wxProcess::Kill(..., wxSIGTERM) is not the same as taskkill /F ..., probably it is the same as calling taskkill without /F.
See here: http://docs.wxwidgets.org/trunk/group__group__funcmacro__procctrl.html#gae3a7de5a29e70c77463dd1bc38fb98cf
Unfortunately the docu just talks about Kill, not taskkill. Can only say from what I obverved from the behaviour that both taskkill and taskkill /F closed the application ok, the CB Abort button, aka the wxProcess::Kill did not. And from you feedback I take it that you would be surprised by different behaviour as well.
I'm a little confused by your last post. Can you explain in details what are you trying to do?
I agree it can be a bit confusing by description in pieces only. Maybe that helps
I have this crash effect on application exit and have no idea what happens. I was just trying to find a way to reproduce it in the debugger. But since only the run abort from CB caused that crash, first I was looking for another way to cause that crash. My first idea was to find out what the abort button does and just do the same from the console.
With your help I can see now that CB does what you would expect, send a sigterm to the window. But that is exactly what I think the task manager kill and taskkill also do. But I still get that crash only with the Abort button version, not when stopping the App any other way.
The last post was about an alternative approach I did not see before. Its basically the way I can debug what the Build/Abort button does live. If you want I run the C:B debugging in parallel to CB running the application via the Build/Run command.
So now I can debug the problem. I did not find the bug yet, but I can see that I did not anticipate the way an application is closed by calling wxApp:OnExit() directly, i.e. before any windows know about the app being closed. But thats a pure wxWidgets programming issue and not C:B related.
To make a long story short. We can close that thread, because the issue itself is answered. I know what C::B does.
I still do not know why my App behaves differently between C::B Abort and taskkill but thats out of scope with this thread.
Should I find out sth which relates to C:B I shall add a short Info to this post.
Thanks for your help.