duplicate train marker

[jendrikw]

https://api.transitous.org/?tripId=20251217_21%3A44_de-DELFI_3060997859

and

https://api.transitous.org/?tripId=20251217_21%3A44_de-VBN_472481302

are the same train, but with different real-time information (although they differ only by one minute at most at the time of writing).

You can't find them by searching for a route, you have to click on the map markers.

[jendrikw]

Same thing happens with RE 60 / RE 70 between Braunschweig and Hannover. VBN has shapes, but DELFI has the better name and color

[jbruechert]

The DELFI color is injected by a lua script, we could do that with VBN as well.

What exactly would you like to see here? In my opinion, this can mostly be solved with better fuzzy merging in MOTIS. The only alternative is to rely only on DELFI, but the same problem also exists with trips from the Dutch, Swiss etc. datasets.

[jendrikw]

You're right, this probably belongs to MOTIS. I will open an issue over there.

I'll also add a lua script for VBN.

[felixguendling]

I would say that in this case, MOTIS works as expected.

The stop positions are more than link_stop_distance (which is configured to 100m in Transitous) apart.

Since this is also used to create footpaths and meta stop, we probably need another method and can't use the equivalences we use elsewhere.