Re: Gnome logout script

[Rodrigo Moya <rodrigo gnome-db org>]

On Mon, 2003-10-20 at 09:37, Mariano Suarez-Alvarez wrote:
> On Mon, 2003-10-20 at 04:28, Sean Atkinson wrote:
> > > > Sure, unless there's some reason I don't understand that some of its
> > > > processes choose to hang around.  However that's a bigger problem for
> > > > the longer term.  Also there are other reasons to want logout scripts
> > > > and tidy up the user's environment, so I think it's worth looking into.
> > > > 
> > > evolution-alarm-notify has to be kept around, for calendar alarms, and
> > > since it uses the wombat for the calendar, it also keeps
> > > evolution-wombat around. So, those 2 processes are saved in the session,
> > > which is what they are supposed to do.
> > > 
> > > There is a bug in Ximian's bugzilla about making it exit if there are no
> > > alarms. But if so, we need a way to restart the alarm daemon at some
> > > interval in the future to check for new alarms.
> > 
> > Looking into the processes left after a clean logout, I find
> > evolution-wombat remains without any alarm (I didn't set any). 
> > Presumably as a result of this bonobo-activation-server and gconfd-2
> > also remain.  Previously I remember seeing oafd too.
> > 
> > The file locks left by gconfd-2 are probably the biggest problem since
> > they prevent login from another machine sharing the same /home, but I'm
> > looking into that separately.
> 
> Actually, it prevents me from logging back in the same machine (at
> least, from logging back in with a working gconf setup)
> 
> I wonders what is the point to have alarms be executed when the user is
> not logged in: can't all pending alamns just be run the next time the
> session starts (I guess that in order to do that, wombat or whatever
> should find a way to get included in the session even if I don't
> explicitely save it, though)
> 
as I said, the alarm daemon is supposed to die with the session. So, if
it's being kept around, it's a bug.

I'll look at fixing this.

cheers