Talk:Proposed features/through service

From OpenStreetMap Wiki
Jump to: navigation, search

Through routes with an overlapping segment

I have mapped some train routes ABCD where the route is described in two separate timetables ABC and BCD with different reference numbers, and where trains and platforms display the destination as D all the way from A, but no route numbers. So far, I've used semicolons in the ref= (simple example, more complex example), but something like this proposal might be a better idea, with one ABC and one BCD relation and a superrelation of the other two (although I think the type= of the superrelation should be named something else than route). Would it be enough for hypothetical applications to realize that an overlap exists by comparing the relations, or should this be tagged more explicitly? After all, even in the examples given here, the final stop of the first route will be the initial stop of the second route and thus included in both relations. //Essin (talk) 18:08, 27 April 2017 (UTC)

In these cases I mainly relied on when the vehicle changes the route display, but in your case as it is not available, I would consider what numbers are printed on the ticket if you try to buy one from B to C, A to C or B to D, etc. Maybe after the voting we can discuss how to deal with these cases. -Miklcct (talk) 23:35, 27 April 2017 (UTC).
Tickets don't solve my case, it's a zonal fare system and no tickets are sold on the trains. The only reference numbers for the entire stretch I know are the numbers of individual train runs, but listing them in OSM would lead to sets of ~20 relations with the same members which also probably would infringe on the database rights of the timetable.
I think it would strengthen the proposal to consider applications, since one of the criticisms of the Public Transport schema has been that it requires unnecessary processing downstream. (I lack the coding skills to judge how well-deserved that criticism is though, and I generally just try to implement the PT schema as faithfully as possible.) //Essin (talk) 19:35, 2 May 2017 (UTC)