Yes, we've been through this before.

In his 1991 classic Writing Spaces Jay David Bolter noted that:

An electronic text never needs to end. (p. 87)

I've quoted this before, both in various presentations and in these pages as well. The idea is rather clear, though 25 years ago it had a different ring that it has today. After all, even if a text links to pages that link to pages that link to ... even that never-ending text ends somewhere along the line - probably when the reader simply feels that he or she has tangented enough already, or when the ends get so loose or frayed they're no longer meaningful. When we move from theory to practice, however, and someone (let's say me) writes a hypertextual document, it's clear that there are going to be some boundaries. Yes, Boidem columns are more or less self-inclusive, meaning that someone can read a single column, (hopefully including the basic linked pages) and know that they've read something from beginning to end. It's also true, however, that each column's main page links back to the entire project's main page, meaning that from there one can get to all of the columns. I wouldn't go so far as to say that that's never-ending, but it certainly substantially expands the limits of the text.

Writing a column, however, is different from reading one, and the main difference may be that no matter how many tangential links I create, I still have a clear idea of where I think the column is going to end, or at least where it should.

Go to: Get to the point already!, or maybe Isn't that a bit oxymoronic?, or
Go to: How to write a Boidem column.