Re: Launch/Spawn to current application

Andreas Kraft (kraft@fokus.gmd.de)
Fri, 22 Aug 1997 12:00:16 +0200 (MET DST)

Hi Wataru,

thank you for you clarification. Your remarks help a lot (as always).

On Thu, 21 Aug 1997, Wataru KAMEYAMA wrote:

> For me, it's not so confusing because PoU of Spawn() clearly says that
> the Target shall not be an active Application, and there is a
> principle that we have only one active Application in MHEG-5.

You are right. What confused me was the following part of the PoU

"...while there is currently an active Application object."

For me, there is always an Application active, what else should be the
source of a Spawn action? But this is maybe implementation dependend?

> > "Check that the target application is different from the active
> > application; if it is not, ignore the action."
>
> Yes, you may. But you should notice that this causes a slight
> semantics change for Launch() and Spawn(). Without your proposal,
> if Launch() and Spawn() are targeted to an active Application, this is
> forbidden by the standard and the behavior of the engine is unknown.
> With your proposal, if targeted so, the behavior of the engine to this
> error condition is defined by the standard.

Oops, you are correct. Some discussions might be necessary on this topic.

> In essential, with your proposal or without your proposal, I agree
> with you that Launch() and Spawn() shall not be targeted to an active
> Application.

Thank you again for your clarification.

Very best regards,

Andreas

--
  o  _     Andreas Kraft
 (\_|_)      GMD FOKUS, kraft@fokus.gmd.de, +49 30 3463-7232
 T> ] [        The sky is the limit