osm-ToniE
commented
3 weeks ago Thanks, there is actually a bug, but not related to handling several GTFS files.
Bug: inside a single GTFS file, PTNA does not correctly calculate the earliest day when a trip or route provides service to the public.
What you might report here is related to multiple GTFS files though:
- gtfs:feed=FR-BRE-Star being set and gtfs:release_date not being set here: results in PTNA taking the latest available GTFS data at PTNA, not available at the owner of the GTFS data.
- the GTFS data at the owner might be updated on a daily basis. PTNA does not support daily updates. PTNA takes one (the first) GTFS update of a month. However, I can manually update at any time: on request.
For the reported "C2" GTFS route:
Analysis:
- GTFS data version2024-20-28: the earliest day of service is reported as
- 2024-11-21 on GTFS overview
- 2024-11-21 on GTFS route
- 2024-11-04 on GTFS trip # 1 and GTFS trip # 2 which is not the same as for the two dates above (bug)
So for this GTFS data version 2024-10-28 best trip is still in the future: 2024-11-04, the report in the image above (made based on GTFS data as of 2010-28) is correct, despite the error in the calculation.
However for the GTFS data version 2024-10-31 (just updated that), the earliest day is given as 2024-10-29 but calculated as 2024-11-06 (bug) which should have an impact on the image above.
When there are several gtfs files, you should choose the file that covers the current day