Hi, thanks for your reply. > (...) > is called via a widget action and during Exit() this widget itself is > destroyed, on return from Exit() this widget is then still used which > might also cause a BadWindow. O.K. It is clear that we don't know where the problem comes from ... So, can I protect myself somehow ? For example, what if I in the beginning of xFrame destructor place : gVirtualX->UnmapWindow(GetId()); and then delete everything that I expect to be deleted. Would it "cure" the problem ? Will the "UnmapWindow" be executed immidiately ? Or should I place : gVirtualX->UnmapWindow(GetId()); gSystem->ProcessEvents(); in the beginning of the destructor ? > historyFrame = 0; // no cast needed Of course one does not need the "cast" for the compiler. I always try to be "explicit" : historyFrame = ((G4RootTextFrame *)0); So, looking at this line I immidiately know that the pointer that was destroyed was a "G4RootTextFrame". Otherwise I would have to dig through all sources in order to find what a hell the "historyFrame" was ... Thanks in advance, Jacek.
This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:50:48 MET