Adriana Ricchiuti <[EMAIL PROTECTED]> writes: > I' m working on workspace. I viewed the PTII source code, and I note > that a new workspace is created when a director is added to a > composite actor. Is it right? Is there a Workspace for every > director? Where can I find more information about workspace > behaviour (without investigate in the code ;-) )?
Hi Adriana, Other than the source, you could look at the Ptolemy II Design doc at http://ptolemy.eecs.berkeley.edu/papers/05/ptIIdesign1-intro There are three volumes, you might find chapter 1 of volume 2, "The Kernel," to be the most relevant: > 1.3.3 Workspace > > Workspace is a concrete class that implements the Nameable interface, > as shown in figure 1.4. All objects in a Ptolemy II model are > associated with a workspace, and almost all operations that involve > multiple objects are only supported for objects in the same > workspace. This constraint is exploited to ensure thread safety, as > explained in section 1.6 below. Ptolemy uses hierarchy. Every object that is in a model (director, actor, port in an actor, parameter etc) has a Workspace. I think if a Workspace is not specified at construction time, then one is created. Usually Directors go in a container (a top level) that has a Workspace. > I'm working on actor's ports and I discovered that it is possible > setting a port to be neither output nor input. What is the meaning of > this port? As Edward said, we have not yet found a reason to use this, but that is not a reason not to have it available. We also don't have much of a need for level crossing links, yet we have limited support for those. There has not been much work done with actors that are both input and output. I suppose input/output ports could be used to simulate a bus, where we want to check for contention? In other words, ports that are neither input nor output might have a use in a model of computation that we have not yet developed. > I'm sorting actors in PTII. I'd like to know if there is a > functional documentation about actors. I mean a documentation in > which, for every actor, there is a complete and user friendly > description about behaviour, data token allowable, etc. There is a little bit of documentation in Chapter 4, "Actor Libraries" of Volume 1 of the Ptolemy II Design Doc. The source code includes information about port types. The source code is processed by javdoc to produce html. See http://ptolemy.eecs.berkeley.edu/ptolemyII/ptII5.0/ptII5.0.1/doc/codeDoc/index.html You might want to start with what is in ptolemy.actor.lib. The Javadoc for ptolemy.kernel.util.Workspace might also be of interest. Professor Lee has developed a documentation viewer that displays html documentation for actors and composite actors. The cool thing is that the documentation for each object is customizable and follows the inheritance tree. This code is in the Ptolemy II CVS tree now, if you right click on an actor and select "Documentation -> then the documentation is displayed. I've been working with him on this code, we now have a Doclet that generates xml code for use with the displayer. I've also hacked up a index tool that lists what demonstrations an actor is used in. The Kepler folks are also very interested in user level documentation and have a system in place that I have not tried recently. We hope to hash out a common system next week. I was thinking it would be cool to have a Wiki or something the the Php website user annotations available for actors where users could annotate actor documentation and discuss gotchas and hints. _Christopher Christopher Brooks (cxh at eecs berkeley edu) University of California Programmer/Analyst Chess/Ptolemy/Trust US Mail: 558 Cory Hall #1770 ph: 510.643.9841 fax:510.642.2739 Berkeley, CA 94720-1770 home: (F-Tu) 707.665.0131 (W-F) 510.655.5480 (office: 400A Cory) ---------------------------------------------------------------------------- Posted to the ptolemy-hackers mailing list. Please send administrative mail for this list to: [EMAIL PROTECTED]
