- mkgmap-dev - The mkgmap lists

Address index incomplete
by Alessio Medici 05 Dec '24

05 Dec '24

Hi, I am trying to create maps with address index, but it seems the index contains only a few (something like 10) cities. >From geofabrik.de I have downloaded the .osm.pbf file for Europe than I have run splitter and mkgmap using the following command lines: java -jar "/home/Alessio/bin/splitter/splitter.jar" \ --description="Italia" \ --output-dir=. \ --geonames-file="../Italia.txt" \ --wanted-admin-level="11" \ --mapid="13900000" \ --max-nodes="5000000" \ --polygon-file="../Italia.poly" \ "/home/Alessio/GarminMapsCreation/planetfile/europe-latest.osm.pbf" java -jar "/home/Alessio/bin/mkgmap/mkgmap.jar" \ -Dlog.config=logging.properties \ --max-jobs \ --gmapsupp \ --output-dir="gmapsupp" \ --country-name="Italia" \ --family-name="Italia" \ --family-id="42" \ --area-name="Italia" \ --description="Italia" \ --series-name="OSM generic routable(Italia_20240728)" \ --product-id="42" \ --mapname="13900000" \ --country-abbr="" \ --draw-priority="26" \ --precomp-sea=../../sea-latest.zip \ --generate-sea \ --route \ --drive-on=detect,right \ --report-roundabout-issues \ --fix-roundabout-direction \ --process-destination \ --process-exits \ --name-tag-list=int_name,name \ --remove-ovm-work-files=true \ --bounds=../../bounds-latest.zip \ --verbose \ --location-autofill=is_in,nearest \ --make-poi-index \ --index \ --split-name-index \ --housenumbers \ --preserve-element-order \ --dem="/home/Alessio/GarminMapsCreation/hgt" \ --style-file=../../default_style \ --keep-going \ --improve-overview \ --min-size-polygon=15 \ -c "gmapsupp/template.args" The java version I use is: java 23 2024-09-17 Java(TM) SE Runtime Environment (build 23+37-2369) Java HotSpot(TM) 64-Bit Server VM (build 23+37-2369, mixed mode, sharing) on Ubuntu 24.04 The map is OK and I can use it for routing, but I have to choose manually the start and end point directly on the map. Can someone point me out where I am wrong? Thank you very much Alessio

2 2

Re: [mkgmap-dev] Problem with newer garmin device text labels not appearing
by scott taggart 03 Dec '24

03 Dec '24

Hi Felix, Thanks for the response. I am just getting started on my debugging and, in fact, one of the things I was going to try was changing the POI "types". I will let you know how it goes. Regarding line "types", I generate custom line types and supply the line pattern as opposed to using any of the well-known types. Maybe this could be a work-around for you? I could supply examples for you, if needed. The whole visibility of objects at different zoom levels has always been a challenge and is very device dependent, unfortunately. For what I do, I generally don't run into it too much (because I don't generate references to device-internal icons, etc.). As a whole, Garmin's UI and compatibility between devices is terrible and unfortunate. I think it stems for poor design decisions, limited memory, etc. I've worked at software companies for 4 decades (as a developer) and I've seen this kind of sloppy design and poor management. It is what it is and someday the phones will eclipse all this poor hardware and software. I saw you registered at my website. Please watch the (poor and outdated) videos and try it out. Let me know if you need any help and I am always willing to share all that I have. Oh, I just received a Tread and it also does not display the text for a "POI" type. I will start debugging now... Best of luck on your project. Scott

1 0

Problem with newer garmin device text labels not appearing
by scott taggart 02 Dec '24

02 Dec '24

Hello, Please forgive any missing information and clarity - my first post here. I have a (free, primitive) website ("TrickTracks <https://mapmaker.tricktracks.org/server/trick_tracks/file_uploader.php>") that allows users to "drop" a gpx file onto a webpage and receive a garmin .img map file as a download. Under the covers my My web server converts the gpx file to an MP file and then feeds mkgmap (currently r4000) to produce the .img file. I've had this website live for about 5 years and it's been well received by my users. The users are primarily motorcycle and bicycle riders who create tracks for their rides and use the maps on their various Garmin devices. My software does a bit of "special" processing on the gpx files in that certain waypoints are converted to large polygon targets and the gpx track lines have mileage marker polygons added at the user requested intervals. There are various web-page configuration options, etc. that let users specify map line width, etc. This has worked well on almost all garmin devices for many years. I now have a problem with some of the newer garmin devices (XT2 and Treads) where the labels for the polygons are not showing up. I'm looking for ideas on where to start with this issue. I have a few ideas and would like ideas on how to proceed: * First, is this a know problem in this community? Have the mkgmap/OSM developers already tackled this issues and has mkgmap been changed in some way to deal with it? I will admit I have not upgrade my server yet to the latest mkgmap. * I am assuming garmin changed something in their rendering software to now show labels that used to work (I suppose that's obvious). I have tried to see if there was a device setting that affected this but have not been able to find any that work. * Is there some limitation with using MP file input to mkgmap that will, in the future, hinder generating .img file for these newer devices. I do not generate and JOSM and have not researched that as it seems like overkill for my application. * Is there perhaps something in my TYP input to mkgmap that could cause this (I self generate TYP files on the fly but they have worked for years and still do on the XT, Montana, etc.). * As a side-note, the labels on my MP polylines do display correctly. * I can provide the MP file contents I am generating for the failing waypoint labels if that would help. Thanks you for any input and help you can give me. Scott

