On 2010-06-11 at 10:57:28 [+0200], Axel Dörfler <axeld@xxxxxxxxxxxxxxxx> wrote: > Stephan Assmus <superstippi@xxxxxx> wrote: > > On 2010-06-11 at 09:56:51 [+0200], axeld@xxxxxxxxxxxxxxxx wrote: > > > Log: > > > * Before stippi starts clobbering the sources everywhere with > > > superfluous > > > work-arounds, I applied the work-around where the problem > > > actually occurs, > > > until someone takes the time to look into it. > > Thanks! Will remove it in LaunchBox. IIRC the only other place is in > > WebPositive. > > I think so, too :-) > While that maybe lowers the motivation to look into the bug, I think > work arounds should be installed as deep as possible, so that as few > components as possible are affected. > > One might also want to add a test case that shows the misbehaviour to > the tool tip test application :-) Like I said in the ticket, I've already looked into the solution back then, and I think the problem is that fVisibleToolTip is never reset and HideToolTip() is never called either. But fVisibleToolTip is returned first, which makes another part of the code think the tool tip didn't change, IIRC. This may already be the problem, but I didn't understand from your replies in the ticket if you followed my reasoning or not. Best regards, -Stephan