[Koha-bugs] [Bug 18963] Some checkins have been moved to old_issues without a returndate

[bugzilla-daemon at bugs.koha-community.org]

https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=18963

Ray Delahunty <r.delahunty at arts.ac.uk> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |r.delahunty at arts.ac.uk

--- Comment #4 from Ray Delahunty <r.delahunty at arts.ac.uk> ---
Not sure if these comments help but…

This problem is a very serious one for us as from Wednesday 1 November
absolutely zero lines in the old_issues table have the returndate field
populated. And overall, we now have upwards of 20,000 items showing up in the
Circulation History (both OPAC and Koha intranet) as “issued” when in reality
they have long since been returned (including a period when we were on 16.05).
This is very confusing for both staff and students. The problem appears to have
got gradually worse over many months. I ran the simple query SELECT itemnumber,
timestamp, returndate FROM old_issues WHERE timestamp > '2015-08-30%' ORDER BY
timestamp ASC and browsed down for empty returndate fields. (UAL migrated 3
million lines of historic loans from out old LMS and want to review the problem
from Day 1 of Koha only, hence the August 2015 limit).

At first I see a small number of problems appearing after the end of many of
the day’s opening hours. Even back on 16.05 in June of this year we have
transactions occurring in the middle of the night when the library is closed. I
have no idea where these are coming from- the building is locked so users don’t
have access even to the 3M returns chute (use of which would remove the item
from the account). Oddly a lot of the lines are all timestamped in the 23:xx
range and all are without a returndate.  

The problem gets steadily worse as the weeks go by, and my take on it is that
the problem explodes in 17.05.01 (8/2017 for us) where the count of the lines
rockets. By the 4th October we are seeing upwards of 100 lines appearing, all
timestamped in the 22:xx range, and all with no returndate. The library is
closed at that time. And by 30th we have 1,205 lines with a timestamp (all now
in *both* the 22:xx and 23:xx ranges) but no returndate, again at a time when
the building is locked up for the night. The last line in the table with a
returndate in is Tuesday 31 October at 21:55 when the library physically
closed. Lines in the table are timestamped all through the (now 24 hour a day)
opening time, but there are no return dates. We have currently 15,000
sequential lines without the returndate and this is making Circulation History
a mess. 

Would a crude fix be to copy the timestamp over the returndate where the return
date is empty? I can of course add the 737,000 line spreadsheet containing the
data I have reviewed if necessary.

Ray Delahunty
University of the Arts London

-- 
You are receiving this mail because:
You are watching all bug changes.
You are the assignee for the bug.