September 2013 - mkgmap-dev - The mkgmap lists

mergeroads branch
by WanMil 05 Oct '13

05 Oct '13

Hi, the mergeroads branch should be stable now so I invite all to test it. Here are the changes compared to trunk: * Roads now become merged before the routing network is created. This might improve routing on long distances (I have seen some). The number of roads in the routing network is reduced by 10-20% (in germany). * refs are now composed in the style sheet and not in java code. The new tag mkgmap:ref is used for that (see inc/refs in default style). * access restrictions are now also handled in the style sheet and not in java code. There are some new tags for that: mkgmap:access:emergency mkgmap:access:delivery mkgmap:access:car mkgmap:access:bus mkgmap:access:taxi mkgmap:access:foot mkgmap:access:bike mkgmap:access:truck mkgmap:access:carpool If a tag is set to 'no' the given access type is restricted in the map. All other values are mapped to 'yes'. I have tried to convert the java code to the style sheet. This was quite complicated but I hope it's rather correct. See new include file inc/access. * Instead of using hardcoded rules for maxspeed it is now possible to control that via style file. Set the mkgmap:road-speed-class tag to the desired class value (0 to 7) and use the new functions maxspeedkmh() and maxspeedmph() to parse the maxspeed tag. * Two new actions that seems useful to me: echotags 'Hello' prints out the OSM element id plus all tags of the element plus the text. This seems to be useful for style debugging. deletealltags removes all tags from an element and therefore stops its processing. I would be happy to get a lot of feedback! Have fun! WanMil

7 44

highway count not fixed yet... - merge-roads-branch
by Felix Hartmann 03 Oct '13

03 Oct '13

Just cannot find the topic on the merge-roads-branch. Is it known that the highway count error is not fully fixed yet? I still get loads of them.

5 31

Commit: r2724: Make sure that remove-short-arcs with no argument is evaluated with the default value (5.0)
by svn commit 03 Oct '13

03 Oct '13

Version 2724 was committed by gerd on Fri, 27 Sep 2013 Make sure that remove-short-arcs with no argument is evaluated with the default value (5.0)

3 4

Provincial capitals
by Bernhard Hiller 30 Sep '13

30 Sep '13

How can provincial capitals be detected by a style rule? I want to show them at low zoom level, even if their population is rather little. Example: Phangnga town (South Thailand) has a population of some 10,000 people, but it is the capital of Phangnga province (see http://www.openstreetmap.org/browse/node/984029824) There is a relation for the province boundary: http://www.openstreetmap.org/browse/relation/1908799 which has Phangnga set as its admin_centre; admin_level = 4. I tried in the points file: place=* & mkgmap:admin_level4=* & cityxx!=yes {set cityxx=yes} [0x0100 level 6 continue with_actions ] but that does not help: Phangnga city is shown at a high zoom level only, instead of at the same zoom level with Bangkok (ok, that's not useful, I do so temporarily to find out how things work). I tried place=city or place=village instead of place=*, same result. Only place=* & is_capital=province & cityxx!=yes {set cityxx=yes} [0x0100 level 6 continue with_actions ] works as expected. Fortunately, a mapper added the is_capital tag to all Thai province capitals just a few days ago... But I think using the relation is a better option. Only: how can I do that? Tried that with: mkgmap 2724, Java 7 u40 (64bit), Windows 7 (64bit) Thanks for your hints.

