Bad tile around East London / Greenwich

Hi all, since some time my map has a bad tile around East London (Greenwich). I´ve checked earlier revisions of splitter and mkgmap, and decreased splitter´s max-nodes - without success. Problem seems to be in my style, which is based on OpenFietsMap. With the original OFM-style the error is rendered as well, but not with the mkgmap default style. But I´ve got no clue where to look for the mistake - anyone has got a hint for me? https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg Jan

Hi Jan, Any error message from mkgmap? I can try to reproduce if you upload the input osm file to https://files.mkgmap.org.uk/ (or post another download link) and the options that you use. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von jan meisters <jan_m23@gmx.net> Gesendet: Freitag, 6. August 2021 19:35 An: Development list for mkgmap Betreff: [mkgmap-dev] Bad tile around East London / Greenwich Hi all, since some time my map has a bad tile around East London (Greenwich). I´ve checked earlier revisions of splitter and mkgmap, and decreased splitter´s max-nodes - without success. Problem seems to be in my style, which is based on OpenFietsMap. With the original OFM-style the error is rendered as well, but not with the mkgmap default style. But I´ve got no clue where to look for the mistake - anyone has got a hint for me? https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg Jan _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Gerd, yes, but only on my own style: SCHWERWIEGEND (global): Bad file format: 10700105.osm.pbf File uploaded here <https://files.mkgmap.org.uk/download/517/10700105.osm.pbf>. It is splitter-generated from a geofabrik great-britain-latest with max-node=1400000. Splitter 615 and mkgmap 4802. For my own and and the default I have the same splitter options: java -Xmx26000m -jar splitter.jar --output=pbf --output-dir=/great_britain-splitter --max-nodes=1400000 --mapid=10750001 --geonames-file=PRECOMP/cities15000.txt --polygon-file=ARGS/great_britain.poly DATA/great_britain-latest.osm.pbf And mkgmap then: java -Xmx26000m -jar mkgmap.jar --output-dir=great_britain-output -c ARGS/great_britain-HGT.args -c great_britain-splitter/template.args ARGS/10000.txt The mkgmap ARGS for my style are: family-id: 10700 product-id: 1 draw-priority: 30 series-name: Great_Britain-HGT family-name: Great_Britain-HGT description: Great_Britain-HGT copyright-message: openstreetmap.org | earthexplorer.usgs.gov | viewfinderpanoramas.org | openfietsmap.nl area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: Styles overview-mapname: 10700000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip location-autofill: is_in,nearest housenumbers split-name-index tdbfile latin1 name-tag-list: name,name:de,int_name,place_name code-page: 1252 dem: HGT_NASADEM_Europe dem-dists: 3314,4971,6628,13256,29826 overview-dem-dist=89478 show-profiles: 1 merge-lines road-name-config=examples/roadNameConfig.txt add-pois-to-areas pois-to-areas-placement add-pois-to-lines link-pois-to-ways poi-excl-index=0x3200-0x331f nearby-poi-rules=*/named:5,*/unnamed:10,0x2b07:50,0x2f00-0x2f03:50,0x2f04-0x2f09:25,0x2f0a-0x2f0b:25,0x2f0c:50,0x2f0d-0x2f0f:25,0x2f10-0x2f16:50,0x2f17-0x2f18:25,0x2f19:50,0x2f1a-0x2f1f:50,0x3200-0x331f:50,0x13700-0x13702:50,0x13703-0x13707:25,0x13708-0x13709:50,0x1370a-0x1370f:50 min-size-polygon: 4 polygon-size-limits=16:2, 14:0 preserve-element-order keep-going route index And mkgmap ARGS for default: family-id: 10750 product-id: 1 draw-priority: 30 series-name: Great_Britain-default family-name: Great_Britain-default description: Great_Britain-default area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: examples/styles/default overview-mapname: 10750000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip # location-autofill: is_in,nearest # housenumbers # split-name-index # tdbfile # latin1 # name-tag-list: name,name:de,int_name,place_name code-page: 1252 # dem: HGT_NASADEM_Europe # dem-dists: 3314,4971,6628,13256,29826 # overview-dem-dist=89478 # show-profiles: 1 # merge-lines # road-name-config=examples/roadNameConfig.txt add-pois-to-areas # pois-to-areas-placement add-pois-to-lines link-pois-to-ways # min-size-polygon: 4 # polygon-size-limits=16:2, 14:0 # preserve-element-order keep-going route index
Am 06.08.2021 um 19:58 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com>:
Hi Jan,
Any error message from mkgmap? I can try to reproduce if you upload the input osm file to https://files.mkgmap.org.uk/ (or post another download link) and the options that you use.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von jan meisters <jan_m23@gmx.net> Gesendet: Freitag, 6. August 2021 19:35 An: Development list for mkgmap Betreff: [mkgmap-dev] Bad tile around East London / Greenwich
Hi all,
since some time my map has a bad tile around East London (Greenwich). I´ve checked earlier revisions of splitter and mkgmap, and decreased splitter´s max-nodes - without success.
Problem seems to be in my style, which is based on OpenFietsMap. With the original OFM-style the error is rendered as well, but not with the mkgmap default style. But I´ve got no clue where to look for the mistake - anyone has got a hint for me?
https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg
Jan
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Jan Have you tried creating the map without your TYP file as the picture reminds me of a possible issue with a POI. r Nick On 06/08/2021 21:41, jan meisters wrote:
Hi Gerd,
yes, but only on my own style: SCHWERWIEGEND (global): Bad file format: 10700105.osm.pbf
File uploaded here <https://files.mkgmap.org.uk/download/517/10700105.osm.pbf>. It is splitter-generated from a geofabrik great-britain-latest with max-node=1400000. Splitter 615 and mkgmap 4802.
