A Semantic IDE for Constructing Stratified Outlines

Outlines are bi-dimensional texts that can be situated on a coordinate plane such that each heading node has a column and row position. Legal-style heading numbering partially captures this structure. For example, a node numbered  indicates that it is in the fourth column of the outline, since each period of the heading number marks an absolute column division. This is the depth of that particular node. The node number gives as well some vague information about the row position: the number of a given column is a count of the nodes preceding it in that column. Thus indicates that there are two other headings in the first column (1. and 2.), one in the second (3.1.), etc. This is the (relative) height of the node.

We will define a stratified outline as having two properties:

·        It is column regular: all branches have the same depth. For example, if an stratified outline has a node labelled, then there is likewise at least and In other words, all paths from the top node lead to the maximal bottom node.

·        The content of each node is such that higher nodes are summaries of the information contained in the lower nodes: higher nodes condense the information in the lower nodes; lower nodes expand on the information in the higher nodes.

Stratified outlines are read simultaneously in two directions. One reading is restricted to a single column; each heading of that column becomes a part of a coherent text made up of the series of headings at that column. At the same time, each column is an abstract of the lower columns: lower nodes expand on the information of higher nodes; higher nodes summarize.

In this talk we will consider both the stratified browser (described at  www.fhis.ubc.ca/winder/me/mutanda) that displays in a flexible fashion a stratified outline and the computational and methodological tools that can be used to construct stratified outlines. Some of the principles of a future semantic integrated development environment will be… outlined.