Hi, first of all, THANK YOU for the great new navigation over public transport lines!
That is a great addition to the already existing navigation modes.
Now, I noticed this problem. When the start and end point of the wanted route is away from a public transport stop (say a bus stop) there are computed segments of the route where you have to go by foot and OsmAnd correctly uses the existing footways and roads to get to the bus stop. However, the computed foot route goes to the public_transport=stop_position node of the bus stop. That is where the vehicle stops, not where the passengers need to go. To achieve this node, OsmAnd calculates needlessly complicated route with diverts and going over roads. It should actually route to the public_transport=platform node of that bus stop.
Also notice on OSM wiki, that public_transport=stop_position nodes are optional and may not exist in the public transport relations you use for routing (even in PTv2). So using the public_transport=platform nodes is necessary either way.
Yes, this would create a physical gap between the spot where the passenger arrives and where the vehicle departs and continues along its segment of the route along roads.
You just need to solve that somehow visually, as otherwise there is no gap logically (in the OSM model).
Of course this also applies to places where you exit a vehicle of public transport. The route by foot should continue from a platform node, not stop_position on the road.
Hi, first of all, THANK YOU for the great new navigation over public transport lines! That is a great addition to the already existing navigation modes.
Now, I noticed this problem. When the start and end point of the wanted route is away from a public transport stop (say a bus stop) there are computed segments of the route where you have to go by foot and OsmAnd correctly uses the existing footways and roads to get to the bus stop. However, the computed foot route goes to the public_transport=stop_position node of the bus stop. That is where the vehicle stops, not where the passengers need to go. To achieve this node, OsmAnd calculates needlessly complicated route with diverts and going over roads. It should actually route to the public_transport=platform node of that bus stop.
Also notice on OSM wiki, that public_transport=stop_position nodes are optional and may not exist in the public transport relations you use for routing (even in PTv2). So using the public_transport=platform nodes is necessary either way.
Yes, this would create a physical gap between the spot where the passenger arrives and where the vehicle departs and continues along its segment of the route along roads. You just need to solve that somehow visually, as otherwise there is no gap logically (in the OSM model).
Thanks.