General Issues

The following general points may be of interest to all. Please note that not all points are equally relevant to every submission, and you should not feel obliged to incorporate any of the features/utilities that I have mentioned.

• Be sure that you are primarily attending to the meta-level issues involved in the application of EM, rather than writing about a specific domain of application. Because you are writing for the Warwick Electronic Bulletin on Empirical Modelling, the principal focus of your paper should be on some aspect of EM. (Although it's a good thing to write with insight about an application field, it would be unfair to give disproportionate credit for such insight except insofar as it relates to the application of EM! Bear in mind also that writing an application in tkeden that might just as well have been written in Java does not qualify as an application of EM.)

• It is important to be sure that the reviewers of your paper can access and use your models as necessary when evaluating your work. You may find it useful to prepare a tkeden presentation, similar to those that have been presented in lectures, that incorporates relevant illustration of your model in use. In some circumstances, the organising committee may call upon you to demonstrate your model in action, if it is otherwise difficult to evaluate your contribution properly.

• Don't overlook the usefulness of figures and screenshots in communicating information and ideas.

• In many contexts, it may be helpful to develop (simple) special-purpose notations using the Agent Oriented Parser.

• The use of the Dependency Modelling Tool can be helpful in clarifying relationships, both in documenting the model, and in explaining issues that relate to model-building.

• If you are using eddi in tkeden 1.46 (the latest Windows version), don't forget to import the version of the eddi garbage collector that is used in the later versions of tkeden that are now the default on linux. You may otherwise experience problems when using eden in conjunction with eddi in certain circumstances. The eden function you need to redefine is garbage_collect, for which the relevant code is:

  proc garbage_collect {
          auto i, j, dCAT;
          for (i=2; i<=CATALOGUE#; i++) {
                  dCAT = dependants(CATALOGUE[i][1]);
                  /* writeln("i = ", i); writeln(dCAT); writeln((dCAT[2])#); */
                  for (j=1; j<=dCAT[2]#; j++) {
                          /* writeln(dependants(dCAT[2][j])); */
                          if ((dependants(dCAT[2][j])[1])==[]) {
                                  /* writeln(dCAT[2][j]); */
                                  /* this line changed on WMB's request 14 January 2004 [Ash] */
                                  if (substr(dCAT[2][j], 1, 4) == "var_") self_asgn(dCAT[2][j]);
                          }
                  }
          }
  }