Tracking whether a connection is alive

[Kay Sievers]

On Mon, 2004-07-12 at 15:12 +0200, David Zeuthen wrote:
> On Mon, Jul 12, 2004 at 01:49:28PM +0100, Mike Hearn wrote:
> > On Mon, 12 Jul 2004 12:29:38 +0200, Owen Fraser-Green wrote:
> > > Anyway though, even if there's some mechanism such that HAL can be 100%
> > > sure that an application has hung, will it always be able to release the
> > > lock without the application first being killed? In that case, again, I
> > > think dealing with hanging apps is out of HALs scope.
> > 
> > Yeah, I don't think we want to code too defensively here, deadlock
> > detection is notoriously tricky. For apps that it is a problem for, the
> > solution is "don't hang" :)
> > 
> 
> Right. I'll just implement the snippet Daniel Silverstone sent and we'll
> have to rely on existing mechanism to take care of broken apps. Which
> sounds kind of right to me.

What about some other kind of "ping", where the app holding the lock
_must_ return a string, which describes, what the device is actually
doing?

If another application tries to claim the device it will receive the
current holder of the lock and what's currently going on with the device
and can notify the user with something like:
 'CD' is used by 'burner-app' currently 'writing CD Data_2004-08-23'

Wouldn't it be nice and friendly?

Kay