[concurrency-interest] Coordinated Runnables

Alexei Kaigorodov alexei.kaigorodov at gmail.com
Thu Aug 29 13:03:38 EDT 2013


Okay, I was wrong and Akka is closer to the Hewit's original concept. On
the other hand, I do not propose to implement pure Petri nets - indeed,
message passing should be asynchronous. The closest concepts are dataflow
programming and http://en.wikipedia.org/wiki/Dataflow_architecture .


On Thu, Aug 29, 2013 at 11:40 PM, Roland Kuhn <rk at rkuhn.info> wrote:

> Hi Alexei,
>
> Akka implements the Actor Model as described by Carl Hewitt in 1973 quite
> closely (the deviations are not relevant for this discussion); what you
> describe is closer to CSP or the (synchronous) π calculus. Asynchronous
> message passing results in a better decoupling of sender and receiver of a
> communication, which is the important distinction here.
>
> Petri nets can be used to model anything you want, including Actors or
> CSP, but directly implementing them will probably not achieve good
> scalability due to their synchronous (and therefore synchronized)
> transition semantics.
>
> Regards,
>
> *Dr. Roland Kuhn*
> *Akka Tech Lead*
> Typesafe – Reactive apps on the JVM
> twitter: @rolandkuhn
>
> 29 aug 2013 kl. 18:24 skrev Alexei Kaigorodov:
>
>
> Place is a term in Petri Net model - place where tokens wait to be
> processed. "Of type queue" means that tokens are stored in a queue. Another
> possible type, for "colorless" tokens, is just a counter for events, like
> semaphore. Single means actors in Scala and Akka can have only one input
> place (queue) for messages, while Petri Net Nodes (transitions) can have
> several, and this is very  natural for asynchronous programming, as
> illustrated at the picture in the referenced googledoc paper.
>
> On Thu, Aug 29, 2013 at 11:04 PM, √iktor Ҡlang <viktor.klang at gmail.com>wrote:
>
>> Single place of type queue <-- wdym?
>>
>>
>> On Thu, Aug 29, 2013 at 7:42 AM, Alexei Kaigorodov <
>> alexei.kaigorodov at gmail.com> wrote:
>>
>>> Yes it is an actor, in its initial meaning. But unfortunately, the term
>>> "actor" today mainly means restricted actor of Scala or Akka style (with
>>> single place of type queue), and I am in doubt if "actor" would not cause
>>> misunderstanding.
>>>
>>>
>>> On Thu, Aug 29, 2013 at 9:26 PM, Zhong Yu <zhong.j.yu at gmail.com> wrote:
>>>
>>>> Hi Alexei, if in your model one place can be connected to only one
>>>> transition, can't we model one transition with its input places as one
>>>> actor?
>>>>
>>>> Zhong Yu  ://cs.oswego.edu/mailman/listinfo/concurrency-interest<http://cs.oswego.edu/mailman/listinfo/concurrency-interest>
>>>>
>>>
>>>
>>
>>
>> --
>> *Viktor Klang*
>> *Director of Engineering*
>> Typesafe <http://www.typesafe.com/>
>>
>> Twitter: @viktorklang
>>
>
> _______________________________________________
> Concurrency-interest mailing list
> Concurrency-interest at cs.oswego.edu
> http://cs.oswego.edu/mailman/listinfo/concurrency-interest
>
>
> --
> Simplicity and elegance are unpopular because they require hard work and
> discipline to achieve and education to be appreciated.
>   -- Dijkstra
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs.oswego.edu/pipermail/concurrency-interest/attachments/20130830/11e5bbb3/attachment.html>


More information about the Concurrency-interest mailing list