Hi Andrzej, Yes, very long lines are no problem regarding optics, but may cause trouble when the bresenham algo iterates very often. Or do you see a simple way to limit the iteration to a part of the way near the wanted point?
My current algo tries to split the line when the address search for a house returns a point that is more than 40 m away from the closest possible point. If that is not possible without visible distortion, it may also add an empty segment. Maybe 20m are also reasonable, but a much smaller value doesn't improve much while it will incease img size a lot. Maybe I add an option to configure the threshold values. Anyway, it seems that there is no simple solution so I'll continue with what I have. A first usable version should be ready within the next days. It will also support the typically random numbers from addr:place=* tags. Gerd popej wrote > Hi Gerd, > > I have attached picture with 2 lines split around middle point. Upper > line is about 50m, lower 500m. Both have delta_lat equal to 1 grid step. > I don't think that splitting make big problem here. > > If you allow for 10m of offset for splitting point, then Bresenham's > algorithm has to calculate only 8 points. For me it looks like a good > solution. > > Are you going to insert multiple points into a line? > > -- > Best regards, > Andrzej > > > _______________________________________________ > mkgmap-dev mailing list > [email protected] > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev > > split.png (16K) > <http://gis.19327.n5.nabble.com/attachment/5833378/0/split.png> -- View this message in context: http://gis.19327.n5.nabble.com/help-needed-for-graphical-problem-tp5833271p5833386.html Sent from the Mkgmap Development mailing list archive at Nabble.com. _______________________________________________ mkgmap-dev mailing list [email protected] http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
