Hi Andrzej, yes, mkgmap evaluates the tag "junction=roundabout", no matter what type the road gets. In trunk this is only used to decide if check-roundabout must be used, in the branch it is used to group all roundabouts at the beginning of table A. That's why I hoped that the type 0xc is not needed to get hints for the right exit point,but it did not work. Steve is looking at this. I'll add a check to make sure that class is in the range of 0-4 and speed goes from 0-7 . Gerd
Date: Sun, 9 Mar 2014 15:53:37 +0100 From: popej@poczta.onet.pl To: mkgmap-dev@lists.mkgmap.org.uk Subject: Re: [mkgmap-dev] new branch NOD127
Hi Gerd,
If I comment the roundabout block in the lines file
See my first mail, I have tested roundabouts. Commenting the roundabout block clears problem, but if you instead change objects 0x0c into some other road, then problem remains.
More precisely, I think route is correct, when roundabout get the same object as main road, but if roundabout is a different object (not necessary 0x0C), then calculation is wrong. Maybe reason is a turn angle?
In my test detours contain roundabouts too. And if I set avoidance to roundabouts in BaseCamp, then detour changes, which suggest that roundabouts aren't the main reason of detour.
I think mkgmap processes tag "junction=roundabout", even if there is no roundabout object 0x0C. After commenting the roundabout block, avoidance in BaseCamp still works. But avoidance doesn't work with following line in style:
junction=roundabout & highway=* {delete junction}
Side note: there is no verification of "class=" value. If I set "class=6" then mkgmap crashes with array index out of bounds exception.
-- Best regards, Andrzej _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev