Structural or semantic markup identifies the role or meaning of a portion of text. Presentational markup specifies how it looks. The key advantage of the former is that the formatting can be changed to fit the circumstances, platform, and user preferences. However, real-world text doesn't lend itself to completely structural markup. Two different portions of text that have the same essential role may be presented differently in the original.
If forced to choose between structural and presentational markup, I would certainly choose the former -- and I think it makes the best baseline. However, I think it's worth adding sufficient presentation details to the "master" document so that a very close approximation to the original format can be achieved.
Here's a simple example, a portion of the table of contents from The Wonderful Wizard of Oz:
--Introduction-- 1. The Cyclone 2. The Council with the Munchkins
When the above is omitted from the master file and the TOC is generated, the result looks something like this:
Introduction 1. The Cyclone 2. The Council with the Munchkins
i.e. the m-dashes and leading whitespace are omitted.
Granted, this is no great loss, but it illustrates the type of format information that some would like to keep.
Would the following work in TEI? I didn't get a chance to test it.
I'm sure there are more examples. Time permitting, I'll add them here as I encounter them.
Posted Oct. 28, 2004
Classicosm is a Product Architect site.