Hi Ticker, I tried to reach you with a PM but that was returned, so I try again via list... Gerd ________________________________________ Von: Gerd Petermann Gesendet: Mittwoch, 1. Februar 2017 16:09 An: Ticker Berkin Betreff: Please help Hi Ticker, I try to understand why ShapeSplitter complains about a polygon created by my new algo. Attached is a patch and an input file. Command: java -Dlog.config=mkgmap_log.props -ea -jar dist\mkgmap.jar f:\osm\test.osm.pbf My mkgmap_log.props contains .level=WARNING and I see this warning: WARN: uk.me.parabola.util.ShapeSplitter f:\osm\test.osm.pbf: Direction of shape and hole conflict; probably self-intersecting polygon this: fp=146086642 lp=146173971 area=19324813903 #=134 http://www.openstreetmap.org/?mlat=48.975820&mlon=13.355620&zoom=17 http://www.openstreetmap.org/?mlat=48.988628&mlon=13.389160&zoom=17 other: fp=146153055 lp=146161739 area=22313626 #=6 http://www.openstreetmap.org/?mlat=49.002597&mlon=13.353453&zoom=17 http://www.openstreetmap.org/?mlat=49.003250&mlon=13.352696&zoom=17 I don't see what's wrong here. Gerd
Hi Gerd I've plotted the outer data. Not quite well enough to see all the directions of lines and if they emerge into each hole without crossing, or any slight overlaps, however: On the bottom edge, there are two areas in the middle that look as if they should be indentations into the overall area but also have a bottom edge that follows the bounding box Probably not a problem - on the left edge there are two holes where the top one is closer to the edge than the hole it is joined from Are diagonal cuts OK - on some devices you can see where this is happening (for some reason the original OSM data did this) Let me know if you want me to look in more detail. Ticker On Wed, 2017-02-01 at 15:24 +0000, Gerd Petermann wrote:
Hi Ticker,
I tried to reach you with a PM but that was returned, so I try again via list...
Gerd
________________________________________ Von: Gerd Petermann Gesendet: Mittwoch, 1. Februar 2017 16:09 An: Ticker Berkin Betreff: Please help
Hi Ticker,
I try to understand why ShapeSplitter complains about a polygon created by my new algo. Attached is a patch and an input file. Command: java -Dlog.config=mkgmap_log.props -ea -jar dist\mkgmap.jar f:\osm\test.osm.pbf My mkgmap_log.props contains .level=WARNING and I see this warning: WARN: uk.me.parabola.util.ShapeSplitter f:\osm\test.osm.pbf: Direction of shape and hole conflict; probably self-intersecting polygon this: fp=146086642 lp=146173971 area=19324813903 #=134 http://www.openstreetmap.org/?mlat=48.975820&mlon=13.355620&zoom=17 http://www.openstreetmap.org/?mlat=48.988628&mlon=13.389160&zoom=17 o ther: fp=146153055 lp=146161739 area=22313626 #=6 http://www.openstreetmap.org/?mlat=49.002597&mlon=13.353453&zoom=17 http://www.openstreetmap.org/?mlat=49.003250&mlon=13.352696&zoom=17
I don't see what's wrong here.
Gerd _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Ticker, if diagonal cuts cause problems I can stop this. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 1. Februar 2017 17:04:36 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] WG: Please help Hi Gerd I've plotted the outer data. Not quite well enough to see all the directions of lines and if they emerge into each hole without crossing, or any slight overlaps, however: On the bottom edge, there are two areas in the middle that look as if they should be indentations into the overall area but also have a bottom edge that follows the bounding box Probably not a problem - on the left edge there are two holes where the top one is closer to the edge than the hole it is joined from Are diagonal cuts OK - on some devices you can see where this is happening (for some reason the original OSM data did this) Let me know if you want me to look in more detail. Ticker On Wed, 2017-02-01 at 15:24 +0000, Gerd Petermann wrote:
Hi Ticker,
I tried to reach you with a PM but that was returned, so I try again via list...
Gerd
________________________________________ Von: Gerd Petermann Gesendet: Mittwoch, 1. Februar 2017 16:09 An: Ticker Berkin Betreff: Please help
Hi Ticker,
I try to understand why ShapeSplitter complains about a polygon created by my new algo. Attached is a patch and an input file. Command: java -Dlog.config=mkgmap_log.props -ea -jar dist\mkgmap.jar f:\osm\test.osm.pbf My mkgmap_log.props contains .level=WARNING and I see this warning: WARN: uk.me.parabola.util.ShapeSplitter f:\osm\test.osm.pbf: Direction of shape and hole conflict; probably self-intersecting polygon this: fp=146086642 lp=146173971 area=19324813903 #=134 http://www.openstreetmap.org/?mlat=48.975820&mlon=13.355620&zoom=17 http://www.openstreetmap.org/?mlat=48.988628&mlon=13.389160&zoom=17 o ther: fp=146153055 lp=146161739 area=22313626 #=6 http://www.openstreetmap.org/?mlat=49.002597&mlon=13.353453&zoom=17 http://www.openstreetmap.org/?mlat=49.003250&mlon=13.352696&zoom=17
I don't see what's wrong here.