For my own and and the default I have the same splitter options: java -Xmx26000m -jar splitter.jar --output=pbf --output-dir=/great_britain-splitter --max-nodes=1400000 --mapid=10750001 --geonames-file=PRECOMP/cities15000.txt --polygon-file=ARGS/great_britain.poly DATA/great_britain-latest.osm.pbf
And mkgmap then: java -Xmx26000m -jar mkgmap.jar --output-dir=great_britain-output -c ARGS/great_britain-HGT.args -c great_britain-splitter/template.args ARGS/10000.txt
The mkgmap ARGS for my style are: family-id: 10700 product-id: 1 draw-priority: 30 series-name: Great_Britain-HGT family-name: Great_Britain-HGT description: Great_Britain-HGT copyright-message: openstreetmap.org <http://openstreetmap.org> | earthexplorer.usgs.gov <http://earthexplorer.usgs.gov> | viewfinderpanoramas.org <http://viewfinderpanoramas.org> | openfietsmap.nl <http://openfietsmap.nl> area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: Styles overview-mapname: 10700000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip location-autofill: is_in,nearest housenumbers split-name-index tdbfile latin1 name-tag-list: name,name:de,int_name,place_name code-page: 1252 dem: HGT_NASADEM_Europe dem-dists: 3314,4971,6628,13256,29826 overview-dem-dist=89478 show-profiles: 1 merge-lines road-name-config=examples/roadNameConfig.txt add-pois-to-areas pois-to-areas-placement add-pois-to-lines link-pois-to-ways poi-excl-index=0x3200-0x331f nearby-poi-rules=*/named:5,*/unnamed:10,0x2b07:50,0x2f00-0x2f03:50,0x2f04-0x2f09:25,0x2f0a-0x2f0b:25,0x2f0c:50,0x2f0d-0x2f0f:25,0x2f10-0x2f16:50,0x2f17-0x2f18:25,0x2f19:50,0x2f1a-0x2f1f:50,0x3200-0x331f:50,0x13700-0x13702:50,0x13703-0x13707:25,0x13708-0x13709:50,0x1370a-0x1370f:50 min-size-polygon: 4 polygon-size-limits=16:2, 14:0 preserve-element-order keep-going route index
And mkgmap ARGS for default: family-id: 10750 product-id: 1 draw-priority: 30 series-name: Great_Britain-default family-name: Great_Britain-default description: Great_Britain-default area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: examples/styles/default overview-mapname: 10750000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip # location-autofill: is_in,nearest # housenumbers # split-name-index # tdbfile # latin1 # name-tag-list: name,name:de,int_name,place_name code-page: 1252 # dem: HGT_NASADEM_Europe # dem-dists: 3314,4971,6628,13256,29826 # overview-dem-dist=89478 # show-profiles: 1 # merge-lines # road-name-config=examples/roadNameConfig.txt add-pois-to-areas # pois-to-areas-placement add-pois-to-lines link-pois-to-ways # min-size-polygon: 4 # polygon-size-limits=16:2, 14:0 # preserve-element-order keep-going route index
Am 06.08.2021 um 19:58 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com <mailto:gpetermann_muenchen@hotmail.com>>:
Hi Jan,
Any error message from mkgmap? I can try to reproduce if you upload the input osm file to https://files.mkgmap.org.uk/ <https://files.mkgmap.org.uk/> (or post another download link) and the options that you use.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk <mailto:mkgmap-dev-bounces@lists.mkgmap.org.uk>> im Auftrag von jan meisters <jan_m23@gmx.net <mailto:jan_m23@gmx.net>> Gesendet: Freitag, 6. August 2021 19:35 An: Development list for mkgmap Betreff: [mkgmap-dev] Bad tile around East London / Greenwich
Hi all,
since some time my map has a bad tile around East London (Greenwich). I´ve checked earlier revisions of splitter and mkgmap, and decreased splitter´s max-nodes - without success.
Problem seems to be in my style, which is based on OpenFietsMap. With the original OFM-style the error is rendered as well, but not with the mkgmap default style. But I´ve got no clue where to look for the mistake - anyone has got a hint for me?
https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg <https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg>
Jan
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Jan, I can reproduce the problem with the OFM full style. With the just released r4804 I see this: SCHWERWIEGEND (global): Bad file format: e:\Jan\10700105.osm.pbf uk.me.parabola.imgfmt.FormatException: too many restrictions at uk.me.parabola.imgfmt.app.net.TableC.getOffsetSize(TableC.java:84) at uk.me.parabola.imgfmt.app.net.TableC.propagateSizeBytes(TableC.java:88) at uk.me.parabola.imgfmt.app.net.RouteCenter.updateOffsets(RouteCenter.java:87) at uk.me.parabola.imgfmt.app.net.RouteCenter.write(RouteCenter.java:98) at uk.me.parabola.imgfmt.app.net.NODFile.writeNodes(NODFile.java:101) at uk.me.parabola.imgfmt.app.net.NODFile.write(NODFile.java:71) at uk.me.parabola.mkgmap.build.MapBuilder.makeMap(MapBuilder.java:352) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:114) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:70) at uk.me.parabola.mkgmap.main.Main.lambda$1(Main.java:290) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748) I think this problem is caused by an error in mkgmap, it seems to put too many restrictions into one so called RouteCenter. Have to check the details... Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von jan meisters <jan_m23@gmx.net> Gesendet: Freitag, 6. August 2021 22:41 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Bad tile around East London / Greenwich Hi Gerd, yes, but only on my own style: SCHWERWIEGEND (global): Bad file format: 10700105.osm.pbf File uploaded here<https://files.mkgmap.org.uk/download/517/10700105.osm.pbf>. It is splitter-generated from a geofabrik great-britain-latest with max-node=1400000. Splitter 615 and mkgmap 4802. For my own and and the default I have the same splitter options: java -Xmx26000m -jar splitter.jar --output=pbf --output-dir=/great_britain-splitter --max-nodes=1400000 --mapid=10750001 --geonames-file=PRECOMP/cities15000.txt --polygon-file=ARGS/great_britain.poly DATA/great_britain-latest.osm.pbf And mkgmap then: java -Xmx26000m -jar mkgmap.jar --output-dir=great_britain-output -c ARGS/great_britain-HGT.args -c great_britain-splitter/template.args ARGS/10000.txt The mkgmap ARGS for my style are: family-id: 10700 product-id: 1 draw-priority: 30 series-name: Great_Britain-HGT family-name: Great_Britain-HGT description: Great_Britain-HGT copyright-message: openstreetmap.org<http://openstreetmap.org> | earthexplorer.usgs.gov<http://earthexplorer.usgs.gov> | viewfinderpanoramas.org<http://viewfinderpanoramas.org> | openfietsmap.nl<http://openfietsmap.nl> area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: Styles overview-mapname: 10700000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip location-autofill: is_in,nearest housenumbers split-name-index tdbfile latin1 name-tag-list: name,name:de,int_name,place_name code-page: 1252 dem: HGT_NASADEM_Europe dem-dists: 3314,4971,6628,13256,29826 overview-dem-dist=89478 show-profiles: 1 merge-lines road-name-config=examples/roadNameConfig.txt add-pois-to-areas pois-to-areas-placement add-pois-to-lines link-pois-to-ways poi-excl-index=0x3200-0x331f nearby-poi-rules=*/named:5,*/unnamed:10,0x2b07:50,0x2f00-0x2f03:50,0x2f04-0x2f09:25,0x2f0a-0x2f0b:25,0x2f0c:50,0x2f0d-0x2f0f:25,0x2f10-0x2f16:50,0x2f17-0x2f18:25,0x2f19:50,0x2f1a-0x2f1f:50,0x3200-0x331f:50,0x13700-0x13702:50,0x13703-0x13707:25,0x13708-0x13709:50,0x1370a-0x1370f:50 min-size-polygon: 4 polygon-size-limits=16:2, 14:0 preserve-element-order keep-going route index And mkgmap ARGS for default: family-id: 10750 product-id: 1 draw-priority: 30 series-name: Great_Britain-default family-name: Great_Britain-default description: Great_Britain-default area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: examples/styles/default overview-mapname: 10750000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip # location-autofill: is_in,nearest # housenumbers # split-name-index # tdbfile # latin1 # name-tag-list: name,name:de,int_name,place_name code-page: 1252 # dem: HGT_NASADEM_Europe # dem-dists: 3314,4971,6628,13256,29826 # overview-dem-dist=89478 # show-profiles: 1 # merge-lines # road-name-config=examples/roadNameConfig.txt add-pois-to-areas # pois-to-areas-placement add-pois-to-lines link-pois-to-ways # min-size-polygon: 4 # polygon-size-limits=16:2, 14:0 # preserve-element-order keep-going route index Am 06.08.2021 um 19:58 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com<mailto:gpetermann_muenchen@hotmail.com>>: Hi Jan, Any error message from mkgmap? I can try to reproduce if you upload the input osm file to https://files.mkgmap.org.uk/ (or post another download link) and the options that you use. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk<mailto:mkgmap-dev-bounces@lists.mkgmap.org.uk>> im Auftrag von jan meisters <jan_m23@gmx.net<mailto:jan_m23@gmx.net>> Gesendet: Freitag, 6. August 2021 19:35 An: Development list for mkgmap Betreff: [mkgmap-dev] Bad tile around East London / Greenwich Hi all, since some time my map has a bad tile around East London (Greenwich). I´ve checked earlier revisions of splitter and mkgmap, and decreased splitter´s max-nodes - without success. Problem seems to be in my style, which is based on OpenFietsMap. With the original OFM-style the error is rendered as well, but not with the mkgmap default style. But I´ve got no clue where to look for the mistake - anyone has got a hint for me? https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg Jan _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Jan, I think r4805 fixes the problem. I've not analysed why the default style doesn't show the problem. The routing data is very different. It seems that the OFM style produces a lot more restrictions. Gerd ________________________________________ Von: Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Samstag, 7. August 2021 09:47 An: Development list for mkgmap Betreff: AW: [mkgmap-dev] Bad tile around East London / Greenwich Hi Jan, I can reproduce the problem with the OFM full style. With the just released r4804 I see this: SCHWERWIEGEND (global): Bad file format: e:\Jan\10700105.osm.pbf uk.me.parabola.imgfmt.FormatException: too many restrictions at uk.me.parabola.imgfmt.app.net.TableC.getOffsetSize(TableC.java:84) at uk.me.parabola.imgfmt.app.net.TableC.propagateSizeBytes(TableC.java:88) at uk.me.parabola.imgfmt.app.net.RouteCenter.updateOffsets(RouteCenter.java:87) at uk.me.parabola.imgfmt.app.net.RouteCenter.write(RouteCenter.java:98) at uk.me.parabola.imgfmt.app.net.NODFile.writeNodes(NODFile.java:101) at uk.me.parabola.imgfmt.app.net.NODFile.write(NODFile.java:71) at uk.me.parabola.mkgmap.build.MapBuilder.makeMap(MapBuilder.java:352) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:114) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:70) at uk.me.parabola.mkgmap.main.Main.lambda$1(Main.java:290) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748) I think this problem is caused by an error in mkgmap, it seems to put too many restrictions into one so called RouteCenter. Have to check the details... Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von jan meisters <jan_m23@gmx.net> Gesendet: Freitag, 6. August 2021 22:41 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Bad tile around East London / Greenwich Hi Gerd, yes, but only on my own style: SCHWERWIEGEND (global): Bad file format: 10700105.osm.pbf File uploaded here<https://files.mkgmap.org.uk/download/517/10700105.osm.pbf>. It is splitter-generated from a geofabrik great-britain-latest with max-node=1400000. Splitter 615 and mkgmap 4802. For my own and and the default I have the same splitter options: java -Xmx26000m -jar splitter.jar --output=pbf --output-dir=/great_britain-splitter --max-nodes=1400000 --mapid=10750001 --geonames-file=PRECOMP/cities15000.txt --polygon-file=ARGS/great_britain.poly DATA/great_britain-latest.osm.pbf And mkgmap then: java -Xmx26000m -jar mkgmap.jar --output-dir=great_britain-output -c ARGS/great_britain-HGT.args -c great_britain-splitter/template.args ARGS/10000.txt The mkgmap ARGS for my style are: family-id: 10700 product-id: 1 draw-priority: 30 series-name: Great_Britain-HGT family-name: Great_Britain-HGT description: Great_Britain-HGT copyright-message: openstreetmap.org<http://openstreetmap.org> | earthexplorer.usgs.gov<http://earthexplorer.usgs.gov> | viewfinderpanoramas.org<http://viewfinderpanoramas.org> | openfietsmap.nl<http://openfietsmap.nl> area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: Styles overview-mapname: 10700000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip location-autofill: is_in,nearest housenumbers split-name-index tdbfile latin1 name-tag-list: name,name:de,int_name,place_name code-page: 1252 dem: HGT_NASADEM_Europe dem-dists: 3314,4971,6628,13256,29826 overview-dem-dist=89478 show-profiles: 1 merge-lines road-name-config=examples/roadNameConfig.txt add-pois-to-areas pois-to-areas-placement add-pois-to-lines link-pois-to-ways poi-excl-index=0x3200-0x331f nearby-poi-rules=*/named:5,*/unnamed:10,0x2b07:50,0x2f00-0x2f03:50,0x2f04-0x2f09:25,0x2f0a-0x2f0b:25,0x2f0c:50,0x2f0d-0x2f0f:25,0x2f10-0x2f16:50,0x2f17-0x2f18:25,0x2f19:50,0x2f1a-0x2f1f:50,0x3200-0x331f:50,0x13700-0x13702:50,0x13703-0x13707:25,0x13708-0x13709:50,0x1370a-0x1370f:50 min-size-polygon: 4 polygon-size-limits=16:2, 14:0 preserve-element-order keep-going route index And mkgmap ARGS for default: family-id: 10750 product-id: 1 draw-priority: 30 series-name: Great_Britain-default family-name: Great_Britain-default description: Great_Britain-default area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: examples/styles/default overview-mapname: 10750000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip # location-autofill: is_in,nearest # housenumbers # split-name-index # tdbfile # latin1 # name-tag-list: name,name:de,int_name,place_name code-page: 1252 # dem: HGT_NASADEM_Europe # dem-dists: 3314,4971,6628,13256,29826 # overview-dem-dist=89478 # show-profiles: 1 # merge-lines # road-name-config=examples/roadNameConfig.txt add-pois-to-areas # pois-to-areas-placement add-pois-to-lines link-pois-to-ways # min-size-polygon: 4 # polygon-size-limits=16:2, 14:0 # preserve-element-order keep-going route index Am 06.08.2021 um 19:58 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com<mailto:gpetermann_muenchen@hotmail.com>>: Hi Jan, Any error message from mkgmap? I can try to reproduce if you upload the input osm file to https://files.mkgmap.org.uk/ (or post another download link) and the options that you use. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk<mailto:mkgmap-dev-bounces@lists.mkgmap.org.uk>> im Auftrag von jan meisters <jan_m23@gmx.net<mailto:jan_m23@gmx.net>> Gesendet: Freitag, 6. August 2021 19:35 An: Development list for mkgmap Betreff: [mkgmap-dev] Bad tile around East London / Greenwich Hi all, since some time my map has a bad tile around East London (Greenwich). I´ve checked earlier revisions of splitter and mkgmap, and decreased splitter´s max-nodes - without success. Problem seems to be in my style, which is based on OpenFietsMap. With the original OFM-style the error is rendered as well, but not with the mkgmap default style. But I´ve got no clue where to look for the mistake - anyone has got a hint for me? https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg Jan _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Gerd, yes, 4805 renders correctly. Thanks! @Nick: I tried without typ, but that made no difference. Do you remember the faulty POI? Jan
Am 07.08.2021 um 11:24 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com>:
Hi Jan,
I think r4805 fixes the problem. I've not analysed why the default style doesn't show the problem. The routing data is very different. It seems that the OFM style produces a lot more restrictions.
Gerd
________________________________________ Von: Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Samstag, 7. August 2021 09:47 An: Development list for mkgmap Betreff: AW: [mkgmap-dev] Bad tile around East London / Greenwich
Hi Jan,
I can reproduce the problem with the OFM full style. With the just released r4804 I see this: SCHWERWIEGEND (global): Bad file format: e:\Jan\10700105.osm.pbf uk.me.parabola.imgfmt.FormatException: too many restrictions at uk.me.parabola.imgfmt.app.net.TableC.getOffsetSize(TableC.java:84) at uk.me.parabola.imgfmt.app.net.TableC.propagateSizeBytes(TableC.java:88) at uk.me.parabola.imgfmt.app.net.RouteCenter.updateOffsets(RouteCenter.java:87) at uk.me.parabola.imgfmt.app.net.RouteCenter.write(RouteCenter.java:98) at uk.me.parabola.imgfmt.app.net.NODFile.writeNodes(NODFile.java:101) at uk.me.parabola.imgfmt.app.net.NODFile.write(NODFile.java:71) at uk.me.parabola.mkgmap.build.MapBuilder.makeMap(MapBuilder.java:352) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:114) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:70) at uk.me.parabola.mkgmap.main.Main.lambda$1(Main.java:290) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748)
I think this problem is caused by an error in mkgmap, it seems to put too many restrictions into one so called RouteCenter. Have to check the details...
