Darwin:Gotchas

From Open Rail Data Wiki

A list of various non-obvious things Darwin (and LDBWS, in some cases) clients need to consider:

  • Trains can have multiple associations at the same station.
    • For example, the Inverness, Fort William, and Aberdeen portions of the Caledonian Sleeper all merge at Edinburgh.
  • Passenger-relevant associations may occur at non-passenger stations.
    • Both the Highland and Lowland Caledonian Sleepers do this, at Edinburgh and Carstairs respectively - though, on the highland sleeper, Edinburgh is a passenger stop for the Fort William portion only.
  • An intermediate station can become a destination if the train is cancelled short. This means:
    • If you only display one time for the stop, it should be the arrival time, not the (no longer meaningful) departure time.
    • Any future TS messages will only be identified by the pta/wta, not the ptd/wtd.
    • I haven't encountered this, but similar considerations likely apply for stations becoming origins if trains start short.
  • Not all reference data is complete, and will cause you problems if you base foreign key constraints on them
    • In particular, you will receive messages from CISes not listed in CISSource elements (e.g. NA01 and XRRC), and from operators not included in TocRef elements (e.g. LS)
    • LocationRef, in contrast, is a complete list of valid TIPLOCs