[kepler-users] Zero-delay loop in DE domain

[Edward A. Lee]

Which director are you using? It sounds like it might be
DE, in which case, you need either a TimedDelay or a Register
actor in the feedback loop.  In PN or SDF, a SampleDelay actor
will do the job...

The technique of using a parameter is probably not desirable
in any case, as it's likely to be nondeterministic.

Edward


At 01:59 PM 1/18/2006 -0600, Maurice Yarrow wrote:
>Content-class: urn:content-classes:message
>Content-Type: multipart/alternative;
>         boundary="----_=_NextPart_001_01C61C69.A7920CD7"
>
>Hello Kepler Users
>
>I have what I call a negotiated loop involving three actors.  The is 
>negotiated in the sense that as a result of a test it can only do one 
>go-around:
>
>ActorA   sends event to   ActorB  which requests info 
>from   ActorC   which responds with the requested info 
>to   ActorB,   which based on this info, sends a request to   ActorA   for 
>an action.
>
>Well, no question about it, there is a loop here, but what do I need to do 
>to avoid this "Zero-delay loop"?
>
>In other such situations, I set a global flag ("parameter") in one actor 
>and from another unconnected graph in that actor, polled and read that 
>flag and then acted on the contained information.   But in this case, 
>immediate servicing of the original event is required.  Polling in not an 
>option here.
>
>Any help most welcome.
>
>Maurice Yarrow
>Nortel Networks Research Labs
>_______________________________________________
>Kepler-users mailing list
>Kepler-users at ecoinformatics.org
>http://mercury.nceas.ucsb.edu/ecoinformatics/mailman/listinfo/kepler-users

------------
Edward A. Lee
Professor, Chair of the EE Division, Associate Chair of EECS
231 Cory Hall, UC Berkeley, Berkeley, CA 94720
phone: 510-642-0253 or 510-642-0455, fax: 510-642-2845
eal at eecs.Berkeley.EDU, http://ptolemy.eecs.berkeley.edu/~eal