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von jan meisters <jan_m23@gmx.net> Gesendet: Freitag, 6. August 2021 22:41 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Bad tile around East London / Greenwich
Hi Gerd,
yes, but only on my own style: SCHWERWIEGEND (global): Bad file format: 10700105.osm.pbf
File uploaded here<https://files.mkgmap.org.uk/download/517/10700105.osm.pbf>. It is splitter-generated from a geofabrik great-britain-latest with max-node=1400000. Splitter 615 and mkgmap 4802.
For my own and and the default I have the same splitter options: java -Xmx26000m -jar splitter.jar --output=pbf --output-dir=/great_britain-splitter --max-nodes=1400000 --mapid=10750001 --geonames-file=PRECOMP/cities15000.txt --polygon-file=ARGS/great_britain.poly DATA/great_britain-latest.osm.pbf
And mkgmap then: java -Xmx26000m -jar mkgmap.jar --output-dir=great_britain-output -c ARGS/great_britain-HGT.args -c great_britain-splitter/template.args ARGS/10000.txt
The mkgmap ARGS for my style are: family-id: 10700 product-id: 1 draw-priority: 30 series-name: Great_Britain-HGT family-name: Great_Britain-HGT description: Great_Britain-HGT copyright-message: openstreetmap.org<http://openstreetmap.org> | earthexplorer.usgs.gov<http://earthexplorer.usgs.gov> | viewfinderpanoramas.org<http://viewfinderpanoramas.org> | openfietsmap.nl<http://openfietsmap.nl> area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: Styles overview-mapname: 10700000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip location-autofill: is_in,nearest housenumbers split-name-index tdbfile latin1 name-tag-list: name,name:de,int_name,place_name code-page: 1252 dem: HGT_NASADEM_Europe dem-dists: 3314,4971,6628,13256,29826 overview-dem-dist=89478 show-profiles: 1 merge-lines road-name-config=examples/roadNameConfig.txt add-pois-to-areas pois-to-areas-placement add-pois-to-lines link-pois-to-ways poi-excl-index=0x3200-0x331f nearby-poi-rules=*/named:5,*/unnamed:10,0x2b07:50,0x2f00-0x2f03:50,0x2f04-0x2f09:25,0x2f0a-0x2f0b:25,0x2f0c:50,0x2f0d-0x2f0f:25,0x2f10-0x2f16:50,0x2f17-0x2f18:25,0x2f19:50,0x2f1a-0x2f1f:50,0x3200-0x331f:50,0x13700-0x13702:50,0x13703-0x13707:25,0x13708-0x13709:50,0x1370a-0x1370f:50 min-size-polygon: 4 polygon-size-limits=16:2, 14:0 preserve-element-order keep-going route index
And mkgmap ARGS for default: family-id: 10750 product-id: 1 draw-priority: 30 series-name: Great_Britain-default family-name: Great_Britain-default description: Great_Britain-default area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: examples/styles/default overview-mapname: 10750000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip # location-autofill: is_in,nearest # housenumbers # split-name-index # tdbfile # latin1 # name-tag-list: name,name:de,int_name,place_name code-page: 1252 # dem: HGT_NASADEM_Europe # dem-dists: 3314,4971,6628,13256,29826 # overview-dem-dist=89478 # show-profiles: 1 # merge-lines # road-name-config=examples/roadNameConfig.txt add-pois-to-areas # pois-to-areas-placement add-pois-to-lines link-pois-to-ways # min-size-polygon: 4 # polygon-size-limits=16:2, 14:0 # preserve-element-order keep-going route index
Am 06.08.2021 um 19:58 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com<mailto:gpetermann_muenchen@hotmail.com>>:
Hi Jan,
Any error message from mkgmap? I can try to reproduce if you upload the input osm file to https://files.mkgmap.org.uk/ (or post another download link) and the options that you use.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk<mailto:mkgmap-dev-bounces@lists.mkgmap.org.uk>> im Auftrag von jan meisters <jan_m23@gmx.net<mailto:jan_m23@gmx.net>> Gesendet: Freitag, 6. August 2021 19:35 An: Development list for mkgmap Betreff: [mkgmap-dev] Bad tile around East London / Greenwich
Hi all,
since some time my map has a bad tile around East London (Greenwich). I´ve checked earlier revisions of splitter and mkgmap, and decreased splitter´s max-nodes - without success.
Problem seems to be in my style, which is based on OpenFietsMap. With the original OFM-style the error is rendered as well, but not with the mkgmap default style. But I´ve got no clue where to look for the mistake - anyone has got a hint for me?
https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg
Jan
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Jan It wasn't so much the type number as the size of a poi bitmap. In the 'olden' days Mapsource would allow pois with lengths and width equalling zero. This would render them transparent, useful for names of places. However, Basecamp would not accept such pois and I think has a problem displaying some pois with length and width set to 1. I'm glad it wasn't the typ file. r Nick On 07/08/2021 14:59, jan meisters wrote:
Hi Gerd,
yes, 4805 renders correctly. Thanks!
@Nick: I tried without typ, but that made no difference. Do you remember the faulty POI?
Jan
Am 07.08.2021 um 11:24 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com>:
Hi Jan,
I think r4805 fixes the problem. I've not analysed why the default style doesn't show the problem. The routing data is very different. It seems that the OFM style produces a lot more restrictions.
