February 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

address search and mixed index
by Gerd Petermann 04 Apr '15

04 Apr '15

Hi all, maybe you can help me with an open problem. I am not sure sure if this is a case of "garbage-in -> garbage-out" or not. I see a lot of houses with plausible tags addr:housenumber and addr:street which are close to a road that has no name or a different name. A typical example is a village like Stühren: http://www.openstreetmap.org/node/378300184#map=17/52.89235/8.71508 Almost all buildings are all tagged with addr:street="Stühren", but the main road through the village is the L340: http://www.openstreetmap.org/way/26701165 which has no name and many buildings are close to it. I think the postal address for these buildings is really something like "Stühren 28, 27211 Bassum,Germany" With trunk and r3477 in the housenumber2 branch I see rather bad results for this because the housenumbers are only added to those roads that have the name "Stühren". I wonder if the housenumber option should add the name "Stühren" to (a part of) L340 ? Or should we try to handle this with bounds containing villages and some style rules? Gerd

4 8

weired housenumbers in Canada
by Gerd Petermann 23 Mar '15

23 Mar '15

Hi all, in Canada, a lot of OSM data comes from imports, e.g. source=NRCan-CanVec-8.0 or NRCan-CanVec-10.0 I think the house number data provided by these imports is more or less garbage. I see many ways like this: http://www.openstreetmap.org/way/133338396 It is an addr:interpolation way connecting two different points which both have addr:housenumber=5 addr:street=20th Sideroad In many cases I see two of these meaningless objects, one with source=NRCan-CanVec-8.0, one with source=NRCan-CanVec-10.0. I think the only way to handle this data is to ignore it, if mkgmap finds an addr:interpolation way connecting two equal numbers, we should ignore the numbers. Right? Gerd

3 2

Why is there a code-page in the .typ-file? And - is there a code-page a gmapsupp.img?
by Felix Hartmann 20 Mar '15

20 Mar '15

Hi Everyone 1. Switching to unicode I ask myself If I should also change the typ-file to code-page=65001 - is this only for the case that you want to label a line/poi/polygon using unicode characters - or should any .typ-file that is "bundled" by a unicode map, also be unicode? Does anyone know a garmin unicode map that has a .typ-file? 2. I couldn't notice any change on compiling a gmapsupp.img regarding a code-page set. Is there actually a code-page for a gmapsupp.img? Especially regarding address index. Or is the code-page only for the map.img files themselves? Felix -- keep on biking and discovering new trails Felix openmtbmap.org & www.velomap.org

4 5

log messages for housenumbers
by GerdP 28 Feb '15

28 Feb '15

Hi all, I wonder what info mkgmap should report regarding housenumbers. The current log messages printed with uk.me.parabola.mkgmap.osmstyle.housenumber.level=INFO are probably not very useful. I think mkgmap should try to report those OSM elements which should be reviewed. The messages containing the string "looks wrong" are candidates. I also think that mkgmap should report OSM elements which cause problems in the algos used to calculate the img data. These elements might be 100% correct, but the messages may show problems in the algos. Any ideas? Gerd -- View this message in context: http://gis.19327.n5.nabble.com/log-messages-for-housenumbers-tp5834339.html Sent from the Mkgmap Development mailing list archive at Nabble.com.

3 29

Commit: r3487: Save memory by eliminating the MultiSortKey from Mdr7.preWriteImpl.
by svn commit 28 Feb '15

28 Feb '15

Version mkgmap-r3487 was committed by steve on Sat, 28 Feb 2015 Save memory by eliminating the MultiSortKey from Mdr7.preWriteImpl. The m.getPartialName() may have to be extended to name.substring(m.getNameOffset()).

1 0

Complex ocean polygon hiding the islands
by Alexandre Loss 28 Feb '15

28 Feb '15

Hi guys, Please see if any of you can help me ... I have an ocean polygon with multiple edges and in the middle of these there are several islands that are in fact holes in the polygon. Here's an example: [image: Imagem inline 1] Showing only the contour lines, the polygon is drawn as follows: [image: Imagem inline 2] Zooming... [image: Imagem inline 3] When I compile the map, the islands/holes are overlapped by the sea. I noticed that when the polygon is less complex, this overlapping phenomena doesn't happens. Its looks like mkgmap simplifies complex polygons. Is there any option that I can use in mkgmap to respect the layout of complex ocean polygons? By the way, if this isn't the right list to ask this kind of question, please let me know cause I'm new here. Thanks. Alexandre

2 2

FW: AW: RE: computing power mdx/mdr
by Gerd Petermann 28 Feb '15

28 Feb '15

Hi all, Arndt confirmed that his memory problem is solved with the Mdr7.patch . If I hear no complains I'll commit it to trunk tomorrow. Gerd Hi Gerd,funktioniert ☺ ich werde zukünftig mehrere Durchläufe mit mkgmap machen.Kacheln - mdx/mdr - gmapsupp -tdb/img (mit den Höhenlinien). Für Europa muss ich mal in richtige Hardware investieren. Vielen Dank für deine Hilfe!!! GrußArndt

4 12

"Initial value too large" with mkgmap-housenumbers2-r3477
by Stéphane MARTIN 27 Feb '15

27 Feb '15

Hi, I made a new map for France (including addresses from french BANO) with r3477 and I got this message : "HOUSE NUMBER RANGE: Initial value too large: 561709099" Anyway, the map seems OK. Regards. Steph

3 3