I have waited several months before posting about this problem in the hope that it would be permanently sorted. In fact it was sorted three weeks ago, only for it to rear its ugly head the folloing week.
TfL publishes bus working timetables (schedules) at https://tfl.gov.uk/corporate/publications-and-reports/bus-schedules, a very welcome development. Typically about 3,200 schedules are available at any one time, though this includes some scheduels which are obsolete but never needed to eb overwritten.
The schedules are clearly recreated every week from some source (judging from the file creation dates in the relevant Data Bucket at http://bus.data.tfl.gov.uk/).
Unfortunately at least 450 (14% plus) of the schedules are currently incorrect. By that I mean that they were replaced at some stage but have managed to get reinstated, overwriting the later versions, or that what should be the current schedules simply have not been loaded.
This has been going on for months. In some cases the schedule available for download had repeatedly oscillated between the correct and one, sometimes two, older versions.
I have pointed this out repeatedly to the web team by e-mail. Although I have never received anything other than an automated reply, this may have had some effect as the 3rd June upload basically corrected everything.
Move on a fortnight to the 18th June update and the problem was back on an even bigger scale.
This week (the 24th June update) nothing was corrected as such and even more incorrect files were reloaded.
This is written up more fully at http://timetablegraveyard.co.uk/new_wtts.html and I will not repeat it all here. Amongst the current lowlights are the 25 apparently still running to Oxford Circus and the N72 via the clsoed Hammersmith Bridge. The new 46 schedules loaded just last week have been overwritten with something older.
On the face of it, this is a simple failure of version control. There must be an “expired” marker (otherwise even more routes would be involved) but either it is not always being set properly or the file creation process is failing to use it correctly.
It clearly can be done right (as on 3rd June) so why is it not always being done right?
I have heard it said that the online schedules are useful to service controllers. If this is true they are using a set of schedules with an alarmingly high error rate.
I appreciate that this a niche interest and also about what is essentially a static source but I would really like someone to take ownership of this issue.