Gerd _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Gerd Did the other comment help you find the problem - it occurred to me later that I was dumping the data on input to the crashing ShapeSplitter and something had probably clipped the shape between multiPolygon generating it and this division. Also my plot shows the area rotated by 90 degrees (Latitude as X = horizontal) Ticker On Wed, 2017-02-01 at 16:39 +0000, Gerd Petermann wrote:
Hi Ticker,
if diagonal cuts cause problems I can stop this.
Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 1. Februar 2017 17:04:36 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] WG: Please help
Hi Gerd
I've plotted the outer data. Not quite well enough to see all the directions of lines and if they emerge into each hole without crossing, or any slight overlaps, however:
On the bottom edge, there are two areas in the middle that look as if they should be indentations into the overall area but also have a bottom edge that follows the bounding box
Probably not a problem - on the left edge there are two holes where the top one is closer to the edge than the hole it is joined from
Are diagonal cuts OK - on some devices you can see where this is happening (for some reason the original OSM data did this)
Let me know if you want me to look in more detail.
Ticker
On Wed, 2017-02-01 at 15:24 +0000, Gerd Petermann wrote:
Hi Ticker,
I tried to reach you with a PM but that was returned, so I try again via list...
Gerd
________________________________________ Von: Gerd Petermann Gesendet: Mittwoch, 1. Februar 2017 16:09 An: Ticker Berkin Betreff: Please help
Hi Ticker,
I try to understand why ShapeSplitter complains about a polygon created by my new algo. Attached is a patch and an input file. Command: java -Dlog.config=mkgmap_log.props -ea -jar dist\mkgmap.jar f:\osm\test.osm.pbf My mkgmap_log.props contains .level=WARNING and I see this warning: WARN: uk.me.parabola.util.ShapeSplitter f:\osm\test.osm.pbf: Direction of shape and hole conflict; probably self-intersecting polygon this: fp=146086642 lp=146173971 area=19324813903 #=134 http://www.openstreetmap.org/?mlat=48.975820&mlon=13.355620&zoom=17 http://www.openstreetmap.org/?mlat=48.988628&mlon=13.389160&zoom=17 o ther: fp=146153055 lp=146161739 area=22313626 #=6 http://www.openstreetmap.org/?mlat=49.002597&mlon=13.353453&zoom=17 http://www.openstreetmap.org/?mlat=49.003250&mlon=13.352696&zoom=17
I don't see what's wrong here.
Gerd _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Ticker, did not yet try to understand the details. I used GPXCreator to analyse the shape before the split and the two parts which are causing the message. I did not understand why they cause a problem, so I wondered if the split algo is wrong or if there is still a bad intersection in the shape produced by the mp-cutter. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 1. Februar 2017 19:05:42 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] WG: Please help Hi Gerd Did the other comment help you find the problem - it occurred to me later that I was dumping the data on input to the crashing ShapeSplitter and something had probably clipped the shape between multiPolygon generating it and this division. Also my plot shows the area rotated by 90 degrees (Latitude as X = horizontal) Ticker On Wed, 2017-02-01 at 16:39 +0000, Gerd Petermann wrote:
Hi Ticker,
if diagonal cuts cause problems I can stop this.
Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Mittwoch, 1. Februar 2017 17:04:36 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] WG: Please help
Hi Gerd
I've plotted the outer data. Not quite well enough to see all the directions of lines and if they emerge into each hole without crossing, or any slight overlaps, however:
On the bottom edge, there are two areas in the middle that look as if they should be indentations into the overall area but also have a bottom edge that follows the bounding box
Probably not a problem - on the left edge there are two holes where the top one is closer to the edge than the hole it is joined from
Are diagonal cuts OK - on some devices you can see where this is happening (for some reason the original OSM data did this)
Let me know if you want me to look in more detail.
Ticker
On Wed, 2017-02-01 at 15:24 +0000, Gerd Petermann wrote:
Hi Ticker,
I tried to reach you with a PM but that was returned, so I try again via list...
Gerd
________________________________________ Von: Gerd Petermann Gesendet: Mittwoch, 1. Februar 2017 16:09 An: Ticker Berkin Betreff: Please help
Hi Ticker,
I try to understand why ShapeSplitter complains about a polygon created by my new algo. Attached is a patch and an input file. Command: java -Dlog.config=mkgmap_log.props -ea -jar dist\mkgmap.jar f:\osm\test.osm.pbf My mkgmap_log.props contains .level=WARNING and I see this warning: WARN: uk.me.parabola.util.ShapeSplitter f:\osm\test.osm.pbf: Direction of shape and hole conflict; probably self-intersecting polygon this: fp=146086642 lp=146173971 area=19324813903 #=134 http://www.openstreetmap.org/?mlat=48.975820&mlon=13.355620&zoom=17 http://www.openstreetmap.org/?mlat=48.988628&mlon=13.389160&zoom=17 o ther: fp=146153055 lp=146161739 area=22313626 #=6 http://www.openstreetmap.org/?mlat=49.002597&mlon=13.353453&zoom=17 http://www.openstreetmap.org/?mlat=49.003250&mlon=13.352696&zoom=17
I don't see what's wrong here.
Gerd _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
participants (2)
-
Gerd Petermann -
Ticker Berkin