Ok, then. Since adding column ordering is trivial, thanks to QT, tomorrow I'll change sorting to jobid desc and enable column sorting. I've sent a patch that should fix the selection bug.
Regards, Riccardo PS: Good luck for the demo! Kern Sibbald wrote: > On Friday 16 May 2008 01:05:50 rghetta wrote: >> Yes, ordering by jobid would nicely show all current jobs first and also >> avoid using a multi-field ordering. >> Having a single field order could also allow making table headers >> clickable and sortable, like Marc proposed (with multi-field there's no >> way to return to the original ordering without using dirty tricks). >> I like it. > > Yes, I agree. It is always better to try to keep it simpler where we can. > >> One caveat: AFAIK ordering by jobid orders by queuing time, while the >> current ordering (by starttime,jobid) gives you jobs by effective start. >> I don't care much, but perhaps others have different needs. >> >> What do you think ? > > That sounds fine to me, but seeing it actually working is what is the > determining factor -- for me, it is often only then that I can actually "see" > all the ramifications. > >> Still, I wonder why you want to keep around old jobs who newer ran, >> instead of just deleting them. That would keep your display nice and >> clean :-) > > Well, that is a good question. I ususally never delete anything for > historical > reasons - the files generally do get deleted when their retention period is > up. However, in this case, since it is a dummy database and since I would > like to demo it looking relatively clean, I went ahead and deleted the > offending records, which brought up another oddity ... > > When I tried to delete a line in the JobList, it would respond that JobId 0 > was deleted. This happened over and over. Finally I double clicked on the > row, and then it deleted the correct JobId. I would appreciate if if you or > Dirk would look at this and see if it is working properly for you. One > should be able to simply select (single click) on one or more rows and the > select the "Delete one job ..." or whatever the menu items and all selected > jobs should be deleted -- if we implement multiple deletion as we should, I > think we need to change the menu item a bit ... > > For the original sorting question: since you are programming it and most > solutions seem reasonable, I will let you choose the one you prefer hopefully > permitting us to later add sorting by clicking on column headings. > > Thanks for working on this. > > Best regards, > > Kern > > PS: My bat demo is tomorrow, and I will be showing the very latest one in the > SVN ... :-) > >> Seriously, I'm fairly new to bacula and perhaps there's some useful >> trick to learn here. >> >> Riccardo >> >> Kern Sibbald wrote: >>> On Thursday 15 May 2008 07:43:50 rghetta wrote: >>>> Hello Kern, >>>> yes, I've changed the ordering from >>>> job.starttime desc, job.jobid desc >>>> to >>>> job.starttime=0 desc, job.starttime desc, job.jobid desc >>>> exactly to show records with starttime zero before the others. >>>> >>>> I've done that because I had to scroll to the bottom to see the >>>> scheduled-but-not-already-started jobs. >>>> I thought that change was an improvement, sorry :-) >>> Nothing to be sorry about. If we don't experiment, we might miss some >>> good ideas. The only question is whether or not the current behavior is >>> what we really want. >>> >>> What you did made sense, and I suppose I wouldn't have a problem seeing >>> scheduled jobs, but I also seem to see all the old jobs that were >>> scheduled but didn't actually run. >>> >>> Perhaps if you modify your sort to put the JobId first in the sort >>> sequence, it would possibly resolve the problem. >>> >>> Another suggestion might be to sort by EndTime possibly pushing any jobs >>> with a zero end time to the end of the list??? >>> >>> Kern ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Bacula-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/bacula-devel
