Hi,

I am reading through CH2 of the book, and I want to make sure I understand the use of Iterators and Arbiters:

1) If you want to call yeild return to suspend a thread from operation until a message is received, you need to place this call in an iterator wrapped function and declare an Arbiter.Receive as non-persistent. Then a delegate (or handler) is called upon reception and the receiver is destroyed.

2) Otherwise, if you want to always allow a receiver to exist, you can call it from a regular (i.e non-iterator wrapped function) using the Activate call (and assume the receiver is declared persistent), and this will allocate a seperate thread to handle the received messages.

Please let me know if I am understanding this correctly...

Thanks,
Scott

Hi,

For 1, yes you can only yield inside an Iterator. In fact, if you try to compile code with a yield statement in a normal method it will cause an error. If you were to yield to a persistent receiver, it would never finish.

In general, whether a receiver is non-persistent or not depends on your application. Maybe I did not quite understand your question.

For 2, usually you declare persistent receivers in an Interleave. This allows you to have exclusive handlers. If you just set up a bunch of persistent receivers separately then they can all run at the same time and you might have issues with concurrent access to data.

Activating a receiver does not allocate a thread directly. Dispatchers are responsible for assigning threads to tasks. You can however create your own dispatchers and specify the number of threads.

Activate can be a little confusing. When you call Activate you are using the CcrServiceBase.Activate method which hides the dispatcher queue from you. You can do the same thing using Arbiter.Activate, but then you have to specify the dispatcher queue.

Trevor