December 2012 - mkgmap-dev - The mkgmap lists

splitter r269
by GerdP 25 Dec '12

25 Dec '12

Hi all, I've committed r269 in the problem-list branch. It contains the changes proposed by WanMil and a few optimizations regarding throughput and memory needs. It is now possible to split europe on a 4GB machine with 64 bit JVM with --keep-complete and --max-areas=2048, no matter what output format you use. I've changed the default "batch_limit" in the pbf writer from 4000 to 1000. This increases the output file size by 2-3%, but it reduces the heap requirements with output=pbf drastically if you have many areas. The limit of 4000 was ok when splitter wasn't able to write more than 100 files at a time. Gerd -- View this message in context: http://gis.19327.n5.nabble.com/splitter-r269-tp5741292.html Sent from the Mkgmap Development mailing list archive at Nabble.com.

2 2

New exit_hints branch
by WanMil 25 Dec '12

25 Dec '12

Hi, I have created a new exit_hints branch which may adds a new interesting workaround feature. When driving on a motorway my Garmin does not give me a hint on which exit I should leave. It just tells me the name of the street connected to the motorway_link but this isn't helpful because this name isn't shown on any sign on the motorway. It would be more helpful if the Garmin tells me the number and the name of the exit. That's what I am trying in the exit_hints branch. How is that achieved? It is activated if the option --process-destination is set. There is an exit node: highway=motorway_junction ref=53 name=Unna-Ost A motoway_link starts at this exit node. highway=motorway_link ... some more tags ... E================= The motorway_link is now cut into three parts: E11122233333333333 Part 1 and 3 are tagged with the same tags of the original motorway_link. Part 2 is tagged additionally with mkgmap:exit_hint=true mkgmap:exit_hint_ref=53 mkgmap:exit_hint_name=Unna-Ost In the style file there are two additional rules in the lines file: # Name the hint part with the name and the ref of the exit highway=motorway_link & mkgmap:exit_hint=true { delete display_name; name 'Exit ${mkgmap:exit_hint_ref} ${mkgmap:exit_hint_name}' | 'Exit ${mkgmap:exit_hint_name}' } # Use the motorway code for the hint part so that the GPS displays the name as hint where to leave the motorway highway=motorway_link & mkgmap:exit_hint=true { add oneway = yes; add bicycle = no; add foot = no; } [0x01 road_class=3 road_speed=2 resolution 20] Please try and comment if you find it useful or if you have ideas how to realize it in a better way! It is not yet complete: If a motoway_link uses reverse oneway direction the code may fail? Have fun! WanMil

5 16

Obsolete(?) road naming using POI names
by WanMil 25 Dec '12

25 Dec '12

Hi, while crawling through the mkgmap sources I found another mystic tagging which I think come from the time before the style system was so powerful. Look at commit r1267: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&path=%2F&rev=12… If the first point of a road (example highway=secondary) has a node that is tagged with secondary_name this value is taken as name for the road. Is this still used? Otherwise I want to remove that. WanMil

1 1

splitter with new fastutil library
by GerdP 25 Dec '12

25 Dec '12

Hi all, I've committed r272 in the problem-list branch. Please note that you need the new fastutil.jar that is contained in the trunk version (splitter-r270) Major changes: -keep-complete is now the default -loops in relations are reported again (this is just informational) -improvements in memory requirements for the keep-complete option -improved parameter checking: show which parameters are interpreted as file names, stop immediately if a given file name is not readable. next on my todo list: add a --precomp-sea parameter to take into account when mkgmap adds nodes from precompiled sea data Gerd -- View this message in context: http://gis.19327.n5.nabble.com/splitter-with-new-fastutil-library-tp5741668… Sent from the Mkgmap Development mailing list archive at Nabble.com.

