Hi Andrzej,

there was a stupid bug in the code, it did not always create routing nodes for 
the added nodes.
I've also changed the code in WrongAngleFixer so that these messages should no 
longer be printed unless something really goes wrong.

I've tested with a smaller set (just Luxembourg and Saarland) and it seems to 
work fine now.

I'll now look at the code that calculates the position of the intersection. For 
now, I copied code from JOSM, but I'd prefer to make old
code from WanMil work for this.

Gerd

________________________________________
Von: mkgmap-dev <[email protected]> im Auftrag von Andrzej 
Popowski <[email protected]>
Gesendet: Mittwoch, 18. Juli 2018 15:00
An: [email protected]
Betreff: Re: [mkgmap-dev] new branch country-borders

Hi Gerd,

yes, there are messages like this:
SEVERE (WrongAngleFixer): pbf\be\71282723.osm.pbf: country boundary node
is replaced by node with non-equal coordinates at
http://www.openstreetmap.org/?mlat=50.920098&mlon=2.593385&zoom=17
SEVERE (WrongAngleFixer): pbf\be\71282721.osm.pbf: country boundary node
is replaced by node with non-equal coordinates at
http://www.openstreetmap.org/?mlat=51.257172&mlon=4.086521&zoom=17

I don't know, how it influence routing. We should check in GPS too.

 > My understanding was that you would create different gmapsupp for each
 > coountry, with different FID etc.

Yes, the idea is that even maps created by different authors would get
matching external nodes. But you won't be able to test routing with
multiple maps on PC. My test actually creates 3 different maps, but
install them as a single item on PC.

--
Best regards,
Andrzej
_______________________________________________
mkgmap-dev mailing list
[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

Reply via email to