> At 08:51 3.9.2000 +0200, you wrote:
> >setState() will trigger message notifications. If this method doesn't
> >wait until all listeners have been notified they will not have time to
> >do graceful shutdowns, since doStopCode() will have executed long before
> >they get the notification....
> >
> >How to solve?
>
> Hmm.. well we have something that supports both synchronous and
> asynchronous messaging. SpyderMQ.
>
> Which leads to an interesting mind game. We've building a fully
> componentized server. In order for all the components to work together it
> has become obvious that messaging between them is the most crucial
> functional requirement. So should a component management
> framework like JMX
> be built on top of messaging services like JMS? And since the server is
> built on top of the JMX framework it leads to it having a very robust
> messaging system at its very core (nice for logging etc etc).
>
> Now instead we're plugging in JMS as an afterthough for EJB 2.0.
>
> Hmm.. where did I leave my pipe?
I got it, man that killer weed from California...
Ok, it is obvious that JMS is going to play a crucial role in the webOS, but
the point now is that we do need synchronous step
"stopping"
doStop()
"stopped"
one solution is to have the step in separate methods with asynch
callbacks... more complicated though, I need to think about it.
marc
>
> -- Juha
>
>
>
>
