> From: David G Anderson [mailto:[log in to unmask]]
> Sent: Tuesday, June 27, 2000 9:39 AM
> I much appreciated Kevin William's edifying comments on
> Elements vs Attributes and would like to know his
> comments on the Text Encoding Initiative (TEI) at Oxford.
I'm familiar with the TEI - I think it's very well done. All of the text of
the original source is preserved as text content:
<q>Frankly, my dear, I don't give a damn.</q>
while annotative information (information not included in the original text)
is included as attributes:
<div id="WN1" n="I" type="book">
When operating from an original text source, this is a good strategy to
adopt - a "markup-ignorant" viewer will still be able to make sense of the
document content. However, when an original text source is not available,
other factors need to be taken into consideration:
- Is there a meaningful text representation of the data, and is the primary
purpose of the designed structures to support that text representation?
- Does the information have an inherent order?
- Are there other constraining factors on design (such as designing for
- Can markup and text appear adjacent to one another (the "mixed" content
But I digress. As an example of a text markup strategy, I think the TEI is