Re: [patch] kernel events layer

[Oliver Neukum <oliver@neukum.org>]

Am Dienstag, 27. Juli 2004 20:15 schrieb Mike Waychison:
> Tim Hockin wrote:
> > On Mon, Jul 26, 2004 at 03:03:05PM -0400, Greg KH wrote:
> >
> >>>On a related note, is this supposed to supersede the current hotplug
> >>>mechanism?
> >>
> >>No, it will not.  At the most, it will report the same information to
> >>make it easier for userspace programs who want to get the other
> >>event information, also get the hotplug stuff through the same
> >>interface, reducing their complexity.
> >>
> >>So the existing hotplug interface is not going away at all.  Do not even
> >>begin to think that :)
> >
> >
> > What about flipping it around and using either hotplug or a hotplug-like
> > mechanism for these events?
> >
> > It solves the issue of events being dropped when there is no listening
> > daemon...
> >
> > These are not going to be high-traffic messages, right, so the overhead is
> > negligible...
> >
> 
> The problem with this is that you'd lose the ability to send the
> messages broadcast, whereby you may have multiple dbus's listening for
> events.

Why is that? Couldn't a hotplug script use dbus?

	Regards
		Oliver