1 0

input file has a different sort order
by Minko 24 Dec '12

24 Dec '12

When I'm trying to combine two img tiles from different maps mkgmap gives this warning: input file ...img has a different sort order Both maps are created with --code-page=1252 and --latin1 compiled with mkgmap-r2379. I even get this warning when combining two (or more) img tiles from the same mapset! This issue is only when I use gmapsupp in my args file and only when processing img's (not with osm.pbf files) and without gmapsupp it is ok. Consequence is that the address index is not readable on the gps device (on Mapsource it's fine). With gmaptool I notice the s-f value is 1, normally this is 2. See the attachment for more headers. My mkgmap args file is as follows: family-name=OFM_DBNL family-id=4931 product-id=1 overview-mapname=mapset draw-priority=20 tdbfile latin1 code-page=1252 show-profiles=1 net route index gmapsupp The last mkgmap version that didn't have this issue was r2328 What is going wrong here?

2 9

Re: [mkgmap-dev] Strange routing directions
by Francisco Moraes 24 Dec '12

24 Dec '12

Any ideas? Francisco On 12/20/2012 11:15 AM, Francisco Moraes wrote: > Hi, > > I've noticed several times this but never remembered to email about > it. My GPS always tells me to turn right on Park Village Dr when I am > driving the full extent of it > (http://www.openstreetmap.org/browse/way/18912622) There seems to be > something near the intersection with Trents Wood Way where it always > says turn right into PV Dr, even when coming from PV Dr itself. I've > noticed similar instructions before, for example when travelling south > on Miami Blvd into

4 4

feature request for style-functions
by Henning Scholland 24 Dec '12

24 Dec '12

Hi, in OSM-data there are several objects having something like disused=yes or abandonned=yes. It's pretty hard and confusing to add disused!=yes and so on to each object in style-file. So I thought about another possibility toget rid of these objects. It would be very easy if it would be possible to stop style-processing of actual object, if it matches one rule. E.g. disused=yes { stop-processing() } Would something like this be useful at all, or am I the only one? Or is there already a workaround? Henning

3 3

Problem with setting output for splitter
by Jaromír Mikeš 24 Dec '12

24 Dec '12

Hello, I just noticed that I can't set pbf or xml output for splitter ... I getting this message: --keep-complete is not supported for multiple input files. Please execute splitter once for each file. But when I run the same without output specified it is set to pbf automatically and everything goes fine. splitter-r264 regards mira $ java -Xmx6144m -jar splitter-r264/splitter.jar --max-nodes=1000000 --description=Finland --mapid=63250001 --keep-complete=true output=pbf --output-dir=CONTOUR srtm.osm Splitter version 264 compiled 2012-12-20T01:45:03+0000 cache= description=Finland geonames-file= keep-complete=true mapid=63250001 max-areas=512 max-nodes=1000000 max-threads=2 (auto) mixed=false no-trim=false output=pbf output-dir=CONTOUR overlap=-1 polygon-file= problem-file= problem-report= resolution=13 split-file= status-freq=120 stop-after=dist write-kml= --keep-complete is not supported for multiple input files. Please execute splitter once for each file.

2 3

new tile splitter released: r263
by GerdP 22 Dec '12

22 Dec '12

Hi all, I decided to merge the problem-list branch back to trunk, so r263 is now a release version. @Steve: Attached is an updated file to replace http://www.mkgmap.org.uk/page/tile-splitter Please check, I am a beginner with html and my editor complains about line 351, whereas html tidy complais about other lines as well. tile-splitter.htm <http://gis.19327.n5.nabble.com/file/n5740649/tile-splitter.htm> I found no page on http://www.mkgmap.org.uk/ that links to http://www.mkgmap.org.uk/page/tile-splitter If that is intended, please let me know where I should place the updated docu. Ciao, Gerd -- View this message in context: http://gis.19327.n5.nabble.com/new-tile-splitter-released-r263-tp5740649.ht… Sent from the Mkgmap Development mailing list archive at Nabble.com.

6 10

location style obsolete?
by WanMil 22 Dec '12

22 Dec '12

The location style contains rules for address mapping that have now been replaced by the new addr style which is included by the new style inclusion. Any objections againg removing that? WanMil

3 5