Gerd
________________________________________ Von: Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Samstag, 7. August 2021 09:47 An: Development list for mkgmap Betreff: AW: [mkgmap-dev] Bad tile around East London / Greenwich
Hi Jan,
I can reproduce the problem with the OFM full style. With the just released r4804 I see this: SCHWERWIEGEND (global): Bad file format: e:\Jan\10700105.osm.pbf uk.me.parabola.imgfmt.FormatException: too many restrictions at uk.me.parabola.imgfmt.app.net.TableC.getOffsetSize(TableC.java:84) at uk.me.parabola.imgfmt.app.net.TableC.propagateSizeBytes(TableC.java:88) at uk.me.parabola.imgfmt.app.net.RouteCenter.updateOffsets(RouteCenter.java:87) at uk.me.parabola.imgfmt.app.net.RouteCenter.write(RouteCenter.java:98) at uk.me.parabola.imgfmt.app.net.NODFile.writeNodes(NODFile.java:101) at uk.me.parabola.imgfmt.app.net.NODFile.write(NODFile.java:71) at uk.me.parabola.mkgmap.build.MapBuilder.makeMap(MapBuilder.java:352) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:114) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:70) at uk.me.parabola.mkgmap.main.Main.lambda$1(Main.java:290) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748)
I think this problem is caused by an error in mkgmap, it seems to put too many restrictions into one so called RouteCenter. Have to check the details...
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von jan meisters <jan_m23@gmx.net> Gesendet: Freitag, 6. August 2021 22:41 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Bad tile around East London / Greenwich
Hi Gerd,
yes, but only on my own style: SCHWERWIEGEND (global): Bad file format: 10700105.osm.pbf
File uploaded here<https://files.mkgmap.org.uk/download/517/10700105.osm.pbf>. It is splitter-generated from a geofabrik great-britain-latest with max-node=1400000. Splitter 615 and mkgmap 4802.
For my own and and the default I have the same splitter options: java -Xmx26000m -jar splitter.jar --output=pbf --output-dir=/great_britain-splitter --max-nodes=1400000 --mapid=10750001 --geonames-file=PRECOMP/cities15000.txt --polygon-file=ARGS/great_britain.poly DATA/great_britain-latest.osm.pbf
And mkgmap then: java -Xmx26000m -jar mkgmap.jar --output-dir=great_britain-output -c ARGS/great_britain-HGT.args -c great_britain-splitter/template.args ARGS/10000.txt
The mkgmap ARGS for my style are: family-id: 10700 product-id: 1 draw-priority: 30 series-name: Great_Britain-HGT family-name: Great_Britain-HGT description: Great_Britain-HGT copyright-message: openstreetmap.org<http://openstreetmap.org> | earthexplorer.usgs.gov<http://earthexplorer.usgs.gov> | viewfinderpanoramas.org<http://viewfinderpanoramas.org> | openfietsmap.nl<http://openfietsmap.nl> area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: Styles overview-mapname: 10700000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip location-autofill: is_in,nearest housenumbers split-name-index tdbfile latin1 name-tag-list: name,name:de,int_name,place_name code-page: 1252 dem: HGT_NASADEM_Europe dem-dists: 3314,4971,6628,13256,29826 overview-dem-dist=89478 show-profiles: 1 merge-lines road-name-config=examples/roadNameConfig.txt add-pois-to-areas pois-to-areas-placement add-pois-to-lines link-pois-to-ways poi-excl-index=0x3200-0x331f nearby-poi-rules=*/named:5,*/unnamed:10,0x2b07:50,0x2f00-0x2f03:50,0x2f04-0x2f09:25,0x2f0a-0x2f0b:25,0x2f0c:50,0x2f0d-0x2f0f:25,0x2f10-0x2f16:50,0x2f17-0x2f18:25,0x2f19:50,0x2f1a-0x2f1f:50,0x3200-0x331f:50,0x13700-0x13702:50,0x13703-0x13707:25,0x13708-0x13709:50,0x1370a-0x1370f:50 min-size-polygon: 4 polygon-size-limits=16:2, 14:0 preserve-element-order keep-going route index
And mkgmap ARGS for default: family-id: 10750 product-id: 1 draw-priority: 30 series-name: Great_Britain-default family-name: Great_Britain-default description: Great_Britain-default area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: examples/styles/default overview-mapname: 10750000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip # location-autofill: is_in,nearest # housenumbers # split-name-index # tdbfile # latin1 # name-tag-list: name,name:de,int_name,place_name code-page: 1252 # dem: HGT_NASADEM_Europe # dem-dists: 3314,4971,6628,13256,29826 # overview-dem-dist=89478 # show-profiles: 1 # merge-lines # road-name-config=examples/roadNameConfig.txt add-pois-to-areas # pois-to-areas-placement add-pois-to-lines link-pois-to-ways # min-size-polygon: 4 # polygon-size-limits=16:2, 14:0 # preserve-element-order keep-going route index
Am 06.08.2021 um 19:58 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com<mailto:gpetermann_muenchen@hotmail.com>>:
Hi Jan,
Any error message from mkgmap? I can try to reproduce if you upload the input osm file to https://files.mkgmap.org.uk/ (or post another download link) and the options that you use.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk<mailto:mkgmap-dev-bounces@lists.mkgmap.org.uk>> im Auftrag von jan meisters <jan_m23@gmx.net<mailto:jan_m23@gmx.net>> Gesendet: Freitag, 6. August 2021 19:35 An: Development list for mkgmap Betreff: [mkgmap-dev] Bad tile around East London / Greenwich
Hi all,
since some time my map has a bad tile around East London (Greenwich). I´ve checked earlier revisions of splitter and mkgmap, and decreased splitter´s max-nodes - without success.
Problem seems to be in my style, which is based on OpenFietsMap. With the original OFM-style the error is rendered as well, but not with the mkgmap default style. But I´ve got no clue where to look for the mistake - anyone has got a hint for me?
https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg
Jan
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Nick, interesting! Indeed I have a lot transparent 1x1-POIs in my typ, but of course for stuff I only want in the index and not to be displayed. As my map is only used by me with BaseCamp and my Oregon 450 I´m not aware of possible problems beside this setup. I´ll keep an eye on it. Jan
Am 07.08.2021 um 17:17 schrieb nick <osm@pinns.co.uk>:
Hi Jan
It wasn't so much the type number as the size of a poi bitmap.
