Questions regarding the wakeup interrupt

[Michael 'Mickey' Lauer]

On Saturday 29 March 2008 23:57:14 Holger Freyther wrote:
> On Friday 28 March 2008 13:01:20 Michael 'Mickey' Lauer wrote:
> > On Friday 28 March 2008 03:48:34 Holger Freyther wrote:
> >
> > Regarding responsibilities, the suspend cycle should be somewhat like:
> >
> > pimd: "please set wakeup time for next appointment"
>
> or atd. But we agree that there needs to be someone to control the RTC
> resource. It depends how broad PIM is, alarm clock could be PIM, scheduling
> of recordings could be done with a calendar/event API as well. On second
> thought I agree, the name PIM confused me. :)
>
> Start this application and use this dbus path to tell me (the application
> scheduling an alarm is not a dbus service though *eeek*). So starting of
> applications and waiting for them to turn up needs to be handled as well.
>
> > eventd: "Ok, this is far away and there's no one playing music or
> > watching a film right now, I guess we can shutdown now"
>
> Eek. This sounds like GNOME. So let me tell you one of the most annoying
> differences between OSX and GNOME.
>
> I listen to music and 'press' the suspend button/close the lid:
> 	OSX: Does sleep
> 	GNOME: Says rhythmbox refuses the suspend because music is played.

I agree, however your example does not apply in this case. We were not 
referring to explicit suspends, were we? If so I missed that, I was talking 
about a suspend triggered by a timeout for non-activity.

:M: