INTRODUCTION

The principal goal of scientific publications is to teach new concepts, show the resulting implications of those concepts in an illustration, and provide enough detail that the work is reproducible. In real life reproducibility is haphazard and variable. We rarely see a seismology PhD thesis being redone at a later date by another person.

The reproducibility problem can be largely overcome by standardized software generally available that is not hard to use. A new form of documentation is coming into existence. We call it an active document (a-doc). Active documents will serve us far better than paper documents. In an a-doc the author provides programs and a command script for every figure. Readers, students, and customers, can thus verify the calculation and adapt it to new circumstances without laboriously recreating the author's environment.

An a-doc could take many forms. It need be no more elaborate than this:

1. Each figure in the document has a name given in the figure caption.

2. Each figure in the document is made by a computer command script of the same name.

3. The document be distributed along with the command scripts and underlying programs, written in languages and using plotting code that is reasonably accessible.

Interactive documentation

Once an a-doc has been prepared another frontier immediately opens up, interactive documentation (i-doc). Interactive documents allow many features, the most obvious of which are (1) pushbuttons in the text allow rapid jumping from page to page, (2) figures can be converted into movies. While these features are dazzling, they are a distraction, I believe, from the truly revolutionary feature of an a-doc which is the ready reproducibility of the results it describes.

From the system programmers' view, the transition from an a-doc to an i-doc is a huge one (described elsewhere in this report), but from the authors' view the transition from an a-doc to an i-doc is easy.