Hi Christopher, Thanks for your reply. The Generic Director had been convenient for running models with arithmetic loops, which other directors would throw exceptions on. We will just use alternative implementations that don't require loops and use a different director.
Thanks, Kevin On Wed, 27 Apr 2005, Christopher Brooks wrote: | Date: Wed, 27 Apr 2005 19:12:27 -0700 | From: Christopher Brooks <[EMAIL PROTECTED]> | To: Kevin Tang <[EMAIL PROTECTED]> | Cc: ptolemy-hackers <ptolemy-hackers@eecs.berkeley.edu> | Subject: Re: inconsistent MoML execution | | Hi Kevin, | | The short answer is that the Generic Director is a hack. As you've | found, it depends on the order of the entities in MoML file. | | To be honest, I think it was a fluke that using a Generic Director | works at all. As I remember it, Professor Lee and I think Jorn | Janneck (probably with others), discovered that the Generic Director | could be made to work without much effort. The Generic Director | does have its role for small models though. | | For more complex models, you probably want to use SDF. | Ptolemy II 5.0-beta, due out in May, will include DDF and HDF: | | "The Dynamic Dataflow (DDF) domain, created by Gang Zhou, is a | superset of the Synchronous Dataflow (SDF) and Boolean dataflow (BDF) | domains. In the SDF domain, an actor consumes and produces a fixed | number of tokens per firing. This static information makes possible | compile-time scheduling. In the DDF domain, an actor could change the | production and consumption rates after each firing. The scheduler | makes no attempt to construct a compile-time schedule, neither does it | attempt to statically answer questions about deadlock and boundedness, | which are fundamentally undecidable. Instead, each actor has a set of | sequential firing rules (patterns) and can be fired if one of them is | satisfied, i.e., one particular firing pattern forms a prefix of | sequences of unconsumed tokens at input ports. The scheduler | dynamically schedules the firing of actors according to some criteria. | The canonical actors in the DDF domain include Select and Switch, | which consume or produce tokens on different channels based on the | token received from the control port." | | "The Heterochronous Dataflow (HDF) Domain, created by Ye Zhou is an | optimization of the SDFDirector and implements the HDF model of | computation. The HDF model of computation is a generalization of | Synchronous Dataflow (SDF). In SDF, the set of port rates of an actor | (called rate signatures) are constant. In HDF, however, rate | signatures are allowed to change between iterations of the HDF | schedule. Within each state, the HDF model behaves like an SDF model." | | It would be nice if someone would create a Boolean Dataflow (BDF) | domain. | | _Christopher | | | | | | -------- | | Hi Hackers, | | I've encountered an inconsistency in the way MoML written by Vergil gets | executed by the generic Director. | | At the end of this message are two simple Ptolemy layouts that are | functionally and structurally identical in Vergil. However, running the | first layout produces a "full mailbox" exception. Is this a bug? | | Reordering entities in the xml file will solve the problem. However, in | more complex layouts more connected entities, the proper order is not | clear. What is the correct ordering of entities in MoML to get the | generic Director to run layout without full mailbox execeptions? | | | I can send the xml files as email attachments as requested. I also have | examples of slightly more complex MoML reproducing the bad behavior and | the fix, and can send those as well. | | | Thank you very much, | | Kevin | | | ---------------------------------------------------------------------------- | Posted to the ptolemy-hackers mailing list. Please send administrative | mail for this list to: [EMAIL PROTECTED] | ---------------------------------------------------------------------------- Posted to the ptolemy-hackers mailing list. Please send administrative mail for this list to: [EMAIL PROTECTED]
