[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: little KProcess problem
From:       Lubos Lunak <l.lunak () sh ! cvut ! cz>
Date:       1999-06-02 6:49:27
[Download RAW message or body]

On Út, 01 čen 1999, Harri Porten wrote :
>Lubos Lunak wrote:
>> 
>> I thought KApplication::shutDown() could solve it, but it seems to be emited
>> only when the app is forced to exit ( is the shutDown() signal really intended
>> to work this way ? ).
>
>Yes. shutDown() is only emitted when the session manager signals the app
>to clean up when the user logs out.
>
>Harri.

 Hmm, now I see how it works. If the app "decides" to quit, it quits using
KApplication::quit() the event loop and exits normally and especially cleanly.
But if it's killed using xkill or by the logout process, it only emits the
shutDown() signal and exit()-s without calling destructors for instances on the
stack, so the slots connected to shutDown() should theoretically duplicate all
the cleanup. A bit ugly, that's why my test app behaves so strangely, and some
other apps leave temporary files. It should be at least mentioned in the docs.
I wonder why it doesn't simply ignore the following X errors and call quit().
 Anyway, a signal emited before destroying KApplication would be IMHO handy.

 Lubos Lunak
 l.lunak@email.cz

[prev in list] [next in list] [prev in thread] [next in thread]