Why does Vehicle Locations trip_stop_sequence not match the data type for GTFS-RT stop_sequence?

[ohrite]

Hi folks!

First, thanks for the excellent specification work you've done so far. We're working on the first reference implementation of TIDES, and we're looking at ways of streamlining conversion of GTFS-RT data into the appropriate format.

When we were reading the spec for Vehicle Locations, we saw that trip_stop_sequence has a minimum value of 1. However, GTFS-RT consistently specifies the uint32 (unsigned 32 bit integer) data type for stop_sequence/current_stop_sequence. We're wondering if there's a story here, since it seems like a minimum value of 0 might be more appropriate to represent a uint32 type.

Thank you!

[lauriemerrell]

I will cross reference some commentary from here: #252 (comment), where @jlstpaul noted that compatibility with GTFS-RT is not a strict requirement in and of itself for TIDES.

However, in this case, agree that compatibility makes sense and I'm not sure why 0 is not allowed

[lauriemerrell]

Actually, the docs kind of allude to an answer for this IMO... I guess trip stop sequence is supposed to be inferred from the actual observed order (like actually incrementing by one in order throughout the trip) and is in fact distinct from scheduled stop sequence which can be 0-indexed and doesn't have to increment by one

[botanize]

That is correct, "scheduled_stop_sequence" is the field that directly matches the meaning and data type of GTFS-realtime stop_sequence/current_stop_sequence.

[lauriemerrell]

I think this might be an ambiguity in the vehicle positions GTFS RT docs -- I would assume that the current stop sequence in GTFS-RT vehicle positions is referring to GTFS schedule? But then it is confusing that it is not called out in those docs as it is for stop times updates

[lauriemerrell]

But I'm looking at a GTFS-RT vehicle positions feed right now from the MBTA and I'm inferring that they at least are using vehicle positions current_stop_sequence to match GTFS schedule (I'm seeing values as high as 650, I'm just going on the assumption that they don't actually have trips making 650 stops)

[lauriemerrell]

I have submitted the intra-GTFS-RT issue here: google/transit#642 because I think this is worth clarifying.