1 1

[PATCH v1] Merge road network
by WanMil 29 Sep '13

29 Sep '13

mkgmap has a LineMergeFilter but that does not merge the road network. Attached patch performs a merging in the OSM data just before the road network is created. This has the advantage that the source code for the merging is quite easy. Turn restrictions and through routes are considered. I tested some data in germany and got a reduction of 5% to 12% (number of roads). I did not find any improvements but also no downgrades in routing. But I did not test long distance routing (> 600km). I think the merge results can be improved if the access handling and some other parts are moved from source code to the style files. This makes it easier for the merger to say that two roads have the same attributes and can be merged without problems. But that's a task for the future... I have already uploaded a compiled version: http://files.mkgmap.org.uk/detail/148 For anyone who is interested in the code: Please check the tag list (variable flatCompareTags) in the new RoadMerger. This taglist gives all tags that must be equal so that two roads can be merged. Is this list complete? Can I remove some tags from it? Have fun! WanMil

6 14

frig-roundabouts doesn't work
by GerdP 28 Sep '13

28 Sep '13

Hi, The option is described in the wiki, and I find a peace of source in the HighwayHook that evaluates it, but it doesn't work. What is missing? Gerd -- View this message in context: http://gis.19327.n5.nabble.com/frig-roundabouts-doesn-t-work-tp5748545.html Sent from the Mkgmap Development mailing list archive at Nabble.com.

2 3

Commit: r2714: Make java 1.7 mandatory and refuse to run on earlier versions.
by svn commit 27 Sep '13

27 Sep '13

Version 2714 was committed by steve on Thu, 19 Sep 2013 Make java 1.7 mandatory and refuse to run on earlier versions.

5 10

merge-roads branch verry bugged...
by Felix Hartmann 26 Sep '13

26 Sep '13

I needed to revert to version 2715. ...2716 is the version which really breaks on my styles. There are for sure quite a lot serious bugs in 2716 and later... I'm sorry that I cannot relly help on tracking down the bugs - I'm currently in China, and haven't got much time, nor anything that we would consider usable internet in Europe (cannot reach my own SVN for example and lots of other things blocked - VPNs working unrealiably for productive tasks, and speed is max 256Kbit up, 2-3Mbit down) -- on the up to date - merge-roads rev 2722 many of my maps fail to compile - the usual is like countless errors due to identical road id (and rev 2721 really made this much worse. With 2717 I only get like 20 for Switzerland instead of uncountable), then (not for all countries) java.lang.assertionError: SEVERE (StyledConverter): c:\openmtbmap\maps\63710216.osm.pbf: Road with identical id: 238840783 ..... uncountable amount of, then: java.lang.AssertionError at uk.me.parabola.mkgmap.osmstyle.housenumber.HousenumberGenerator.match(HousenumberGenerator.java:322) at uk.me.parabola.mkgmap.osmstyle.housenumber.HousenumberGenerator.generate(HousenumberGenerator.java:195) at uk.me.parabola.mkgmap.osmstyle.StyledConverter.end(StyledConverter.java:389) at uk.me.parabola.mkgmap.reader.osm.ElementSaver.convert(ElementSaver.java:244) at uk.me.parabola.mkgmap.reader.osm.bin.OsmBinMapDataSource.load(OsmBinMapDataSource.java:67) at uk.me.parabola.mkgmap.reader.osm.OsmMapDataSource.load(OsmMapDataSource.java:127) at uk.me.parabola.mkgmap.main.MapMaker.loadFromFile(MapMaker.java:167) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:63) at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:243) at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:239) at java.util.concurrent.FutureTask$Sync.innerRun(Unknown Source) at java.util.concurrent.FutureTask.run(Unknown Source) at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source) at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source) at java.lang.Thread.run(Unknown Source) java.lang.AssertionError at uk.me.parabola.mkgmap.osmstyle.housenumber.HousenumberGenerator.match(HousenumberGenerator.java:322) at uk.me.parabola.mkgmap.osmstyle.housenumber.HousenumberGenerator.generate(HousenumberGenerator.java:195) at uk.me.parabola.mkgmap.osmstyle.StyledConverter.end(StyledConverter.java:389) at uk.me.parabola.mkgmap.reader.osm.ElementSaver.convert(ElementSaver.java:244) at uk.me.parabola.mkgmap.reader.osm.bin.OsmBinMapDataSource.load(OsmBinMapDataSource.java:67) at uk.me.parabola.mkgmap.reader.osm.OsmMapDataSource.load(OsmMapDataSource.java:127) at uk.me.parabola.mkgmap.main.MapMaker.loadFromFile(MapMaker.java:167) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:63) at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:243) at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:239) at java.util.concurrent.FutureTask$Sync.innerRun(Unknown Source) at java.util.concurrent.FutureTask.run(Unknown Source) at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source) at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source) at java.lang.Thread.run(Unknown Source) java.lang.AssertionError at uk.me.parabola.mkgmap.osmstyle.housenumber.HousenumberGenerator.match(HousenumberGenerator.java:322) at uk.me.parabola.mkgmap.osmstyle.housenumber.HousenumberGenerator.generate(HousenumberGenerator.java:195) at uk.me.parabola.mkgmap.osmstyle.StyledConverter.end(StyledConverter.java:389) at uk.me.parabola.mkgmap.reader.osm.ElementSaver.convert(ElementSaver.java:244) at uk.me.parabola.mkgmap.reader.osm.bin.OsmBinMapDataSource.load(OsmBinMapDataSource.java:67) at uk.me.parabola.mkgmap.reader.osm.OsmMapDataSource.load(OsmMapDataSource.java:127) at uk.me.parabola.mkgmap.main.MapMaker.loadFromFile(MapMaker.java:167) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:63) at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:243) at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:239) at java.util.concurrent.FutureTask$Sync.innerRun(Unknown Source) at java.util.concurrent.FutureTask.run(Unknown Source) at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source) at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source) at java.lang.Thread.run(Unknown Source)

