You can construct Venapps in any manner that you choose, and the process does tend to be iterative. The most important determinant of what is required in a Venapp are the needs of the user. Be sensitive to their needs and, if appropriate, include them in the construction process.
When building a Venapp it is often desirable to start with the logical flow. Build the screens containing the menus and buttons you want so that you can test the usability of the Venapp. This is quick and can be done before a model is ready.
Input screens can easily become overwhelming, and our experience is that it is best to break complicated input screens into a number of simpler components. Including more than ten distinct input concepts on a screen can be confusing. The best way to break things up depends on purpose, but there are usually some fairly natural hierarchies available.
Output of the model can be limited to a small number of reports and graphs. However, there is a vast leap between prepackaged analysis (showing a small number of predetermined reports and graphs) and active analysis controlled by the user. For pedagogical reasons, we encourage you to give users access to the interactive analysis capabilities possible in a Venapp — this should help to give the underlying model validity in the eyes of the user. See the section on Analysis Screens below.
Venapp are defined using a scripted input file that can be created using a text editor or using Vensim's Venapp Editor which behaves much like it does when working with models. Regardless of the tools you use the actual Venapp code is stored as a text file.
You might wish to base your Venapp on the templates provided and customize it to your needs. You can also write all the code in a Venapp from scratch. However, usually you will find it quicker to copy code, either full screens or just selected lines, from existing Venapps and modify this code for your particular needs.
