@jamesevans do you have any update on crowding data? the static datasets & PowerBI reports are all showing 31/08 still?
I suspect it would help the TfL team if you were to specify which data source(s) you are referring to e.g. Countdown, Unified API, BODS, TfL website, Countdown screens at bus stops etc…
Simon
Bus working timetables are back!
1 Like
Hi @mjcarchive - I was told my colleagues had enabled some batch jobs the other day, but didn’t realise this was one of them.
Many thanks,
James
@jamesevans
I think the upload happened yesterday. It is reported that predictions are back for the routes where they were missing, which is presumably related. The usual issue with the 481D MFSc schedule overwriting the 481 MFSc schedule as part of the process but we are talking gnats compared with elephants there. Christmas has come early, in time for all the Christmas specials to be loaded…
Hi, it looks like the BODS team have now addressed this issue:
1 Like
@jamesevans
Just to note that you can remove the strikeout fonting from Bus Working Timetable updates in the description of outputs affected by the cyberattack. NRather appropriately, no sooner did we have 600+ new ones last week than we had about 300 more loaded at the usual time last Sunday.
1 Like
Hello Leon
Yes that seems to be fixed, thank you! However when looking at the BODS dataset just now, the OriginAimedDepartureTime values tend to be throughout the day.
Look for example the picture below that includes all the buses running for route 12 in London. I have mapped their locations + show their OriginAimedDepartureTime as a label. I took the data at around 11:30 this morning, as you can see, there are many that are set to depart much later in the day.
Second Saturday in a row with no predictions for route 108…
@SJCooper @jamesevans
I can’t help wondering whether the 108’s issues are to do with the existence of a 108D (the double deck element) with the same Sa daytype. Both working timetables get loaded to the WTT site as 108 Sa but my suspicion is that the main one (108) gets overwritten almost instantaneously by the 108D file. This also affects Su but not the various weekday daytypes. The 481 MFSc has a similar issue; I might check whether that is giving predictions tomorrow.
Of course a file not being available to punters does not necessarily mean that
it is unavailable to the prediction system but it’s worth considering.
I have been banging on about this built in ambiguation (if that is the right word) which has been there since the WTT format changes a couple of years ago. Perhaps it can now be looked at. If it is impossible to solve without digging up the system perhaps the main files for the offending route/daytype combos can be made available some other way. FOI is a possibility, I suppose, but that consumes far more resource.
@SJCooper @jamesevans
481 is reporting OK today (MFSc) so that’s another theory exploded.
The general point about two files sharing the same name remains valid though.