Hi all,
First of all I just wanted to say how great this dataset is. It is an excellent addition to the open data in the UK. Thanks to everyone involved.
TAP/TSI messages are a huge contrast to Sweden where all details about consists are offered in a simple free-text field:
E.g. "Vagnsordning Lok-[9-8-7]-6-5-4-3-2"
Finland has its own proprietary JSON format, and Norway uses SIRI-SM XML messages with a TrainFormation extension schema.
Anyway, I'm not always sure if I interpret the data correctly. In particular, the order of vehicles in consist.
Ordering logic I'm using is currently like this (in Postgresql):
ORDER BY
AllocationOriginDateTime,
ResourcePosition,
CASE WHEN Reversed='Y' THEN ResourceGroupPosition END DESC,
CASE WHEN Reversed='N' THEN ResourceGroupPosition END ASC
Looking at the 23:45 Paddington–Penzance 1C50, the result of the query is:
<- 57301 - 10563 - 10596 - 10534 - 10589 - 10616 - 10219 - 12142 - 17173 - 57312
Rear loco leaves them at Reading. This seems correct.
Then, looking at the 17:22 Cardiff Central–Holyhead 1W96, there appears to be a DVT in a very odd place, in between vehicles, while typically DVTs should be at the either end.
This is the order I've interpreted for 1W96 leaving Cardiff Central:
<- 12434 - 12310 - 12211 - 82201 - 11319 - 10318 - 67022
Then the train changes direction at Chester, the next leg has Reversed=Y, and the consist is reversed to reflect that. That seems to be correctly put in.
I don't remember which way round the loco is when the train leaves Cardiff. Last time I was on that service it was all Mk3s. It would be helpful to know. :)
Also, noted that TfW's Coach letters look weird. For this train: C - F - B - P - L - H - (Cl67). ScotRail's HST's have Coach letters that are always obviously wrong - maybe the letters have not been updated for a while?
Kind Regards,
Juhani