4 4

Style processor skips rule with tags added in style
by WanMil 25 Sep '13

25 Sep '13

Hi, I observed an unexpected behaviour of the style processor while checking the new access rules in the merge_roads branch. Assume the following way: highway=path And the lines file: highway=path {add test=yes} test=* [0x8] highway=path [0x9] I would expect that a line with garmin type 0x8 is created. But the 2nd rule is not considered and therefore a line with garmin type 0x9 is created. @Steve: Can you have a look at it? Attached you find the described test case. The merge_roads branch makes extended usage of such a pattern (see the inc/access file). Thanks! WanMil

2 2

special cases with short arcs
by GerdP 25 Sep '13

25 Sep '13

Hi, I am trying to find a better algo for the handling of short arcs. (Garmins routing algo refuses to route over short arcs) The typical case of a short arc is a crossing like here: http://www.openstreetmap.org/browse/node/1449682318 We don't lose much information if we merge the node with node 1303933872, so merging is a good idea here. On the other hand I stumbled across a few cases were I am not sure what to do. I many cases I'd prefer to move one of the two nodes so that the arc gets longer. See e.g. http://www.openstreetmap.org/browse/way/47138169 The first two points of the way build a short arc (2.8 m), while the next arc to point 267185122 is > 33m. The merging of these two nodes will produce a map which implies that you can turn right directly at the end of the bridge http://www.openstreetmap.org/browse/way/47138168 I think it would be better to move the node 627157116 a little bit closer to the next point 267185122. A similar case: The bridge at http://www.openstreetmap.org/browse/way/131860005 has a length of ~ 4.99 m. With the default value 5m, this is considered to be a short arc and short-arc-removal will remove the bridge by merging the two end points. On the other hand, a (service) road with just two points, one point connected to a road, the other not connected to a road should be removed if it is a short arc instead of eventually moving the point connected to the road (as this produces the zig zag lines as posted by popej (road.png): http://gis.19327.n5.nabble.com/Distorted-lines-tp5775761.html (this one is easy to detect and I can create a simple patch) I don't know yet how to implement all this, so I'd like to make sure that I am not investing much time into a wrong solution. Comments? Gerd -- View this message in context: http://gis.19327.n5.nabble.com/special-cases-with-short-arcs-tp5778708.html Sent from the Mkgmap Development mailing list archive at Nabble.com.

5 10