The following appeared on comp.soft-sys.ptolemy, below is my response, I'm sure that others will have some ideas as well.
> From: [EMAIL PROTECTED] (llmercy) > Subject: Next Ptolemy II and DDE enhancements? > Date: 10 Apr 2002 10:43:32 -0700 > > It sounds like the next version of Ptolemy II is getting > close to completion. I have a few quick questions about > Ptolemy II and the DDE domain in the next release: Yep, we are coming down to the wire on the release > 1) Is the DDE still "experimental" or will it be "stable"? It is still experimental. Not much work has happened on dde. Recent changes in the devel tree (not the beta branch) broke dde and we are considering retiring it. dde needs a champion who is willing to develop it further. At this time, I'm not sure if there is anyone on site here at UC Berkeley who is interested in doing this. We would entertain collaborating with someone who wanted to further develop the dde domain and provide source code fixes and enhancements. > 2) At one time there was talk of using RMI to distribute > the actors in DDE. Is this implemented in the next > release and if not, is it still being considered? No, this is not in Ptolemy II 2.0 We are not really considering this at this time > 3) Is there support for doing checkpoints in the new > version? If so, can they be used to restart a > simulation from a checkpoint in the event of a crash? No, the next release does not support checkpoints. Ptolemy II never crashes :-) Seriously though, checkpoints would be nice, but we do not specific plans to implement checkpoints. BTW - I have a simple SDF model that is running in the foyer of the Donald O. Pederson Center where it displays a set of images using the full screen facility in JDK1.4. It has been running for weeks (months?) and has not crashed. > 4) Is there support for running Monte Carlos in the > next release? I think so. The FSM domain as been extended to perform multiple runs of refinement on a model > I would appreciate any help with these questions. We're > currently evaluating Ptolemy II for our work and one of > our requirements is parallel distributed simulations. Any > help in this endeavor would be very helpful. Yep, parallel distributed simulations are pretty tricky. In the past, people have tried to partition simulations and run actors and groups of actors over different processors. The big problem is that most networks, say 100Mb ethernet, have high latency, so this causes problems. Running standalone simulations in parallel seems to work a little better. > Thank you. > > llmercy -Christopher ---------------------------------------------------------------------------- Posted to the ptolemy-hackers mailing list. Please send administrative mail for this list to: [EMAIL PROTECTED]
