (mkgmap:synthesised=yes) then else junction=roundabout & (highway=tertiary | highway=tertiary_link) [0x10804 resolution 21] end I understand, that there is some kind of optimization for keywords, but why one type of condition works while others don't? Mkgmap accepts "keyword=false" but not "keyword=!true" nor "!(keyword=true)". If this is unavoidable, then would be nice to get a warning in manual. Finally I made an error in style and crashed mkgmap. I wrote something like this: if (mkgmap:style-option:test=true) then () echo {"test"} end and got following: Mkgmap version 4902 Time started: Sun May 29 21:06:03 CEST 2022 SEVERE (global): Unexpected error java.lang.NullPointerException: Cannot invoke "uk.me.parabola.mkgmap.osmstyle.eval.Op.isType(uk.me.parabola.mkgmap.osmstyle.eval.NodeType)" because the return value of "uk.me.parabola.mkgmap.osmstyle.eval.Op.getFirst()" is null at uk.me.parabola.mkgmap.osmstyle.ExpressionArranger.selectivity(ExpressionArranger.java:381) at uk.me.parabola.mkgmap.osmstyle.ExpressionArranger.arrangeAndChain(ExpressionArranger.java:348) at uk.me.parabola.mkgmap.osmstyle.ExpressionArranger.arrangeTop(ExpressionArranger.java:95) at uk.me.parabola.mkgmap.osmstyle.ExpressionArranger.arrange(ExpressionArranger.java:74) at uk.me.parabola.mkgmap.osmstyle.RuleFileReader.loadFile(RuleFileReader.java:132) at uk.me.parabola.mkgmap.osmstyle.RuleFileReader.load(RuleFileReader.java:102) at uk.me.parabola.mkgmap.osmstyle.StyleImpl.readRules(StyleImpl.java:303) at uk.me.parabola.mkgmap.osmstyle.StyleImpl.<init>(StyleImpl.java:158) at uk.me.parabola.mkgmap.osmstyle.StyleImpl.readStyle(StyleImpl.java:546) at uk.me.parabola.mkgmap.reader.osm.OsmMapDataSource.createConverter(OsmMapDataSource.java:324) at uk.me.parabola.mkgmap.reader.osm.OsmMapDataSource.setupHandler(OsmMapDataSource.java:206) at uk.me.parabola.mkgmap.reader.osm.OsmMapDataSource.parse(OsmMapDataSource.java:166) at uk.me.parabola.mkgmap.reader.osm.OsmMapDataSource.load(OsmMapDataSource.java:142) at uk.me.parabola.mkgmap.main.MapMaker.loadFromFile(MapMaker.java:165) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:58) at uk.me.parabola.mkgmap.main.Main.lambda$processFilename$1(Main.java:291) at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:264) at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1136) at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:635) at java.base/java.lang.Thread.run(Thread.java:833) SEVERE (global): Exiting due to unexpected error Number of ExitExceptions: 1 -- Best regards, Andrzej

4 10

Errors when using --gmapi-minimal
by dfjkman＠gmail.com 30 May '22

30 May '22

Hi Gerd, I decided to have a go using the --gmapi-minimal option and may have come across a potential bug. When using the include-pattern and some .img files are skipped because they do not match the pattern, but exist in the input folder containing pre written files, these skipped files are still written to the *.tdb file but not included in the map. In other words using the pattern example given --gmapi-minimal=.*4711[0-9]{4}\.img, matching files between 47110000.img and 47119999.img, if there is also a file 47120000.img this will be skipped but there will still be a reference to the skipped in the *.tdb file the same with any overview files that may exist in the input folder. I think having any overview maps in the input folder will generate the following error messages: gmapi-minimal: Skipping file C:\Users\Dave\Documents\Maps\Data\DemImg\Zambia_Dem.img gmapi-minimal: Skipping file C:\Users\Dave\Documents\Maps\Data\ContImg\Zambia_Contours.img WARNING (global): Could not copy 63240000.RGN uk.me.parabola.imgfmt.FileExistsException: File 63240000.RGN already exists WARNING (global): Could not copy 63240000.TRE uk.me.parabola.imgfmt.FileExistsException: File 63240000.TRE already exists WARNING (global): Could not copy 63240000.LBL uk.me.parabola.imgfmt.FileExistsException: File 63240000.LBL already exists SEVERE (global): Invalid mapname for C:\Users\Dave\Documents\Maps\Data\DemImg\Zambia_Dem.img, perhaps it is not a .img file SEVERE (global): Invalid mapname for C:\Users\Dave\Documents\Maps\Data\ContImg\Zambia_Contours.img, perhaps it is not a .img file I discovered the errors while using JaVaWa GMTK when checking the maps for errors. I am creating a map including pre written DEM and contour information to try find badly mapped rivers that go uphill and over ridges in BaseCamp. Kind regards, Dave

2 11

Commit r4904: mp-rendering.patch:
by svn commit 30 May '22

30 May '22

Version mkgmap-r4904 was committed by gerd on Mon, 30 May 2022 mp-rendering.patch: Handle special case with a complex multipolygon that contains the whole tile without intersection. This can lead to large missing rectangular parts of the multipolygon. http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4904

1 0