[mkgmap-dev] Commit r4292: Fix problem with long roads producing the error message "internal error? The nodeCount doesn't match value calculated by RoadNetWork"
From Gerd Petermann gpetermann_muenchen at hotmail.com on Wed Oct 2 10:18:59 BST 2019
Hi IMG experts, the message is produced because we have to different ways to calculate a counter which is written to NET. It is not obvious which one is correct and therefore we have a check that compares the counters. Some special cases only occur with polish (*.mp) input, others only with osm data. I've now started a new branch called "NET-no-NOD" to improve the code so that it - supports writing NET without NOD data from polish input - doesn't write redundant data into RGN when neither housenumbers nor NOD data is present for a road - ignores empty POI instead of crashing (special case for John Thorne) This is complex because in OSM data each point (Node) has a unique ID while in *.mp this is not the case. On the other hand with *.mp input the user can produce roads which have only one routable node in the middle (+-shaped crossing). I think the branch already works better than trunk but testing is complicated. I test these fixed options --index --preserve-element-order --bounds=f:\osm\bounds.zip in combination with 1) --net 2) --route 3) --net --housenumbers 4) --route --housenumbers both with various *.osm and *.mp data. Whereever I find differences in the IMG output I try to find out if the new result is an improvement or not. I use Mapsource to test effects on routing and the display tool to analyse the differences in the RGN, NET and NOD files... Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag von Gerd Petermann <gpetermann_muenchen at hotmail.com> Gesendet: Samstag, 28. September 2019 08:33 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Commit r4292: Fix problem with long roads producing the error message "internal error? The nodeCount doesn't match value calculated by RoadNetWork" Hi Ticker, thanks for the details, I've reverted the problemtatic patch for now, so r4294 should work fine for most users (except John Thorne) Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap at jagit.co.uk> Gesendet: Samstag, 28. September 2019 08:22 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Commit r4292: Fix problem with long roads producing the error message "internal error? The nodeCount doesn't match value calculated by RoadNetWork" Hi I've minimised the options and still get the problem. Downloading this morning from: http://download.geofabrik.de/europe/great-britain/wales.html java -Xmx1540M -ea -jar splitter-r592/splitter.jar --mapid=74320089 - -split-file=tstSep19.list wales-latest.osm.pbf >splitter.log 2>&1 where tstSep19.list contains: 4320089: 2398208,-194560 to 2437120,-155648 # : 51.459961,-4.174805 to 52.294922,-3.339844 Then, with latest bounds downloaded from the link on mkgmap.org.uk java -jar mkgmap-r4293/mkgmap.jar --route --bounds=bounds.zip --input -file=74320089.osm.pbf I get: Time started: Sat Sep 28 07:03:00 BST 2019 SEVERE (RoadDef): 74320089.osm.pbf: internal error? The nodeCount doesn't match value calculated by RoadNetWork: ( http://www.openstreetmap.org/browse/way/26552046) SEVERE (RoadDef): 74320089.osm.pbf: internal error? The nodeCount doesn't match value calculated by RoadNetWork: ( http://www.openstreetmap.org/browse/way/124109672) SEVERE (RoadDef): 74320089.osm.pbf: internal error? The nodeCount doesn't match value calculated by RoadNetWork: ( http://www.openstreetmap.org/browse/way/139314867) SEVERE (RoadDef): 74320089.osm.pbf: internal error? The nodeCount doesn't match value calculated by RoadNetWork: ( http://www.openstreetmap.org/browse/way/24556428) Number of MapFailedExceptions: 0 Number of ExitExceptions: 0 Time finished: Sat Sep 28 07:03:44 BST 2019 Total time taken: 44 seconds Ticker On Fri, 2019-09-27 at 13:48 +0100, Ticker Berkin wrote: > Hi Gerd > > I've just updated my build from 4288 to 4293 and I'm now getting > quite > a few of these errors > > SEVERE (RoadDef): 74220089.osm.pbf: internal error? The nodeCount > doesn't match value calculated by RoadNetWork: ( > http://www.openstreetmap.org/browse/way/26552046) > SEVERE (RoadDef): 74220089.osm.pbf: internal error? The nodeCount > doesn't match value calculated by RoadNetWork: ( > http://www.openstreetmap.org/browse/way/40038526) > > My options include: > --no-housenumbers > --link-pois-to-ways > --add-boundary-nodes-at-admin-boundaries=0 > --style-file=../styles/mystyle > and many others > > I'll try to reproduce with simpler options and the default style > > Regards > Ticker > > > On Wed, 2019-09-25 at 08:10 +0100, svn commit wrote: > > Version mkgmap-r4292 was committed by gerd on Wed, 25 Sep 2019 > > > > Fix problem with long roads producing the error message "internal > > error? The nodeCount doesn't match value calculated by RoadNetWork" > > Problem was introduced with r4291 > > > > http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=429 > > 2 > > _______________________________________________ > > mkgmap-dev mailing list > > mkgmap-dev at lists.mkgmap.org.uk > > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev > _______________________________________________ > mkgmap-dev mailing list > mkgmap-dev at lists.mkgmap.org.uk > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev at lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev at lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
- Previous message: [mkgmap-dev] Commit r4295: avoid conversion to Integer, no funtional change
- Next message: [mkgmap-dev] mkgmap-dev Digest, Vol 135, Issue 1
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the mkgmap-dev mailing list