January 2012 - mkgmap-dev - The mkgmap lists

Address search: feature request
by Carlos Dávila 12 Jan '12

12 Jan '12

First of all, a big thank you Steve and all how have contributed to get address search working on mkgmap generated gmapsupp. I think it's a big improvement. In a further step, it would be great to be able to search for partial street names, so that for languages in which street names begin with a word like Street/Avenue/whatever you don't need to type all it. Apart of the annoyance of having to type those extra letters with the minimalistic keyboard of the device, there's the problem that many times you don't know how exactly the name is written in the map. For example in Spain you can have the same street written as Calle foo, Calle of foo, C/ foo, C. foo, ... Abbreviated forms should be fixed in the data, but in many cases you can see Calle foo and Calle of foo in the signs on the street, so they both are equally valid. About one year ago, Johann Gail gave some information about how it could be achieved. See http://www.mkgmap.org.uk/pipermail/mkgmap-dev/2011q1/009922.html If I can help in any way to get this feature, let me know.

1 0

Help/ideas with address index
by Tomas Straupis 12 Jan '12

12 Jan '12

Hello I'm trying to generate address index, but so far am unable to do that. Maybe you can give me some advices on what should I try? I expect to get index right in the gmapsupp.img as I'm using linux and MapSource is unable to transfer gmapsupp to my Colorado (I'm simply copying gmapsupp.img to the device via usb cabel). I was following instructions I've found on osm wiki. Created bounds info (tried both creating my own using osmfilter/osmconvert/mkgmap as well as downloading pre-created europe bounds 20111202). This is how I run mkgmap: java -jar mkgmap.jar \ --style-file=/home/tomas/garmin/tomas_style \ --gmapsupp \ --tdbfile \ --family-id=1 \ --family-name=OpenStreetMap \ --description=Lietuva`date +%Y%m%d` \ --country-name=Lithuania \ --country-abbr=LTU \ --preserve-element-order \ --route \ --add-pois-to-areas \ --generate-sea=multipolygon \ --coastlinefile=lkrantas.osm \ --report-undefined-nodes \ --check-roundabouts \ --drive-on-right \ --report-dead-ends \ --index \ --location-autofill=bounds \ lithuania_.osm The only error messages I get are about short arcs. This is how I test. Upload the map to my device. Choose "where to" -> "address", press "done" (or enter letter "L") and get an empty list... Note: I'm not using splitter, just taking data from geofabrik and making a map from it directly. mkgmap version 2165. Thank you P.S. Everything else (routing, styles) seems to work ok. -- Tomas Straupis

5 9

Bug in LocationHook?
by GerdP 12 Jan '12

12 Jan '12

WanMil, LocationHook checks these hard coded tag names: private static final String[] LEVEL2_NAMES = new String[]{"name","name:en","int_name"}; but it doesn't add them in getUsedTags() as long as the user doesn't set them in parameter name-tag-list. Is this intended? Ciao, Gerd -- View this message in context: http://gis.638310.n2.nabble.com/Bug-in-LocationHook-tp7157897p7157897.html Sent from the Mkgmap Development mailing list archive at Nabble.com.

4 32

Routing issue with split road
by Francisco Moraes 11 Jan '12

11 Jan '12

Hi, I have seen some strange routing issues with the way below. If I am driving south, my GPS wants me to make a U-turn where the divided highway ends. It seems like the ways are not connected but I checked and I can't find anything wrong in the OSM data. http://www.openstreetmap.org/browse/way/37649065 Any ideas? Francisco

4 9

[PATCH v1] Remove shrinks quadtree
by WanMil 11 Jan '12

11 Jan '12

Hi, I implemented a shrinking remove in the elementquadtree which is used by the LocationHook. If enough elements have been removed the depth of the quadtree is shrinked. Sometimes this improves the performance, sometimes not. The patch seems to be not 100% correct because I do get different numbers of query results from the quadtree (patched finds 24974922 in 66 tiles, unpatched finds 24974934 elements in 66 tiles). WanMil

2 1

LocationHook and tagging ways
by GerdP 11 Jan '12

11 Jan '12

Hi list, as discussed before, the current implementation of LocationHook produces somehow unpredictable results when it locates ways that cross boundaries. A correct solution would be to split the way, but according to WanMil this would be very slow. So, with bad luck, we may find a way with 100 points in a boundary that contains only one of them. While thinking about a test for my optimized version I wondered why we put all points of a way into the quadtree, and not only one (e.g. the 1st, the last, or the one in the middle). Using only one point makes the quadtree much smaller and faster, and the result is (almost) as correct as before. I see only one problem: If the way crosses boundaries, one of the boundaries may contain better information (more complete) than the other(s). If we chose the "wrong" point for searching, we may find only the incomplete information. Gerd -- View this message in context: http://gis.638310.n2.nabble.com/LocationHook-and-tagging-ways-tp7167931p716… Sent from the Mkgmap Development mailing list archive at Nabble.com.

3 2

Commit: r2166: Improve logging in LocationHook
by svn commit 10 Jan '12

10 Jan '12

Version 2166 was commited by wanmil on 2012-01-10 21:33:39 +0000 (Tue, 10 Jan 2012) Improve logging in LocationHook --GerdP

1 0

Convert from gmapsupp to mapsource map or the other way around?
by Manuel Reimer 10 Jan '12

10 Jan '12

Hello, I want to use the latest trunk version to create a gmapsupp.img file with address search. If I want to have both, a "garmin map" for mapsource/qlandkarte and a image file for the GPS device, is it possible to do this in two steps? Create one of the two first and then convert to the second format? Yours Manuel

2 1

[patch] Small error in logging
by GerdP 09 Jan '12

09 Jan '12

Hi, with enabled logging, LocationHook reports multiple times the same tag as added. The attached patch fixes this. http://gis.638310.n2.nabble.com/file/n7167846/LocationHook_logging.patch LocationHook_logging.patch Gerd -- View this message in context: http://gis.638310.n2.nabble.com/patch-Small-error-in-logging-tp7167846p7167… Sent from the Mkgmap Development mailing list archive at Nabble.com.

1 0

Base style sub-styles
by Bennie du Plessis 08 Jan '12

08 Jan '12

Marko (mainly), How does the sub styles idea work.? In commit 1209, where is the sub-styles? at \examples\styles I can see the default and noname styles. In the info file of the default style I can see base-style=contours_ft base-style=location base-style=waters Is this the sub-styles that will be included in the default style? Where are these sub-styles though? Suppose an included base-style treats an element a certain way, will the main style (default in this case) then ignore this element? So the base-style treatment overrides the main style? Will a continue statement in the base style then reverse this, so that the main style treatment will prevail? Then, in the case of 3 sub styles, if they all have actions for the same element: the first style will be used, and the rest ignored unless a continue statement is used in the previous style? A continue statement will have no effect if the previous sub style did not have a continue statement? Or should conflicting rules be avoided? Argh, I hope someone understand what I mean? I don't use the default style, but a style based on the default with a couple of tweaks for custom points & lines etc. Still I like to keep updating regularly to benefit from Marko's improvement in the style. So I copy my tweaks into the latest default styles manually. The sub style idea might make my manual operation unnecessary, if I can just include my own style into the default. Before any other included styles, if I understand it correctly. Bennie

4 13