Routing issue on N10 (France)

Hello,

I think I have found an issue with the route calculated by Kurviger
I read the corresponding documentation.

Here is a url to a short example that shows the issue:

If you reverse the Kurviger route, routing is now ok.
OSM data seems to be ok (not 100% sure though :slight_smile: ).

Routing is also correct with openrouteservice.

Thanks for letting us know. I think there is a turn restriction that you changed recently?

Maybe the issue was this turn restrictions. Please note, it takes a few days for Kurviger to pick-up the new data, it’s not updated immediately.

Yes, I fixed some turn restrictions yesterday. But routing was already ok with openrouteservice and osm/direction before my changes. Anyway, we’ll wait a few days for Kurviger to pick-up the new data and we’ll check if my changes fix this issue.
Any clue on when the new data will be available on Kurviger?

Did you make sure that they use turn restrictions? Because some services don’t use turn restrictions or they need to be enabled before. If you change the routing engine in OSM, the same detour appears: https://www.openstreetmap.org/directions?engine=graphhopper_car&route=48.01786%2C1.28195%3B48.01680%2C1.27905#map=16/48.0124/1.2852

Anyway, let’s wait until the data is updated, if the issue persist, we can think check what might be wrong here.

Hi @boldtrn,

Do you know if OSM data have been updated on Kurviger? The routing issue is now solved on Kurviger, but OSM with the GraphHopper routing engine still have the initial routing issue. Could this be a bug with GrahHopper rather than OSM data?

You can check that in website using the menu “Date OSM database”.
Currently is shows “Routing: 2020-08-04T11:00:00Z”.

Do you mean the GraphHopper offline routing?

Thank you. So OSM data are up to date on Kurgiver.

No, I mean the GraphHopper routing engine used on the https://www.openstreetmap.org/directions website. Please compare the same route which give different results with OSMR and GraphHopper.

I was just curious whether the routing issue was fixed on Kurviger with my change on OSM data to fix turn restrictions, or something else.