End of ways inside area's

[koch-t]

See this situation inside OSM -> http://www.openstreetmap.org/?lat=52.387318&lon=4.638266&zoom=18&layers=M

There are a bunch of unconnected way's ending in the middle of the square/plaza including the busstation. This is done by a lot of OSM'ers who presume this is OK, but ofcourse is a huge pain in the ass when trying to route.

It could be possible to connect the end points with the square geometries but would it result in unintended behaviour?
At this point it often results in detours because it can't walk to the busstation from the trainstation directly.

[clinct]

Does the problem below have to do with this issue as well?

Planning from a location near the train station of Delft will result in a route where you first take a bus going from one side of the station to another and only then let's you walk to the specific track you need to be: http://demo.planner.plannerstack.com/?module=planner&fromPlace=52.00766382326151,4.360456466674805&toPlace=51.93742594489926,4.475780725479126&time=8:00pm&date=01-22-2015&mode=TRANSIT,WALK&maxWalkDistance=804.672&arriveBy=false&wheelchair=false&showIntermediateStops=false&itinIndex=0

This screenshot shows what's being returned, the blue line indicates the expected walking route.

Checking the debug layers shows that the walking route is there, but that it ends in a walk area:

[abyrd]

Those debug layers are a great help. Looking at the Delft station in OSM it is quite a clutter of areas and ways. I guess the walkable path that you expect it to use is an area-routing edge generated by OTP because there are no simple ways passing through the bicycle parking area. Judging by the result OTP does return, the Delft station transit stops are linked directly to one end of Coenderstraat instead of the platform areas and the pedestrian underpass on the north end of the station. The first question I'd ask is why those platforms are not present in the OTP graph -- they do share nodes with the bike parking pedestrian area, which is present.

[clinct]

It's indeed a mess, but also in real life ;) New train station will be opened within a month or so. Guess we should organise an OSM mapping party by then.

Just to be sure that it is clear what the expected walking route would be:

• Walking on the green path, then jumping across the walking area to the smaller green path, then walking up the stairs, walking down the stairs and onto the platform.

Concerning the missing platforms, you mean the train platforms?

[abyrd]

On Thu, Jan 22, 2015, at 14:38, Jasper Hartong wrote:

It's indeed a mess, but also in real life ;) New train station will be
opened within a month or so. Guess we should organise an OSM mapping
party by then.

That's actually a good idea. Maybe contact the people who entered the
existing objects and we can clean it all up after the station opens.

[abyrd]

On Thu, Jan 22, 2015, at 14:38, Jasper Hartong wrote:

Concerning the missing platforms, you mean the train platforms?

Yes, the train station platforms do not show up in the debug layer, and
the GTFS stops should be connected to them. To take the path you are
describing, you'd walk a short distance along the train platform to get
to the elevated walkway and stairs.

[abyrd]

The platforms are tagged area=yes railway=platform, and are meticulously connected to neighboring areas and stairs. Not sure why they have not been edge-ized.

[clinct]

This could be similar then to an issue I'd seen in Utrecht Station, edges for both the stairs and platforms are missing in the dashed blue circle.

We're going to look at cleaning up the Delft OSM, could be a nice team building activity :)

[andreyz]

This is probably related to #1776

[abyrd]

We certainly want to retain platforms and properly connect roads to areas. But we have new issues for tracking progress on related points in OTP2, including #3152 and #3201. Last time we discussed this we were hoping to rewrite the area routing code.