Frequently Asked Questions

General: * What is Design/CPN?
* What is Coloured Petri Nets?
* Why is it called Coloured Petri Nets?
Editor: * How do I move the endpoints on arcs and connectors?
Simulator: *
Occurrence Graph Tool: *
ML-engine Daemon: * What is the cpnmld daemon good for?
Diagrams: * What are the Three Diagram Files?
* Can diagrams be moved between platforms?
Implementation: * Is the source code available?
* What is the design of Design/CPN?
Documentation: * Are there any errata on the documentation?
* Where are the examples for the tutorial located?

General

Why is it called Coloured Petri Nets?

The original idea behind the first versions of high-level Petri nets is that the transitions can distinguish between different kinds of tokens, e.g., three different kinds describing three different processes. In this way we can have one net, instead of having three identical subnets.

Since the tokens of ordinary Petri nets often are drawn as black dots, high-level net tokens were referred to as being coloured - in the sense that they could be distinguished from each other.

Gradually it was realised that instead of having a few colours, totally general types could be used. With the modern form of high-level nets it is much more appropriate to talk about Petri net with types - and we usually have a very large number of different "colours values" - often an infinite number (in terms of physics you may think about the continuum of all existing colours).

Refer to Kurt Jensen's book Coloured Petri Nets, Basic Concepts, volume 1, for more information.

Editor

How do I move the endpoints on arcs and connectors?

Do the following in order to move the source endpoint:

Choose the relevant arc/connector.
Choose Drag from the Makeup menu.
Hold down the left mouse button while moving the mouse in order to drag the endpoint to a new location.

You move the destination endpoint by additionally holding down the <ctrl> key (in item 2 above). Short-cuts can also be used, i.e., use respectively <alt>+2 and <alt>+<ctrl>+2.

A slightly faster alternative is to do the following:

Choose the relevant arc/connector.
Hold down the <shift> key and left mouse button with the mouse pointer on the endpoint to move.
Hold down the key+button while moving the endpoint by means of the mouse.

The latter has the disadvantage that the pointer obscures the endpoint so it is hard to see where the endpoint is located.

When moving endpoints the first time for each arc/connector a region is created. The arc/connector is attached to the region. The father of the region is the object to where the arc/connector originally was attached. Thus moving endpoints really means moving an invisible (endpoint) region.

Simulator

Currently empty.

Occurrence Graph Tool

Currently empty.

ML-engine Daemon

What is the cpnmld daemon good for?

The cpnmld daemon is used to set up the communication between Design/CPN and the simulation engine. The ML Configuration Options dialog (Set menu) is used for setting the specific parameters that Design/CPN needs in order to communicate with cpnmld (i.e., host name, port number, and ML image). When you, e.g., start a syntax check, Design/CPN connects to cpnmld on the specified host and port number. After a login procedure, Design/CPN asks cpnmld to start the ML engine specified by ML image and bind the communication up between Design/CPN and the ML engine.

The cpnmld daemon also writes a lot of information to a log file. This file is called /usr/tmp/cpnmld.log.#, where '#' is the port number.

Diagrams

What are the Three Diagram Files?

When Design/CPN saves a diagram it uses three different files:

"ModelName" contains all the graphical information
"ModelName.DB" contains a data base with all the CPN information (i.e., a mathematical representation of the CP-net)
"ModelName.ML" contains the Standard ML code generated by Design/CPN during the syntax check and the switch to the simulator/occurrence graph tool (this file only exists if the ML compiler has been in use).

The first two files must always be kept. If one of these are lost your Design/CPN diagram is lost. The ML file (which is much larger than the other two files) can be automatically recreated (by making a new syntax check and switch). Hence the ML file can be discarded, but it will then of course take some time to recreate it.

The three files must be kept together in the same folder/directory. If one of them is renamed the others must be renamed accordingly. This is not true for the ML file on the Unix platform. There the name is specified in the Save Dialogue and hence it cannot be changed unless a new ML file is saved.

Can diagrams be moved between platforms?

Diagram files can only be moved from the Macintosh to any UNIX platform. When moving files remember to use ftp in binary mode. It is not possible to use UNIX diagrams on the Macintosh. Some diagrams saved on one UNIX system (eg., Solaris) can't be loaded on another (eg., Linux). In almost all cases this problem can be avoided by using the Textual Interchange Format.

Implementation

Is the source code available?

The Design/CPN source code is not freely available.

What is the design of Design/CPN?

Refer to Kurt Jensen's book Coloured Petri Nets, Basic Concepts, volume 1. See chapter 6 which also contains bibliographical remarks.

Documentation

Are there any errata on the documentation?

We are of course very interested in hearing about your problems with the tutorial, manuals, and other documentation on Design/CPN. There is a WWW page about Known Errors in Design/CPN Manuals where you can both browse existing errata, and also report your own findings.

Where are the examples for the tutorial located?

The examples for the Design/CPN tutorial are not together with the release as claimed in the tutorial. However, the tutorial examples can be found in the same section as the tutorial on the WWW page about Design/CPN Online Manuals (see "Diagram and Text Files Used in the Tutorial").

Last modified: Fri Apr 16 09:47:22 MET DST 1999 -- Design/CPN Online admin