August 2015 - mkgmap-dev - The mkgmap lists

Mapsource crashes while calculating a route
by Andrzej Popowski 29 Sep '16

29 Sep '16

Hi, I have observed crashes for some time. Mapsource 6.16.3 indicates a bug in module MPL_MAP.CPP. Basecamp reports an error or goes into infinite loop. Now I have extracted minimal source data, which creates faulty map. Removing any road or even a crossing point restore routing. Bug exist in mkgmap r2815, r2857, r2889, r2975 and r2976. Map created by mkgmap r2795 works correctly. I'm attaching here files to reproduce problem: mkgmap.bat - batch, which I have used to create the map, test-a.pbf - very simple OSM source data, route.gpx - short route, which causes a crash in Mapsource. -- Best regards, Andrzej

4 15

process-exits vs process-destination
by Franco Bez 08 Oct '15

08 Oct '15

Hi all, is there a way to combine the features process-exits and process-destination in a single style ? Based on my experience process-desitnation does a marvellous job when entering a motorway ( does it also work with also trunk roads ? ), while process-extis is perfect when leaving a motorway or trunk road. For instance the roundabout here http://www.openstreetmap.org/?lat=47.89665&lon=10.64805&zoom=17 has two exits towards the B12. Routing from the "Sudetenstraße" going North (maybe towards Augsburg) I would like the Garmin device to announce "5 th Exit - B12 (München)". On the other hand routing to the South (maybe Kempten) it should announce "2nd exit - B12 (Kempten)". Using process-exits this does not work, it's only "5th Exit - B12" resp. "2nd Exit - B12". I would have to use process-destination instead to get the desired messages displayed. Coming home from the trip, routing from the B12 towards "Sudetenstraße" I would like "exit to the right - Exit Kaufbeuren". Using process-exits this works perfectly. Would I use process-destination I would get "exit to the right towards B12 Verteilerring" or even "exit to the right towards Sudetenstraße" a really useless announcement as the sign on the trunk road reads "Kaufbeuren / Neugablonz / Mauerstetten". So this is why I would like to have an option that combines the functionality of process-exits and process-destination. Do you see any way to achieve this in one single style ? Ciao, Franco

7 6

[Patch v2] sharp angles
by Gerd Petermann 02 Sep '15

02 Sep '15

Hi all, attached is v2 of the patch, the binary is here: Correction: the previous versions did not change angles like this: /\ Improvement: For bicycle maps, the angle is changed up to 90° for arcs with high road speeds. Options (--x-fix-sharp-angles and --x-cycle-map) are the same. @Felix: I wait for your feedback now, as I have no way to test with your style. Gerd

3 9

Commit: r3639: further improve handling of addr:interpolation ways
by svn commit 31 Aug '15

31 Aug '15

Version mkgmap-r3639 was committed by gerd on Mon, 31 Aug 2015 further improve handling of addr:interpolation ways When an interpolated address is the same as an address from a single OSM element, try to include the OSM element in the addr:interpolation way. When that happens, the interpolation is done again. As a result, the interpolated addresses near the OSM element are in the correct order.

1 0

Commit: r3638: fix two small problems in the unit tests:
by svn commit 31 Aug '15

31 Aug '15

Version mkgmap-r3638 was committed by gerd on Mon, 31 Aug 2015 fix two small problems in the unit tests: 1) an osmmap.img was remaining 2) a file args was not closed and produced this error message: [junit] Running uk.me.parabola.mkgmap.CommandArgsTest [junit] Tests run: 6, Failures: 0, Errors: 0, Time elapsed: 0,071 sec [junit] Exception in thread "Thread-1" java.lang.AssertionError: delete existing file: args [junit] at org.junit.Assert.fail(Assert.java:88) [junit] at org.junit.Assert.assertTrue(Assert.java:41) [junit] at func.lib.TestUtils.deleteOutputFiles(TestUtils.java:75) [junit] at func.lib.TestUtils$1.run(TestUtils.java:59) [junit] at java.lang.Thread.run(Thread.java:745)

1 0

[Patch v1] unit tests
by Gerd Petermann 31 Aug '15

31 Aug '15

Hi Steve, please review the attached patch. It fixes two small problems in the unit tests: 1) an osmmap.img was remaining 2) a file args was not closed and produced this error message: [junit] Running uk.me.parabola.mkgmap.CommandArgsTest [junit] Tests run: 6, Failures: 0, Errors: 0, Time elapsed: 0,071 sec [junit] Exception in thread "Thread-1" java.lang.AssertionError: delete existing file: args [junit] at org.junit.Assert.fail(Assert.java:88) [junit] at org.junit.Assert.assertTrue(Assert.java:41) [junit] at func.lib.TestUtils.deleteOutputFiles(TestUtils.java:75) [junit] at func.lib.TestUtils$1.run(TestUtils.java:59) [junit] at java.lang.Thread.run(Thread.java:745) Gerd

2 1

Commit: r3637: improve checks for raddr:interpolation ways
by svn commit 30 Aug '15

30 Aug '15

Version mkgmap-r3637 was committed by gerd on Sun, 30 Aug 2015 improve checks for raddr:interpolation ways - log a warning message when city or zip code is different, no effect on output - ignore duplicate addresses produced by interpolated houses when the distance is < 100 (was < 40) - don't ignore duplicate addresses produced by interpolated houses when the distance is > 150 (this check was added by mistake with r3635) - improve logging: report more details about the elements that cause a conflict - only ignore addr:interpolation ways with equal street and start and end numbers when they are in the same city, if so, report both ways in the "found additional addr:interpolation way with same meaning ... " message

1 0

Commit: r3636: remove dead code
by svn commit 30 Aug '15

30 Aug '15

Version mkgmap-r3636 was committed by gerd on Sun, 30 Aug 2015 remove dead code

1 0

Commit: r3635: corrections to addr:interpolation way checks
by svn commit 29 Aug '15

29 Aug '15

Version mkgmap-r3635 was committed by gerd on Fri, 28 Aug 2015 corrections to addr:interpolation way checks - The checks did not notice when a "duplicated number" belongs to a different cityInfo - The checks allowed only 5m distance between an existing address and the interpolated one, this is changed to 40m - some improvements in logging

3 3

[Patch v1] mkgmap ignores addr:interpolation ways
by Gerd Petermann 28 Aug '15

28 Aug '15

Hi all, see this thread : http://gis.19327.n5.nabble.com/Commit-r3632-fix-possible-message-internal-e… I think Anor Carlos pointed out that mkgmap should not ignore a more or less valid addr:interpolation way. This can happen when someone adds a POI with addr:housenumber and addr:street and doesn't connect the POI node to the addr:interpolation way. Or when someone adds the way without recognizing that the POI has a valid address. Up to now mkgmap prefers to ignore the complete addr:interpolation way when such a node is too far from the interpolated position and "too far" means 5 m, which is probably too restrictive. I guess this is likely to happen and I've changed the code to add a message to the log with r3634. Attached is a patch that changes the tolerated distance to 40 m and further improves logging. If I hear no complains I'll commit this patch on monday. @Anor: Please try the binary at http://files.mkgmap.org.uk/download/279/mkgmap.jar with enabled logging, e.g. uk.me.parabola.mkgmap.osmstyle.housenumber.level=INFO Gerd

1 1