Hi Jakob,
the problem depends on the order of shapes which is more or less unpredictable unless you use --preserve-element-order. Anyway, the problem is caused by the ways 457279543 and two mp-relations https://www.openstreetmap.org/relation/6756850 and <https://www.openstreetmap.org/relation/> https://www.openstreetmap.org/relation/1945484 . The OSM data is not correct, but mkgmap should be able to handle that. Not sure what exactly the problem is. Gerd ________________________________ Von: mkgmap-dev <[email protected]> im Auftrag von Jakob Mühldorfer <[email protected]> Gesendet: Montag, 19. Dezember 2016 20:08:44 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Crash "Illegal Capacity" when converting map Thank god, it was so wonky, I was afraid that it might not work As I said, I have a map for the holidays on my device, no rush Am 19.12.2016 um 19:55 schrieb Gerd Petermann: Hi Jakob, I can reproduce the problem. I hope I can reduce it to a few files tomorrow. Gerd ________________________________ Von: mkgmap-dev <[email protected]><mailto:[email protected]> im Auftrag von Jakob Mühldorfer <[email protected]><mailto:[email protected]> Gesendet: Montag, 19. Dezember 2016 19:05:06 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Crash "Illegal Capacity" when converting map Yes, sorry for the europe-latest thing, but I have a crop polygon that includes many countries and sadly I could not narrow it down. I wondered if resolution and max nodes does not contradict a bit, thanks for letting me know it really does Will use a lower resolution Am 19.12.2016 um 18:58 schrieb Gerd Petermann: Hi Jakob, okay, that will take a while. Did not expect europe-latest as input file ;-) BTW: Why do you use resolution=15 in combination with max-nodes=2500000? This seems to require lots of memory and CPU without much benefit. Gerd ________________________________ Von: mkgmap-dev <[email protected]><mailto:[email protected]> im Auftrag von Jakob Mühldorfer <[email protected]><mailto:[email protected]> Gesendet: Montag, 19. Dezember 2016 18:43:24 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Crash "Illegal Capacity" when converting map Hi Gerd, you can find the workfolder and script here: https://www.jmuehldorfer.de/owncloud/index.php/s/yAKF6E47Y9pyxAd Sorry for the chaos in the options, the script was created piece by piece, they are not in logical order You also need to add a copy of bounds.zip and sea.zip into the preload folder, I did not include that because you likely already have them on your machine And you might need to adjust java ram settings in the script Hope you can trigger the bug, thanks for looking into this! Luckily one recent map converted just fine, and I don't need mkgmap again until end of January Jakob Am 19.12.2016 um 18:33 schrieb Gerd Petermann: Hi Jakob, that sound like we have a field which is not properly initiated [:(] Please post your scripts etc., maybe I can reproduce the problem. Gerd ________________________________ Von: mkgmap-dev <[email protected]><mailto:[email protected]> im Auftrag von Jakob Mühldorfer <[email protected]><mailto:[email protected]> Gesendet: Montag, 19. Dezember 2016 18:27:30 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Crash "Illegal Capacity" when converting map It is not that simple I do get an empty tile, the way you recommended, but when i manually start conversion of just that tile, it works and outputs a 3.5MB img file for this tile It is some kind of effect where you need to convert multiple tiles at once to trigger the bug Am 19.12.2016 um 16:31 schrieb Gerd Petermann: Hi Jakob, yes, send the scripts. You should be able to find the problem tile if you add option --keep-going to mkgmap. When it finishes you should see an empty *.img file, the corresponding input file is probably the cause. Gerd ________________________________ Von: mkgmap-dev <[email protected]><mailto:[email protected]> im Auftrag von Jakob Mühldorfer <[email protected]><mailto:[email protected]> Gesendet: Montag, 19. Dezember 2016 16:18:48 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Crash "Illegal Capacity" when converting map Hi Gerd, I am trying to reproduce it, but not getting very far Different splitter parameters, that did nothing Deleting 2 tiles that need to be converted, still the same, delete just one other tile, and the bug was gone. Unfortunately I am leaving this PC over christmas and can't try much more, can I send you the folder with the script used to convert the maps maybe? Jakob Am 19.12.2016 um 06:41 schrieb Gerd Petermann: Hi Jakob, this seems to be a special case in the (OSM) data. I can probably add a check to avoid the crash but I'd like to know what happens and I was not yet able to reproduce it. I assume it happens only for one input file? Please post a link to that file and maybe your style and options if needed. Gerd ________________________________ Von: mkgmap-dev <[email protected]><mailto:[email protected]> im Auftrag von Jakob Mühldorfer <[email protected]><mailto:[email protected]> Gesendet: Montag, 19. Dezember 2016 02:17:20 An: [email protected]<mailto:[email protected]> Betreff: [mkgmap-dev] Crash "Illegal Capacity" when converting map Hi, currently I am getting a ton of crashes like this: > java.lang.IllegalArgumentException: Illegal Capacity: -1 > at java.util.ArrayList.<init>(ArrayList.java:156) > at > uk.me.parabola.mkgmap.filters.ShapeMergeFilter.mergeLongestSequence(ShapeMergeFilter.java:363) > at > uk.me.parabola.mkgmap.filters.ShapeMergeFilter.tryMerge(ShapeMergeFilter.java:226) > at > uk.me.parabola.mkgmap.filters.ShapeMergeFilter.addWithConnectedHoles(ShapeMergeFilter.java:167) > at > uk.me.parabola.mkgmap.filters.ShapeMergeFilter.merge(ShapeMergeFilter.java:98) > at > uk.me.parabola.mkgmap.build.MapBuilder.processShapes(MapBuilder.java:1124) > at > uk.me.parabola.mkgmap.build.MapBuilder.makeSubdivision(MapBuilder.java:791) > at > uk.me.parabola.mkgmap.build.MapBuilder.makeMapAreas(MapBuilder.java:701) > at uk.me.parabola.mkgmap.build.MapBuilder.makeMap(MapBuilder.java:232) > at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:107) > at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:69) > at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:256) > at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:252) > at java.util.concurrent.FutureTask.run(FutureTask.java:266) > at > java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142) > at > java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) > at java.lang.Thread.run(Thread.java:745) > Exiting - if you want to carry on regardless, use the --keep-going option Any idea what might be wrong? It happened in 3706 to 3730 Sometimes the exact same conversion process (splitting+mkgmap) on the exact same OSM file works, but that is rarely the case If you need any more information, or an upload of the style and split tiles that cause the crash, let me know. Hope you can help Jakob _______________________________________________ mkgmap-dev mailing list [email protected]<mailto:[email protected]> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list [email protected]<mailto:[email protected]> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list [email protected]<mailto:[email protected]> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list [email protected]<mailto:[email protected]> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list [email protected]<mailto:[email protected]> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list [email protected]<mailto:[email protected]> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________ mkgmap-dev mailing list [email protected] http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