In the 'olden' days Mapsource would allow pois with lengths and width equalling zero. This would render them transparent, useful for names of places.
However, Basecamp would not accept such pois and I think has a problem displaying some pois with length and width set to 1.
I'm glad it wasn't the typ file.
r
Nick
On 07/08/2021 14:59, jan meisters wrote:
Hi Gerd,
yes, 4805 renders correctly. Thanks!
@Nick: I tried without typ, but that made no difference. Do you remember the faulty POI?
Jan
Am 07.08.2021 um 11:24 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com>:
Hi Jan,
I think r4805 fixes the problem. I've not analysed why the default style doesn't show the problem. The routing data is very different. It seems that the OFM style produces a lot more restrictions.
Gerd
________________________________________ Von: Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Samstag, 7. August 2021 09:47 An: Development list for mkgmap Betreff: AW: [mkgmap-dev] Bad tile around East London / Greenwich
Hi Jan,
I can reproduce the problem with the OFM full style. With the just released r4804 I see this: SCHWERWIEGEND (global): Bad file format: e:\Jan\10700105.osm.pbf uk.me.parabola.imgfmt.FormatException: too many restrictions at uk.me.parabola.imgfmt.app.net.TableC.getOffsetSize(TableC.java:84) at uk.me.parabola.imgfmt.app.net.TableC.propagateSizeBytes(TableC.java:88) at uk.me.parabola.imgfmt.app.net.RouteCenter.updateOffsets(RouteCenter.java:87) at uk.me.parabola.imgfmt.app.net.RouteCenter.write(RouteCenter.java:98) at uk.me.parabola.imgfmt.app.net.NODFile.writeNodes(NODFile.java:101) at uk.me.parabola.imgfmt.app.net.NODFile.write(NODFile.java:71) at uk.me.parabola.mkgmap.build.MapBuilder.makeMap(MapBuilder.java:352) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:114) at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:70) at uk.me.parabola.mkgmap.main.Main.lambda$1(Main.java:290) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748)
I think this problem is caused by an error in mkgmap, it seems to put too many restrictions into one so called RouteCenter. Have to check the details...
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von jan meisters <jan_m23@gmx.net> Gesendet: Freitag, 6. August 2021 22:41 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Bad tile around East London / Greenwich
Hi Gerd,
yes, but only on my own style: SCHWERWIEGEND (global): Bad file format: 10700105.osm.pbf
File uploaded here<https://files.mkgmap.org.uk/download/517/10700105.osm.pbf>. It is splitter-generated from a geofabrik great-britain-latest with max-node=1400000. Splitter 615 and mkgmap 4802.
For my own and and the default I have the same splitter options: java -Xmx26000m -jar splitter.jar --output=pbf --output-dir=/great_britain-splitter --max-nodes=1400000 --mapid=10750001 --geonames-file=PRECOMP/cities15000.txt --polygon-file=ARGS/great_britain.poly DATA/great_britain-latest.osm.pbf
And mkgmap then: java -Xmx26000m -jar mkgmap.jar --output-dir=great_britain-output -c ARGS/great_britain-HGT.args -c great_britain-splitter/template.args ARGS/10000.txt
The mkgmap ARGS for my style are: family-id: 10700 product-id: 1 draw-priority: 30 series-name: Great_Britain-HGT family-name: Great_Britain-HGT description: Great_Britain-HGT copyright-message: openstreetmap.org<http://openstreetmap.org> | earthexplorer.usgs.gov<http://earthexplorer.usgs.gov> | viewfinderpanoramas.org<http://viewfinderpanoramas.org> | openfietsmap.nl<http://openfietsmap.nl> area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: Styles overview-mapname: 10700000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip location-autofill: is_in,nearest housenumbers split-name-index tdbfile latin1 name-tag-list: name,name:de,int_name,place_name code-page: 1252 dem: HGT_NASADEM_Europe dem-dists: 3314,4971,6628,13256,29826 overview-dem-dist=89478 show-profiles: 1 merge-lines road-name-config=examples/roadNameConfig.txt add-pois-to-areas pois-to-areas-placement add-pois-to-lines link-pois-to-ways poi-excl-index=0x3200-0x331f nearby-poi-rules=*/named:5,*/unnamed:10,0x2b07:50,0x2f00-0x2f03:50,0x2f04-0x2f09:25,0x2f0a-0x2f0b:25,0x2f0c:50,0x2f0d-0x2f0f:25,0x2f10-0x2f16:50,0x2f17-0x2f18:25,0x2f19:50,0x2f1a-0x2f1f:50,0x3200-0x331f:50,0x13700-0x13702:50,0x13703-0x13707:25,0x13708-0x13709:50,0x1370a-0x1370f:50 min-size-polygon: 4 polygon-size-limits=16:2, 14:0 preserve-element-order keep-going route index
And mkgmap ARGS for default: family-id: 10750 product-id: 1 draw-priority: 30 series-name: Great_Britain-default family-name: Great_Britain-default description: Great_Britain-default area-name: Great_Britain country-name: Great_Britain country-abbr: Great_Britain style-file: examples/styles/default overview-mapname: 10750000 precomp-sea: PRECOMP/sea-latest.zip generate-sea: land-tag=natural=background bounds: PRECOMP/bounds-latest.zip # location-autofill: is_in,nearest # housenumbers # split-name-index # tdbfile # latin1 # name-tag-list: name,name:de,int_name,place_name code-page: 1252 # dem: HGT_NASADEM_Europe # dem-dists: 3314,4971,6628,13256,29826 # overview-dem-dist=89478 # show-profiles: 1 # merge-lines # road-name-config=examples/roadNameConfig.txt add-pois-to-areas # pois-to-areas-placement add-pois-to-lines link-pois-to-ways # min-size-polygon: 4 # polygon-size-limits=16:2, 14:0 # preserve-element-order keep-going route index
Am 06.08.2021 um 19:58 schrieb Gerd Petermann <gpetermann_muenchen@hotmail.com<mailto:gpetermann_muenchen@hotmail.com>>:
Hi Jan,
Any error message from mkgmap? I can try to reproduce if you upload the input osm file to https://files.mkgmap.org.uk/ (or post another download link) and the options that you use.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk<mailto:mkgmap-dev-bounces@lists.mkgmap.org.uk>> im Auftrag von jan meisters <jan_m23@gmx.net<mailto:jan_m23@gmx.net>> Gesendet: Freitag, 6. August 2021 19:35 An: Development list for mkgmap Betreff: [mkgmap-dev] Bad tile around East London / Greenwich
Hi all,
since some time my map has a bad tile around East London (Greenwich). I´ve checked earlier revisions of splitter and mkgmap, and decreased splitter´s max-nodes - without success.
