Having found that the data bucket http://bus.data.tfl.gov.uk/ that I use came to a juddering halt after the WTTs for route 187, I checked with the more regular way in to the WTTs, namely Bus schedules - Transport for London.
Sure enough, the schedule for 187 are there but those for 188 and above are missing, so the data bucket is just reporting the reality.
Can this be looked at please?
I had been hoping that this would be sorted overnight but no, still mostly missing.
For clarification, it is not just 188 and above that are missing but also 2 and above, 19 and above and everything with a text letter prefix. We have less than one hundred routes with WTTs.
Any chance of an acknowlegement (or, even better, solution)?
Seems like everyone is too busy to respond that they are too busy to do anything about it.
hi @mjcarchive - apologies I’ve not had time to look into this yet. Hopefully Monday will be quieter!
Thanks James. It is possible that the next update will have occurred by then and - who knows? - it might sort itself out.
You’re in luck, James. That’s exactly what happened early this morning. Seemingly all present and correct again.
I obviously have psychic powers…
Though you could always try applying them to the 353, which still, months after the change of format, has four files of 15B each (so obviously naff). Or the 143 MFSc schedule (which returns just the double deck 143D schedule) or the N29 FrNt and SaNt (which return just the extras VN29).
The second problem is presumably about duplicate file names, with both being created but (for example) 143D overwriting 143. If the new regime had been in force when loads of routes had Covid extras with a V prefix, far more than three of the regular schedules would have become unavailable. In principle this seems a bigger problem than the files for one route for some reason not being created properly.
Just to note that the 353 WTTs are now OK, thanks to the timetable having changed ratgher than correction of the old naff files.
The 143/143D and N29/VN29 issues remain unresolved.
This week’s update has brought us WTTs for the 481D - double deck school day journeys. Unfortunately, and entirely predictably, this means that the school day schedules for the main 481 service have disappeared down the plughole. I say “entirely predictably” because this mirrors what had already happened for the 143/143D and N29/VN29.
Seems to me someone simply did not think through the file naming convention for the new format files. Somehow I doubt that the problems I am seeing are the only ones with the new system; I suspect there may be other more, probably more important, issues which are not visible from the outside.
Another update. N29 is now OK, presumably because the VN29 extras have been withdrawn. 143/143D and 481/481D are still wrong.
… but now the VN29 extras are back.
Could I make a suggestion? Would it not be possible to use a different day type code for additional schedules (MFScEx, FrNtEx etc)? Then both would be shown in the listing for that route on the TfL site, rather than the more important one being lost.
New day type codes can clearly be added, as CxSa has recently appeared for the post-Christmas working day schedules, which used to appear as sMT, sMF, sSa or whatever.
We have a recurrence of the issue with the WTT update falling over part way through. Thus happened last week but was correct by the following day (Monday). When I saw the incomplete nature of the update yesterday I thought it would be corrected this morning but presumably it has not been spotted, hence the need to report it now.
Specifically, the Data Bucket links for routes 206 and above have not been updated. The links for routes 206 to 212 have been lost altogether from the bucket. For higher number routes (including, for example, 3 and 21 because it is an alpha sort) either the daytime links (which is all links for most routes) have disappeared or just the night ones are there, though even those no longer work.
The last point illustrates that it is the actual file set which has been messed up, not just the data bucket or the links to the files on the TfL website.
The only change this morning this morning so far is that the links that don’t work have been wiped from the Data Bucket, which is consistent but not a lot of use.
So, all that is available as of now is the daytime WTTs for routes 1-2, 10-20, 100-205. Nothing else.
Any chance of it being sorted please? Pretty please?
@jamesevans @neamanshafiq @GerardButler
No change this morning. Even if it can’t be sorted quickly, could someone please provide a progress report, ort at least some sort of acknowledgement?
On top of that the usual upload of a zip file of Journey Planner data to Datastore appears not to have taken place. It typically happens some time on a Tuesday.
The apparent failure of two information sources makes me wonder whether something more fundamental is going wrong but without any feedback that is just speculation.
@mjcarchive we think there might be an issue with the system that exports the timetables. We do have an automated task which is deletes files older than 7 days - I have disabled this for now to rule out if this is causing any problems. A transfer of new files will happen at 6am tomorrow morning so I’m hoping we’ll have the full dataset.
Re Journey Planner data - I believe there was an issue this week affecting the export of the data. i can see a file has been uploaded now so hopefully the datastore will be able to show this.
Thanks, Neaman. My guess would be that the automated task can’t itself be to blame as it would not delete files if they had been updated (as they should have been) a few hours or whatever previously. In a way disabling it conceals the problem, though it would leave a file for the link to access. It just wouldn’t necessarily be the one it should be, if a real service revision had taken place. Fingers crossed for tomorrow.
I can’t yet see the new Journey Planner file on either of the links which usually access it, that is
Journey Planner file now available, files within it dated up to 1800 hours.
Fingers crossed for tomorrow morning.
Still the same. Did the process run and fail or did it never run at all?
This is a particularly bad time for things to fall over as a large number of new seasonal schedules is due. I trust that that itself is not part of the problem.