> On Sept. 10, 2013, 7:20 a.m., Cristian Oneț wrote: > > kmymoney/dialogs/investactivities.cpp, line 204 > > <http://git.reviewboard.kde.org/r/112364/diff/1/?file=185610#file185610line204> > > > > Also I don't see a reason why the timeouts passed to QTimer::signleShot > > vary that much, 5, 100, 150, is there a reason for it? > > Allan Anderson wrote: > I can't give an explanation for why the should differ. Having previously > encountered this same problem some why ago, Thomas advised using the timer. > As before, I started off using a small value, but found that this didn't work > in the new near-identical code area. So, I started to increase the delay. In > come cases, I had to use a large delay in order to see that the widget did > finally respond. I then started to reduce the value until failure, and then > back-off. Having got a particular widget finally responding, after further > work I found that the problem was still sometimes there and had to make > further adjustments. In many cases, this 'tuning' was revisited. > When finally all appeared OK, I then found that during actual window > resizing, some widgets would flicker quite badly during their repositioning, > depending upon the speed of mouse movement. The only way I could resolve > that was by zeroing the widget height instead of hiding it.
There must be a simpler way to do this. If this approach is so timing dependent it could be that the solution does not even work on a slower or faster machine/system. - Cristian ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://git.reviewboard.kde.org/r/112364/#review39691 ----------------------------------------------------------- On Aug. 29, 2013, 5:12 p.m., Allan Anderson wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://git.reviewboard.kde.org/r/112364/ > ----------------------------------------------------------- > > (Updated Aug. 29, 2013, 5:12 p.m.) > > > Review request for KMymoney. > > > Description > ------- > > If I choose to use a complex system for check numbers, such that the whole > number is not visible, > the only way I have available is to stretch the whole window. However, even > that doesn't help, > as the whole of the increase is grabbed by the Details column. I accept that > it is likely that that > column is going to need to be the widest. Then, why are the Payment and > Deposit columns twice the > width of the Balance column, when that column may be likely to have the > greatest value? Ditto for > the Date. > > This fix allows modification of column widths, but also resizes the > individual columns to more suitable widths. > > I found that Thomas had started to implement something similar some while > ago, so I have built upon and expanded that. > > I found that the edit widgets were particularly troublesome, in failing to > appear/disappear with the show() and hide() > methods, which I'd previously found when last in this area. Then, when the > screen was being resized, they flickered > more than acceptable. Eventually, where necessary, I resorted to > zeroing/resetting the height instead, which resolved > the issue, although with some complication. > > > This addresses bugs 312816 and 322768. > http://bugs.kde.org/show_bug.cgi?id=312816 > http://bugs.kde.org/show_bug.cgi?id=322768 > > > Diffs > ----- > > kmymoney/dialogs/investactivities.cpp 50f33ed > kmymoney/dialogs/investtransactioneditor.h 3e62c2a > kmymoney/dialogs/investtransactioneditor.cpp e9f87fb > kmymoney/dialogs/transactioneditor.h f07dafb > kmymoney/dialogs/transactioneditor.cpp 39049cf > kmymoney/views/kgloballedgerview.h 04a6303 > kmymoney/views/kgloballedgerview.cpp 78d98b2 > kmymoney/widgets/register.h eebe78d > kmymoney/widgets/register.cpp 1bdf5bd > kmymoney/widgets/transactionform.cpp 642e98f > > Diff: http://git.reviewboard.kde.org/r/112364/diff/ > > > Testing > ------- > > Extensive editing of sample files, and changing back and forth between > different activity types, which tended to show > problem areas. atype run. > > > Thanks, > > Allan Anderson > >
_______________________________________________ KMyMoney-devel mailing list KMyMoney-devel@kde.org https://mail.kde.org/mailman/listinfo/kmymoney-devel
