Hi, When I am going south on Davis Dr and want to merge into US-64 West, the GPS and MapSource choose to turn left into Old Jenks Rd, right into N Salem St and then get the ramp to enter US 64W instead of the direct route from Davis Dr, turn right into the ramp. Any ideas why this happens? I don't see anything wrong in the OSM data. Here's the area in OSM: http://www.openstreetmap.org/?lat=35.7558&lon=-78.85157&zoom=16&layers=M Francisco
On Fri, Sep 09, 2011 at 09:04:14AM -0400, Francisco Moraes wrote:
When I am going south on Davis Dr and want to merge into US-64 West, the GPS and MapSource choose to turn left into Old Jenks Rd, right into N Salem St and then get the ramp to enter US 64W instead of the direct route from Davis Dr, turn right into the ramp. Any ideas why this happens? I don't see anything wrong in the OSM data.
The data looks OK to me too. There are not even any turn restrictions, and the only oneway roads are the ramps and the US64W.
Here's the area in OSM:
http://www.openstreetmap.org/?lat=35.7558&lon=-78.85157&zoom=16&layers=M
Is this just a Garmin bug? Can you try routing a shorter distance? For example, start from the Davis Dr close to the Old Jenkins Rd and ask for a route to the US64W ramp. On a related note, I don't know if this is a useful trick for your GPS receiver, but when I keep my Edge 705 in the basement (away from windows) it will eventually give up its search for satellites and offer an option to choose the location on the map. If I do that and enter demo mode, I can play around with the routing. One more thing: In 2010, Mark Burton implemented support for a through_route relation that could be useful in cases like this. I do not know the specifics; I only know that bicycle routing generally produces stupid routes because Garmin is cheating (ignoring cycleways) in the middle part of long routes. There is a route=bicycle relation on the Davis Drive, but it should not affect the routing behavior, as far as I understand. Best regards, Marko
On 2:59 PM, Marko Mäkelä wrote:
Is this just a Garmin bug? Can you try routing a shorter distance? For example, start from the Davis Dr close to the Old Jenkins Rd and ask for a route to the US64W ramp. I tried this. There several points south of Old Jenks Rd that still want to do that funny detour to get to the ramp. Only when I get close to that service road marked with access=private that it starts to route normally. I am thinking of trying to change that service road and see if it makes any difference. One more thing: In 2010, Mark Burton implemented support for a through_route relation that could be useful in cases like this. I do not know the specifics; I only know that bicycle routing generally produces stupid routes because Garmin is cheating (ignoring cycleways) in the middle part of long routes. There is a route=bicycle relation on the Davis Drive, but it should not affect the routing behavior, as far as I understand. Well, I hope someone else has a different idea on this one. I tried with a Garmin map and there is no such a problem, so the issue is exclusively on the OSM map created via mkgmap.
Thanks for the reply. Francisco
participants (2)
-
Francisco Moraes -
Marko Mäkelä