Hi Ticker, okay, I wait for yor results. Besides that I think the de-duplication should better be done in StyledConverter. Gerd ________________________________________ Von: Ticker Berkin <RIBerkin@jagIT.co.uk> Gesendet: Donnerstag, 2. Februar 2017 11:36:31 An: GPetermann_muenchen@hotmail.com Betreff: Re: [mkgmap-dev] Problems with r3781 Hi Gerd Having electrical problem at the moment so don't have correct computer/sources/email system with correct from email for mkgmap dev list etc, so sending direct to you. Please reply to list. 1/ with the last patch, I changed the behaviour slightly of when lines and shape (unless OrderBy...) get shifted into their own MapArea/subDivision. When this happens, shapeMergeFilter won't merge them. I plan to think about this a bit more. 2/ Behaviour of when highPrecEqual coords get set to the same coord has changed slightly during shape splitting operations. Old logic tended to save a hash of all points, do the operation, then look up each point in the list and use that. New logic only does something similar for points exactly on the boundary (normally created by the split). Ticker -----Original Message----- From: Gerd Petermann Sent: Thursday, February 02, 2017 8:45 AM To: Development list for mkgmap Subject: Re: [mkgmap-dev] Problems with r3781 Hi Ticker, I did some more tests with r3782, also with other styles like that for openfietsmap.nl (by Minko aka ligfietser) http://mijndev.openstreetmap.nl/%7Eligfietser/openfietsmap/Scripts/Styles/ I've compared the results with trunk version r3773 and noticed that the branch produces fewer messages from ShapeMergeFilter like "ignoring duplicate shape with id 144075839 at http://www.openstreetmap.org/?mlat=54.606792&mlon=8.374727&zoom=17 with type 0x53 for resolution 22" and in fact the branch writes those duplicated shapes two times while trunk doesn't. The style adds two equal 0x53 polygons for the way 144075839 , one because it has natural=beach, one because it has surface=sand. The trunk version seems to filter those duplicated shapes while the branch doesn't. The result is an increased img size for this style, while the branch typically creates smaller files for the default style. Do you see why the this happens? 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 17:47:59 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] Problems with r3781 Hi Gerd Fixed this and committed - I think the problem was always there but excess MapArea splitting made it very unlikely to happen. There is still a little bit of old logic that I'm not sure about, but I've removed for now which is safe. Some time I'll think if something similar needs to be put back Ticker On Wed, 2017-02-01 at 11:32 +0000, Ticker Berkin wrote:
Hi Gerd
Sorry - I wasn't paying attention and missed the vital bit of your mail: northwest-territories-latest.osm.pbf
Will try with the correct file!
Ticker
On Wed, 2017-02-01 at 10:17 +0000, Ticker Berkin wrote:
Hi Gerd
Using your new areas.list file and Splitter version 573 compiled 2017-01-12T17:28:57+0000 on: 3094949146 Jan 13 10:00 germany-latest.osm.pbf
all the output files osm.pbf files are only 106 bytes long!
No obvious errors from the splitter.
I'll investigate further
Ticker
On Wed, 2017-02-01 at 09:17 +0000, Ticker Berkin wrote:
Hi Gerd
I'l check this out
Ticker
On Wed, 2017-02-01 at 09:11 +0000, Gerd Petermann wrote:
Hi Ticker,
I see some error messages when creating a map for northwest -territories-latest.osm.pbf: SCHW: uk.me.parabola.imgfmt.app.trergn.Polyline e:\osm_out_work\northwest -territories\20170123_113606\63240008.osm.pbf: Problem writing line (class uk.me.parabola.imgfmt.app.trergn.Polyline) of type 0x1d containing 3 points and starting at http://www.openstreetmap.org/?mlat=58.143592&mlon=-110.006011&z oo m= 17 SCHW: uk.me.parabola.imgfmt.app.trergn.Polyline e:\osm_out_work\northwest -territories\20170123_113606\63240008.osm.pbf: Subdivision shift is 0 and its centre is at http://www.openstreetmap.org/?mlat=58.538117&mlon=-110.791583&z oo m= 17 SCHW: uk.me.parabola.imgfmt.app.trergn.Polyline e:\osm_out_work\northwest -territories\20170123_113606\63240008.osm.pbf: deltaLong = 36610 SCHW: uk.me.parabola.imgfmt.app.trergn.Polyline e:\osm_out_work\northwest -territories\20170123_113606\63240008.osm.pbf: Problem writing line (class uk.me.parabola.imgfmt.app.trergn.Polyline) of type 0x1d containing 3 points and starting at http://www.openstreetmap.org/?mlat=58.143592&mlon=-110.006011&z oo m= 17 SCHW: uk.me.parabola.imgfmt.app.trergn.Polyline e:\osm_out_work\northwest -territories\20170123_113606\63240008.osm.pbf: Subdivision shift is 0 and its centre is at http://www.openstreetmap.org/?mlat=58.538117&mlon=-110.791583&z oo m= 17 ...
You should be able to reproduce it with this command: java -ea -jar d:\mkgmap\dist\mkgmap.jar 63240008.osm.pbf using the attached split file.
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
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