2 1

Dealing with JOSM exports
by Felix Herwegh 22 Nov '24

22 Nov '24

Hi, for some special application I'm parsing gpx-files to osm-files, identical to importing the .gpx into JOSM, convert them to datalayers and finally export to .osm. These files are then processed by splitter to .o5m. I'd like to add a timestamp to the .o5m using "osmconvert --timestamp" next. While that normally works for split-files based on OSM data, for the above files osmconvert allways throws 6 warnings like osmconvert Warning: wrong sequence at node -324213 osmconvert Warning: next object is node -324214 osmconvert Warning: wrong sequence at node -324214 osmconvert Warning: next object is node -324215 osmconvert Warning: wrong sequence at node -324215 osmconvert Warning: next object is node -324216 I guess, osmconvert is expecting ordered increasing positive IDs only. Iirc the decreasing negative IDs are correct here nevertheless, implemented long ago to indicate dealing with non OSM objects. mkgmap seems to work perfectly fine on such split-files. Did I miss any means to avoid these warnings and have the timestamp added other than converting the IDs to positive? Thanks //Felix

2 1

Internal Error Failed to render Multipolygon
by Thomas Morgenstern 02 Sep '24

02 Sep '24

Hello , I use mkgmap-4921 with patch from Ticker Berkin -Angelchecker-V6. With that i get inScandinavia following error : Internal Error Failed to render MultiPolygonRelation 1754729 and 304507. Rel. 1754729 is 'Ladoga'-sea with 5 226 Elements. Rel. 304507 is Multipolygon=water, 171 Elements. I assume the error ist not specific to the patch...and more common.. How to avoid this ? regards Thomas

3 2

Commit r4922: improve AngleChecker to fix wrong or missing routing hints where roads are spltting or crossing at sharp angles.
by svn commit 31 Aug '24

31 Aug '24

Version mkgmap-r4922 was committed by gerd on Sat, 31 Aug 2024 improve AngleChecker to fix wrong or missing routing hints where roads are spltting or crossing at sharp angles. Patch AngleChecker_v6.patch by Ticker Berkin http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4922

1 0

AngleChecker and --cycle-map option
by Ticker Berkin 23 Aug '24

23 Aug '24

Hi Gerd I'm confused by some logic in imgfmt/app/net/AngleChecker.java and wondered if you could help. If not a cycle-map, sharpAnglesCheckMask is set to ACCESS...BIKE = 0x2 (~line 134) Later (~line 294) when the code is looking at an angle and the arcs either side it ignores a sharp angle (ie won't try and expand it) if BIKE access isn't common to both arcs. This could result in time penalties for cars based If it is a cycle-map, sharpAnglesCheckMask is set to all but ACCESS...FOOT, so that if the common access of the arcs is FOOT only, sharp angles will be ignored, but there is logic to do this anyway. Ticker

2 1

Patch to add AppStream metadata
by Sebastiaan Couwenberg 22 Aug '24

22 Aug '24

Petter Reinholdtsen provided a patch in Debian Bug #1078746 to add AppStream metadata for mkgmap. " Here is a patch to add Appstream metainfo XML announcing the hardware handled by this package. Including this information in the package will ensure programs mapping hardware to packages using Appstream information, like the isenkram package, will know that this package is useful on machines where the hardware is discovered. The appstream metadata file can be checked using this command after package build: appstreamcli validate-tree --no-net --explain debian/mkgmap " https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1078746 Kind Regards, Bas -- GPG Key ID: 4096R/6750F10AE88D4AF1 Fingerprint: 8182 DE41 7056 408D 6146 50D1 6750 F10A E88D 4AF1

5 5

some address related errors
by 7770 17 Jun '24

17 Jun '24

Hi. I only observed today that when i run mkgmap (4921) i get lots of messages such as the below. I can't say if it existed earlier, i simply didn't watch to much in the logs. SEVERE (HousenumberGroup): full/splitted/77700746.osm.pbf: Internal error: ExtNumbers.rasterLineNearPoint() found no point, cannot improve address search for road id=52829593, North Farm Road SEVERE (ExtNumbers): full/splitted/77700746.osm.pbf: internal error, distance to road too large, road id=52829669, Oak Tree Drive house 3(1) http:// www.openstreetmap.org/node/2867200804 Is it something to worry about? Regards Karl

3 3

Geofabrik
by osm 16 Jun '24

16 Jun '24

Looks like Geofabrik today is causing maps to look like a spider's web! Have tried UK and Luxembourg. Regards Nick

3 3