Problem seems to be in my style, which is based on OpenFietsMap. With the original OFM-style the error is rendered as well, but not with the mkgmap default style. But I´ve got no clue where to look for the mistake - anyone has got a hint for me?
https://files.mkgmap.org.uk/download/516/bad_tile-east_london.jpg
Jan
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi, I've tried it too with my OFM full style and mkgmap v 4802 but it doesn't show any issues at all. Maybe some parameters are different? I don't use: pois-to-areas-placement poi-excl-index nearby-poi-rules

Hi Minko, I just tried with the latest version of the OFM style and I can still reproduce the problem with r4802. I adapted he paths in Jans args and commented only the DEM options. I can also reproduce it with these few options: java -jar mkgmap.jar --route --link-pois-to-ways --style-file="d:\git_repos\minko\styles\Openfietsmap full" e:\jan\10700105.osm.pbf Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von lig fietser <ligfietser@hotmail.com> Gesendet: Samstag, 7. August 2021 17:24 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Bad tile around East London / Greenwich Hi, I've tried it too with my OFM full style and mkgmap v 4802 but it doesn't show any issues at all. Maybe some parameters are different? I don't use: pois-to-areas-placement poi-excl-index nearby-poi-rules

Hi Gerd and Jan, I have found the difference that caused this issue for you and not for me: I have used the parameter ignore-turn-restrictions . Now I can remember I have seen this issue too in a tile in the UK as well, and by adding this it was solved. Does this make sense Gerd? Mfrgr, Minko ________________________________ Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> namens Gerd Petermann <gpetermann_muenchen@hotmail.com> Verzonden: zaterdag 7 augustus 2021 08:51 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] Bad tile around East London / Greenwich Hi Minko, I just tried with the latest version of the OFM style and I can still reproduce the problem with r4802. I adapted he paths in Jans args and commented only the DEM options. I can also reproduce it with these few options: java -jar mkgmap.jar --route --link-pois-to-ways --style-file="d:\git_repos\minko\styles\Openfietsmap full" e:\jan\10700105.osm.pbf Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von lig fietser <ligfietser@hotmail.com> Gesendet: Samstag, 7. August 2021 17:24 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Bad tile around East London / Greenwich Hi, I've tried it too with my OFM full style and mkgmap v 4802 but it doesn't show any issues at all. Maybe some parameters are different? I don't use: pois-to-areas-placement poi-excl-index nearby-poi-rules _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Minko, well, yes, the problem is caused by an extreme density of (Garmin) route restrictions. Those are created for OSM turn restrictions and also for barriers when --link-pois-to-ways is used. Maybe this area contains many type=restriction relations with via ways, those require more bytes compared to simple turn restrictions with one via node. I think the fix in r4805 is the better alternative compared to using --ignore-turn-restrictions. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von lig fietser <ligfietser@hotmail.com> Gesendet: Samstag, 7. August 2021 18:57 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Bad tile around East London / Greenwich Hi Gerd and Jan, I have found the difference that caused this issue for you and not for me: I have used the parameter ignore-turn-restrictions . Now I can remember I have seen this issue too in a tile in the UK as well, and by adding this it was solved. Does this make sense Gerd? Mfrgr, Minko ________________________________ Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> namens Gerd Petermann <gpetermann_muenchen@hotmail.com> Verzonden: zaterdag 7 augustus 2021 08:51 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] Bad tile around East London / Greenwich Hi Minko, I just tried with the latest version of the OFM style and I can still reproduce the problem with r4802. I adapted he paths in Jans args and commented only the DEM options. I can also reproduce it with these few options: java -jar mkgmap.jar --route --link-pois-to-ways --style-file="d:\git_repos\minko\styles\Openfietsmap full" e:\jan\10700105.osm.pbf Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von lig fietser <ligfietser@hotmail.com> Gesendet: Samstag, 7. August 2021 17:24 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] Bad tile around East London / Greenwich Hi, I've tried it too with my OFM full style and mkgmap v 4802 but it doesn't show any issues at all. Maybe some parameters are different? I don't use: pois-to-areas-placement poi-excl-index nearby-poi-rules _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Thanks for the explanation and the fix Gerd! ________________________________ Hi Minko, well, yes, the problem is caused by an extreme density of (Garmin) route restrictions. Those are created for OSM turn restrictions and also for barriers when --link-pois-to-ways is used. Maybe this area contains many type=restriction relations with via ways, those require more bytes compared to simple turn restrictions with one via node. I think the fix in r4805 is the better alternative compared to using --ignore-turn-restrictions. Gerd
participants (4)
-
Gerd Petermann
-
jan meisters
-
lig fietser
-
nick