In the original formula template method, the templates were represented as: Templates.

The first line contains the source type and description, the second line contained the L style template, the third line contains the F style template, and the fourth line contained the S style template. To create a citation, the appropraite L, F, or S template was used, the variables (between [] and all upper case) were substituted, and a few html tags were used (for italics <i></i>, bold <b></b>, underline <u></u>), and <> surrounded optional parameters and punctuation, and {} surrounded privacy fields and punctuation. This resulted in the rendered citation. I wrote Unix shell scripts that would do this processing to rtf.

Variables are enclosed in square brackets and must be all upper case. They may contain numbers and punctuation. Requiring them to be upper case allows text such as [sic] to appear as simple text, but [SIC] is a variable that the code should prompt for a value.

Presented in this way, individual styles can be represented and shared, e.g. by export/import commands.

The citations were represented as: Citations.

The first line added a citation number or tag (e.g. JHY1) to the source type and description, the second through fourth lines contained the L, F, and S style templates to format the source citation, and the next lines contained a list of the variables, one per line, the variable name first and its value second. By storing citations this way in a SQLite database, one would always know precisely what templates were used to create the output because the templates were stored with the variables.

Presented in this way, citations in this style can be shared, e.g. by export/import commands. The information contained is sufficient to allow software to "mark up" the citation to printer ready output, such as rtf.

The scripts faithfully rendered all 170 QuickCheck Models (L, F, and S styles) to rtf.

Although I defined all 170 QuickCheck Models from Evidence Explained the shell scripts to use it with the SQLite database were command line oriented, not GUI, and not very useful to anyone but me. I want to rewrite it all with Java Swing, making it a GUI, but that would take me some time, if I decide to complete it.

Instead, I decided to parametrize the templates for easy use by programming vendors, and hope that they adopt a model that can accomodate this parametrization. They can even start with my computer ready parametrization to save them a lot of time. In most cases, they only need to generalize slightly how they handle and define sources, then anyone can insert this Evidence Style parametrization.