On Fri, 13 Feb 2004 09:04:36 -0500, David Sewell wrote:
> So the tradeoff is going to be:
> Numbered divs: makes it easier to deal with specific levels in
> the text hierarchy, like all second-level elements
> Unnumbered divs: makes it easier to deal with similar elements that
> may exist at different levels in different documents.
One advantage of relying on the type attributes, at least in part, to
identifying elements of interest is that it makes the queries more
"self-documenting". Looking at
all by itself doesn't tell me much, other than you are interested in
something structural at that level. Compare this to
//div[@type='letter' or @type='diary' or @type='lettergroup' or
which may be "ugly", but reveals more about the intention behind the
// Gregory Murphy. Isopaleocopria.